WO1996036933A9 - Automatic currency processing system - Google Patents

Automatic currency processing system

Info

Publication number
WO1996036933A9
WO1996036933A9 PCT/US1996/006257 US9606257W WO9636933A9 WO 1996036933 A9 WO1996036933 A9 WO 1996036933A9 US 9606257 W US9606257 W US 9606257W WO 9636933 A9 WO9636933 A9 WO 9636933A9
Authority
WO
WIPO (PCT)
Prior art keywords
bill
bills
pattems
die
scanned
Prior art date
Application number
PCT/US1996/006257
Other languages
French (fr)
Other versions
WO1996036933A1 (en
Filing date
Publication date
Application filed filed Critical
Priority to EP96915489A priority Critical patent/EP0830653A4/en
Priority to AU57251/96A priority patent/AU5725196A/en
Publication of WO1996036933A1 publication Critical patent/WO1996036933A1/en
Publication of WO1996036933A9 publication Critical patent/WO1996036933A9/en

Links

Definitions

  • United States patent application Serial No. 08/340,031 filed November 14, 1994, for a "Method And Apparatus For Discriminating and Counting Documents” is a continuation-in-part of pending United States patent applications Serial No. 08/243,807 and Serial No. 08/207,592.
  • United States patent application Serial No. 08/287,882 filed August 9, 1994 for a "Method and Apparatus for Document Identification” is a continuation-in-part of pending United States patent applications Serial No. 08/207,592, Serial No. 08/219,093, and Serial No. 08/127,334 filed September 27, 1993, for a "Method and Apparatus for Currency Discrimination and Counting.”
  • United States patent application Serial No. 08/243,807 filed May 16, 1994, for "Method And Apparatus For Currency Discrimination” is a continuation-in-part of pending United States patent applications Serial No. 08/219,093 and Serial No. 08/127,334.
  • the present invention relates to currency processing systems such as automatic teller machines and currency redemption machines. Summary Of The Invention
  • the primary object of the present invention is to provide an improved automatic teller machine ("ATM”) or currency redemption machine that is capable of processing cash deposits as well as withdrawals.
  • ATM automatic teller machine
  • Another object of this invention is to provide such machines that are capable of accepting and dispensing coins as well as bills.
  • a currency processing machine for receiving and dispensing cash and substantially immediately furnishing an associated cash accounting system with data, including the value of the currency processed, for each transaction.
  • the machine includes a bill dispenser having a bill storage device and controllable transport means for dispensing selected numbers of bills from the storage device, a bill receptacle for receiving stacks of bills to be deposited, and a bill counter and scanner for rapidly removing the bills one at a time from the receptacle and counting the bills while determining the denomination of each bill.
  • the counter and scanner also generates data representing the denomination of each bill, and the number of bills of each denomination, passed through the counter and scanner.
  • a memory receives and stores data representing the number of bills of each denomination passed through the counter and scanner in each transaction, and data representing the total value of the bills passed through the counter and scanner in each transaction.
  • a control system transfers data from the memory to an associated cash accounting system so that the deposits and withdrawals executed at the currency processing machine are entered in the accounting system substantially immediately after the execution of those transactions.
  • the preferred control system checks the genuineness of each bill and coin that is counted, and produces a control signal in response to the detection of a non-genuine bill or coin. The processing of the bill or coin detected to be non- genuine is altered in response to such control signals.
  • FIG. la is a perspective view of an automatic teller machine embodying the present invention.
  • FIG. lb is a diagrammatic side elevation of the machine of FIG. la
  • FIG. lc is a more detailed diagrammatic side elevation of the machine of FIG. la
  • FIG. ld is a flow chart illustrating the sequential procedure involved in the execution of a transaction in the machine of FIG. la;
  • FIG. le is a flow chart illustrating the sequential procedure involved in the execution of a deposit of bills in the machine of FIG. la;
  • FIG. lf is a flow chart illustrating an altemative sequential procedure involved in the execution of a deposit of bills in the machine of FIG. la;
  • FIG. 2a is a functional block diagram of the currency scanning and counting subassembly in the machine of FIG. 1, including a scanhead arranged on each side of a transport path;
  • FIG. 2b is a functional block diagram of a currency scanning and counting device that includes a scanhead arranged on a single side of a transport path;
  • FIG. 2c is a functional block diagram of a currency scanning and counting machine similar to that of FIG. 2b, but adapted to feed and scan bills along their wide dimension;
  • FIG. 2d is a functional block diagram of a currency scanning and counting device similar to those of FIGs. 2a-2c but including a second type of scanhead for detecting a second characteristic of the currency;
  • FIG. 3 is a diagrammatic perspective illustration of the successive areas scanned during the traversing movement of a single bill across an optical sensor according to a preferred embodiment of the primary scanhead;
  • FIGs. 4a and 4b are perspective views of a bill and a preferred area to be optically scanned on the bill;
  • FIGs. 5a and 5b are diagrammatic side elevation views of the preferred areas to be optically scanned on a bill according to a preferred embodiment of the invention;
  • FIG. 6a is a perspective view of a bill showing the preferred area of a first surface to be scanned by one of the two scanheads employed in the preferred embodiment of the present invention
  • FIG. 6b is another perspective view of the bill in FIG. 6a showing the preferred area of a second surface to be scanned by the other of the scanheads employed in the preferred embodiment of the present invention
  • FIG. 6c is a side elevation showing the first surface of a bill scanned by an upper scanhead and the second surface of the bill scanned by a lower scanhead;
  • FIG. 6d is a side elevation showing the first surface of a bill scanned by a lower scanhead and the second surface of the bill scanned by an upper scanhead;
  • FIGs. 7a and 7b form a block diagram illustrating a preferred circuit arrangement for processing and correlating reflectance data according to the optical sensing and counting technique of this invention;
  • FIGs. 8a and 8b comprise a flowchart illustrating the sequence of operations involved in implementing a discrimination and authentication system according to a preferred embodiment of the present invention
  • FIG. 9 is a flow chart illustrating the sequential procedure involved in detecting the presence of a bill adjacent the lower scanhead and the borderline on the side of the bill adjacent to the lower scanhead;
  • FIG. 10 is a flow chart illustrating the sequential procedure involved in detecting the presence of a bill adjacent the upper scanhead and the borderline on the side of the bill adjacent to the upper scanhead;
  • FIG. Ila is a flow chart illustrating the sequential procedure involved in the analog-to-digital conversion routine associated with the lower scanhead
  • FIG. lib is a flow chart illustrating the sequential procedure involved in the analog-to-digital conversion routine associated with the upper scanhead;
  • FIG. 12 is a flow chart illustrating the sequential procedure involved in determining which scanhead is scanning the green side of a U.S. currency bill;
  • FIG. 13 is a flow chart illustrating the sequence of operations involved in dete ⁇ nining the bill denomination from the correlation results;
  • FIG. 14 is a flow chart illustrating the sequential procedure involved in decelerating and stopping the bill transport system in the event of an error
  • FIG. 15a is a graphical illustration of representative characteristic pattems generated by narrow dimension optical scanning of a $1 currency bill in the forward direction
  • FIG. 15b is a graphical illustration of representative characteristic pattems generated by narrow dimension optical scanning of a $2 currency bill in the reverse direction;
  • FIG. 15c is a graphical illustration of representative characteristic pattems generated by narrow dimension optical scanning of a $100 currency bill in the forward direction;
  • FIG. 15d is a graph illustrating component pattems generated by scanning old and new $20 bills according a second method according to a preferred embodiment of the present invention.
  • FIG. 15e is a graph illustrating an pattem for a $20 bill scanned in the forward direction derived by averaging the pattems of FIG. 15d according a second method according to a preferred embodiment of the present invention
  • FIGs. 16a-16e are graphical illustrations of the effect produced on correlation pattem by using the progressive shifting technique, according to an embodiment of this invention
  • FIGs. 17a-17c are a flowchart illustrating a preferred embodiment of a modified pattem generation method according to the present invention
  • FIG. 18a is a flow chart illustrating the sequential procedure involved in the execution of multiple correlations of the scan data from a single bill
  • FIG. 18b is a flow chart illustrating a modified sequential procedure of that of FIG. 18a;
  • FIG. 19a is a flow chart illustrating the sequence of operations involved in determining the bill denomination from the correlation results using data retrieved from the green side of U.S. bills according to one prefened embodiment of the present invention
  • FIGs. 19b and 19c are a flow chart illustrating the sequence of operations involved in determining the bill denomination from the conelation results using data retrieved from the black side of U.S. bills;
  • FIG. 20a is an enlarged vertical section taken approximately through the center of the machine, but showing the various transport rolls in side elevation;
  • FIG. 20b is a top plan view of the interior mechanism of the machine of FIG. 1 for transporting bills across the optical scanheads, and also showing the stacking wheels at the front of the machine;
  • FIG. 21a is an enlarged perspective view of the bill transport mechanism which receives bills from the stripping wheels in the machine of FIG. 1;
  • FIG. 21b is a cross-sectional view of the bill transport mechanism depicted in FIG. 21 along line 21b;
  • FIG. 22 is a side elevation of the machine of FIG. 1, with the side panel of the housing removed;
  • FIG. 23 is an enlarged bottom plan view of the lower support member in the machine of FIG. 1 and the passive transport rolls mounted on that member;
  • FIG. 24 is a sectional view taken across the center of the bottom support member of FIG. 23 across the narrow dimension thereof;
  • FIG. 25 is an end elevation of the upper support member which includes the upper scanhead in the machine of FIG. 1, and the sectional view of the lower support member mounted beneath the upper support member;
  • FIG. 26 is a section taken through the centers of both the upper and lower support members, along the long dimension of the lower support member shown in FIG. 23;
  • FIG. 27 is a top plan view of the upper support member which includes the upper scanhead;
  • FIG. 28 is a bottom plan view of the upper support member which includes the upper scanhead
  • FIG. 29 is an illustration of the light distribution produced about one of the optical scanheads
  • FIGs. 30a and 30b are diagrammatic illustrations of the location of two auxiliary photo sensors relative to a bill passed thereover by the transport and scanning mechanism shown in FIGs. 20a-28;
  • FIG. 31 is a flow chart illustrating the sequential procedure involved in a ramp-up routine for increasing the transport speed of the bill transport mechanism from zero to top speed;
  • FIG. 32 is a flow chart illustrating the sequential procedure involved in a ramp-to-slow-speed routine for decreasing the transport speed of the bill transport mechanism from top speed to slow speed;
  • FIG. 33 is a flow chart illustrating the sequential procedure involved in a ramp-to-zero-speed routine for decreasing the transport speed of the bill transport mechanism to zero;
  • FIG. 34 is a flow chart illustrating the sequential procedure involved in a pause-after-ramp routine for delaying the feedback loop while the bill transport mechanism changes speeds;
  • FIG. 35 is a flow chart illustrating the sequential procedure involved in a feedback loop routine for monitoring and stabilizing the transport speed of the bill transport mechanism
  • FIG. 36 is a flow chart illustrating the sequential procedure involved in a doubles detection routine for detecting overlapped bills
  • FIG. 37 is a flow chart illustrating the sequential procedure involved in a routine for detecting sample data representing dark blemishes on a bill
  • FIG. 38 is a flow chart illustrating the sequential procedure involved in a routine for maintaining a desired readhead voltage level
  • FIG. 39 is a top view of a bill and size determining sensors according to a prefened embodiment of the present invention.
  • FIG. 40 is a top view of a bill illustrating multiple areas to be optically scanned on a bill according to a prefened embodiment of the present invention
  • FIG. 41a is a graph illustrating a scanned patte which is offset from a conesponding master pattem
  • FIG. 41b is a graph illustrating the same pattems of FIG. 41a after the scanned patte is shifted relative to the master pattem;
  • FIG. 42 is a side elevation of a multiple scanhead anangement according to a prefened embodiment of the present invention.
  • FIG. 43 is a side elevation of a multiple scanhead arrangement according to another prefened embodiment of the present invention
  • FIG. 44 is a side elevation of a multiple scanhead arrangement according to another prefened embodiment of the present invention.
  • FIG. 45 is a side elevation of a multiple scanhead arrangement according to another prefened embodiment of the present invention.
  • FIG. 46 is a top view of a staggered scanhead arrangement according to a prefened embodiment of the present invention.
  • FIG. 47a is a top view of a linear array scanhead according to a prefened embodiment of the present invention illustrating a bill being fed in a centered fashion;
  • FIG. 47b is a side view of a linear array scanhead according to a prefened embodiment of the present invention illustrating a bill being fed in a centered fashion
  • FIG. 48 is a top view of a linear array scanhead according to another prefened embodiment of the present invention illustrating a bill being fed in a non- centered fashion
  • FIG. 49 is a top view of a linear array scanhead according to another prefened embodiment of the present invention illustrating a bill being fed in a skewed fashion;
  • FIGs. 50a and 50b are a flowchart of the operation of a currency discrimination system according to a preferred embodiment of the present invention.
  • FIG. 51 is a top view of a triple scanhead arrangement utilized in a discriminating device able to discriminate both Canadian and German bills according to a prefened embodiment of the present invention
  • FIG. 52 is a top view of Canadian bill illustrating the areas scanned by the triple scanhead arrangement of FIG. 51 according to a prefened embodiment of the present invention
  • FIG. 53 is a flowchart of the threshold tests utilized in calling the denomination of a Canadian bill according to a prefened embodiment of the present invention
  • FIG. 54a illustrates the general areas scanned in generating master 10 DM German pattems according to a prefened embodiment of the present invention
  • FIG. 54b illustrates the general areas scanned in generating master 20 DM, 50 DM, and 100 DM German pattems according to a prefened embodiment of the present invention
  • FIG. 55 is a flowchart of the threshold tests utilized in calling the denomination of a German bill
  • FIG. 56 is a functional block diagram illustrating a first embodiment of a document authenticator and discriminator
  • FIG. 57 is a functional block diagram illustrating a second embodiment of a document authenticator and discriminator
  • FIG. 58a is a side view of a document authenticating system utilizing ultraviolet light
  • FIG. 58b is a top view of the system of FIG. 58a along the direction 58b
  • FIG. 58c is a top view of the system of FIG. 58a along the direction 58c
  • FIG. 59 is a functional block diagram of the optical and electronic components of the document authenticating system of FIGS. 58a-58c.
  • FIG. 60 is perspective view of a disc-type coin sorter embodying the present invention, with a top portion thereof broken away to show intemal structure;
  • FIG. 61 is an enlarged horizontal section taken generally along line 61-61 in
  • FIG. 60 is a diagrammatic representation of FIG. 60.
  • FIG. 62 is an enlarged section taken generally along line 62-62 in FIG. 61, showing the coins in full elevation;
  • FIG. 63 is an enlarged section taken generally along line 63-63 in FIG. 61, showing in full elevation a nickel registered with an ejection recess;
  • FIG. 64 is a diagrammatic cross-section of a coin and an improved coin discrimination sensor embodying the invention.
  • FIG. 65 is a schematic circuit diagram of the coin discrimination sensor of FIG. 64;
  • FIG. 66 is a diagrammatic perspective view of the coils in the coin discrimination sensor of FIG. 64;
  • FIG. 67a is a circuit diagram of a detector circuit for use with the discrirnination sensor of this invention;
  • FIG. 67b is a waveform diagram of the input signals supplied to the circuit of FIG. 67a;
  • FIG. 68 is a perspective view of an outboard shunting device embodying the present invention.
  • FIG. 69 is a section taken generally along line 69-69 in FIG. 68
  • FIG. 70 is a section taken generally along line 70-70 in FIG. 68, showing a movable partition in a nondiverting position
  • FIG. 71 is the same section illustrated in FIG. 70, showing the movable portion in a diverting position.
  • FIGs. la, lb and lc there is shown an automatic teller machine ("ATM") having a bill deposit receptacle 1 as well as a bill withdrawal or return slot 2.
  • ATM has the conventional slot 3 for receiving the customer's identification card so that the data on the card can be automatically read by a card reader.
  • a video display 4 provides the customer with a menu of options, and also prompts the customer to carry out the various actions required to execute a transaction, including the use of a keypad 5.
  • the illustrative ATM also has a coin deposit receptacle 6 and a coin return pocket 7.
  • the deposit receptacles 1 and 6 are normally retracted within the machine but are advanced to their open positions (shown in FIG. la) when a customer initiates a transaction and selects a "cash deposit” mode of operation. Bills and coins can then be deposited by the customer into the deposit receptacles 1 and 6, respectively. After the customer has placed a stack of bills into the receptacle 1, the customer is prompted to push that receptacle into the machine, to its retracted position.
  • This inward movement of the receptacle 1 positions the stack of bills at the feed station of a bill scanning and counting module 8 which automatically feeds, counts, scans and authenticates the bills one at a time at a high speed (e.g., at least 350 bills per minute).
  • the bills that are recognized by the scanning module 8 are delivered to a conventional currency canister 9 (FIG. lc) which is periodically removed from the machine and replaced with an empty canister.
  • a diverter 10 is actuated to divert the unidentified bill to the retum slot 2 so that it can be removed from the machine by the customer.
  • unrecognizable bills can be diverted to a separate cunency canister rather than being returned to the customer. Bills that are detected to be counterfeit are treated in the same manner as unrecognizable bills.
  • the bill transport system may also include an escrow holding area where the bills being processed in a pending deposit transaction are held until the transaction is complete. Then if the declared balance entered by the customer does not agree with the amount verified by the machine, the entire stack of bills can be returned to the customer. If desired, this decision can be controlled by the customer via the keypad.
  • the customer When coins are deposited by the customer in the receptacle 6, the customer again is prompted to push that receptacle into the machine. This causes the coins to be fed by gravity into the receiving hopper of a coin-sorting and counting module 11 which physically separates the coins by size (denomination) while separately counting the number of coins of each denomination in each separate transaction.
  • the module 11 also includes a coin discriminator which detects coins that are counterfeit or otherwise non-genuine. These unacceptable coins are discharged from the sorter at a common exit, and the coins from that exit are guided by a tube 12 to the coin return slot 7.
  • the ATM also preferably includes a conventional loose currency dispensing module 13 for dispensing loose bills, and/or a strapped cunency dispensing module 14 for dispensing strapped cunency, into a receptacle 15 at the front of the machine, in response to a withdrawal transaction.
  • a loose coin dispensing module 16 and/or a rolled coin dispensing module 17 may also be included for dispensing coins via the coin return pocket 7.
  • Additional modules that may be included in the ATM or a redemption machine using the same system are modules for verifying and accepting checks, food stamps, tokens and/or tickets containing bar codes.
  • each of the modules 8 and 11 accumulates data representing both the number and the value of each separate cunency item processed by these modules in each separate transaction.
  • this data and the account number for the transaction are downloaded to an associated cash accounting system by a modem link, so that the customer's account can be immediately adjusted to reflect both the deposits and the withdrawals effected by the cunent transaction.
  • the data from the cu ⁇ ency- processing modules and the card reader can be temporarily stored within a temporary memory within the ATM, so that the data can be downloaded at intervals controlled by the computing system on which the cash accounting system is n.
  • FIG. ld is a flow chart of a subroutine for transferring data from the ATM to the cash accounting system.
  • This subroutine is entered at step 10a each time a customer inserts an identification card into the ATM.
  • the customer's account number is stored at step 10b, and step 10c then initiates a transaction by prompting the customer to select from a menu of available deposit or withdrawal transactions, and step lOd then monitors the ATM system to determine when the transaction is complete.
  • the bill deposit amount B ⁇ , the bill withdrawal amount B w , the coin deposit amount C ⁇ j , and the coin withdrawal amount Cw are stored at steps lOe, lOf, lOg and lOh, and then downloaded to the cash accounting system at step IOi. If desired, these amounts may be loaded into a buffer memory for later retrieval by the computer that controls the cash accounting system. The cash accounting system then enters these amounts in the customer's account, and immediately adjusts the balance in that account accordingly.
  • FIG. le A subroutine for executing a cash deposit of bills is shown in FIG. le.
  • the video display prompts the customer to place the stack of bills being deposited into the receptacle 1 and to push that receptacle into the machine.
  • the bill counting and scanning module then automatically withdraws one bill at a time from the bottom of the stack, and scans each bill for denomination and authentication.
  • step 1 la Each successive bill that is withdrawn from the deposit stack is scanned at step 1 la to determine the denomination of the bill, and checked for authentication at step lib. The results of the authentication are checked at step l ie. If the bill cannot be authenticated, it is a counterfeit suspect and thus step 1 lc produces an affirmative answer. This advances the system to step lid, which determines whether the owner of the ATM or redemption machine has opted to retum counterfeit-suspect bills to the customer. If this option has been selected, the suspect bill is returned to the customer at step lie. If the return option has not been selected at step lid, the resulting negative response advances the system to step 1 lf which transports the bill to a suspect bill canister.
  • step l ie advances the system to step 1 lg to check the results of the scanning step.
  • This step determines whether the bill is a "no call,” i.e. , whether it was impossible for the scanning operation to determine the denomination of the bill. If the bill is a "no call,” step l lg produces an affirmative answer, and step l lh determines whether the option to retum "no calls" to the customer has been selected. If the answer is affirmative, the "no call” bill is returned to the customer at step lie. If the answer is negative, the "no call” bill is transported to a "no call” canister at step Hi.
  • the resulting negative response at step 1 lg causes the counter for that particular denomination to be incremented at step llj.
  • the dollar value of that denomination is then added to the verified deposit amount at step I lk to maintain a cunent cumulative total of the cunency deposit that is being processed.
  • the bill is then transported to an escrow holding area for the cunent deposit, at step 111.
  • step 11m determines when the last bill in a deposited stack of bills has been counted. When this step produces an affirmative answer, step 11m then determines whether the final verified deposit amount agrees with the declared balance that was entered by the customer through the key pad. If the answer is affirmative, the deposited bills are transported from the escrow holding area to a verified deposit canister at step l lo. A negative answer at step lln advances the system to step l ip where again the system determines whether a "retum" option has been selected. This option may be preselected by the owner of the ATM or redemption machine, or it may be an option that is available to the customer.
  • the bills are returned to the customer at step llq to enable the customer to determine why the verified deposit amount does not agree with the customer's declared balance.
  • the verified deposit amount is displayed to the customer along with an appropriate message.
  • a negative response at step lip causes the bills to be transported from the escrow holding area to a disputed balance canister at step llr.
  • FIG. lf illustrates a modification of the routine of FIG. le which permits the use of a single storage canister for all of the bills, regardless of whether they are verified bills, no calls, or counterfeit suspects.
  • the various bills are identified within the single canister by placing different colored markers on top of different bills.
  • a first marker e.g., a marker of a first color
  • a second type of marker is inserted at step lit in response to an affirmative response at step 1 lg and a negative answer at step 1 lh, to indicate that the marked bill is a counterfeit suspect.
  • a third type of marker e.g., of a third color, is inserted at step llu in response to negative answers at steps lln and lip, to indicate that the marked batch of bills represents a deposit whose verified amount did not agree with the customer's declared balance. Because this third type of marker identifies a batch of bills instead of a single bill, it is necessary to insert a marker at both the beginning and end of the marked batch.
  • the customer may key in the value and number of such bills and deposit them in an envelope for later verification by the bank.
  • a message on the display screen may advise the customer of this option. For example, if four $10 bills are returned, and then re-deposited by the customer in an envelope, fhe customer may press a "$10" key four times. The customer then receives immediate credit for all the bills denominated and authenticated by the scanner. Credit for the re-deposited "no call” bills is given only after the bank picks up the deposit envelope and manually verifies the amount. Altematively, at least prefened customers can be given full credit immediately, subject to later verification, or immediate credit can be given up to a certain dollar limit. In the case of counterfeit bills that are not retumed to the customer, the customer can be notified of the detection of a counterfeit suspect at the ATM or later by a written notice or personal call, depending upon the preferences of the financial institution.
  • the ATM or redemption machine may also have a "verify mode” in which it simply denominates and totals all the cunency (bills and/or coins) deposited by the customer and retums it all to the customer. If the customer agrees with the amount and wishes to proceed with an actual deposit, the customer selects the "deposit mode” and re-deposits the same batch of cunency in the machine. Altematively, the "verify mode" may hold the initially deposited cunency in an escrow area until the customer decides whether to proceed with an actual deposit.
  • the message display screen advises the customer of the number and value of the cunency items processed prior to the jam.
  • the customer is instructed to retrieve the cunency not yet processed and to manually deposit it in a sealed envelope which is then deposited into the machine for subsequent verification.
  • the machine malfunction is automatically reported via modem to the home office.
  • FIG. 2a there is shown a prefened embodiment of a cunency scanning and counting module 8.
  • the module 8 includes a bill accepting station 12 for receiving stacks of cunency bills from the deposit receptacle 1.
  • a feed mechanism functions to pick out or separate one bill at a time for transfer to a bill transport mechanism 16 (FIG.
  • each scanhead 18a, 18b is an optical scanhead that scans for characteristic information from a scanned bill 17 which is used to identify the denomination of the bill.
  • the scanned bill 17 is then transported to a cassette or bill stacking station 20 where bills so processed are stacked for subsequent removal.
  • Each optical scanhead 18a, 18b preferably comprises a pair of light sources 22 directing light onto the bill transport path so as to illuminate a substantially rectangular light strip 24 upon a cunency bill 17 positioned on the transport path adjacent the scanhead 18.
  • Light reflected off the illuminated strip 24 is sensed by a photodetector 26 positioned between the two light sources.
  • the analog output of the photodetector 26 is converted into a digital signal by means of an analog-to-digital (ADC) convertor unit 28 whose output is fed as a digital input to a central processing unit (CPU) 30.
  • ADC analog-to-digital
  • the scanheads 18a, 18b of FIG. 2a are optical scanheads, it should be understood that the scanheads and the signal processing system may be designed to detect a variety of characteristic information from cunency bills. Additionally, the scanheads may employ a variety of detection means such as magnetic, optical, electrical conductivity, and capacitive sensors. Use of such sensors is discussed in more detail below (see, e.g., FIG. 2d).
  • the bill transport path is defined in such a way that the transport mechanism 16 moves cunency bills with the na ⁇ ow dimension of the bills being parallel to the transport path and the scan direction.
  • the system may be designed to scan bills along their long dimension or along a skewed dimension.
  • the coherent light strip 24 effectively scans the bill across the na ⁇ ow dimension of the bill.
  • the transport path is so arranged that a cunency bill 17 is scanned across a central section of the bill along its na ⁇ ow dimension, as shown in FIG. 2a.
  • Each scanhead functions to detect light reflected from the bill as it moves across the illuminated light strip 24 and to provide an analog representation of the variation in reflected light, which, in turn, represents the variation in the dark and light content of the printed pattem or indicia on the surface of the bill.
  • This variation in light reflected from the na ⁇ ow-dimension scanning of the bills serves as a measure for distinguishing, with a high degree of confidence, among a plurality of cunency denominations which the system is programmed to handle.
  • a series of such detected reflectance signals are obtained across the na ⁇ ow dimension of the bill, or across a selected segment thereof, and the resulting analog signals are digitized under control of the CPU 30 to yield a fixed number of digital reflectance data samples.
  • the data samples are then subjected to a normalizing routine for processing the sampled data for improved conelation and for smithing out variations due to "contrast" fluctuations in the printed pattem existing on the bill surface.
  • the normalized reflectance data represents a characteristic pattem that is unique for a given bill denomination and provides sufficient distinguishing features among characteristic pattems for different cunency denominations.
  • the reflectance sampling process is preferably controlled through the CPU 30 by means of an optical encoder 32 which is linked to the bill transport mechanism 16 and precisely tracks the physical movement of the bill 17 between the scanheads 18a, 18b.
  • the optical encoder 32 is linked to the rotary motion of the drive motor which generates the movement imparted to the bill along the transport path.
  • the mechanics of the feed mechanism ensure that positive contact is maintained between the bill and the transport path, particularly when the bill is being scanned by the scanheads. Under these conditions, the optical encoder 32 is capable of precisely tracking the movement of the bill 17 relative to the light strips 24 generated by the scanheads 18a, 18b by monitoring the rotary motion of the drive motor.
  • the outputs of the photodetectors 26 are monitored by the CPU 30 to initially detect the presence of the bill adjacent the scanheads and, subsequently, to detect the starting point of the printed patte on the bill, as represented by the thin borderline 17a which typically encloses the printed indicia on U.S. cunency bills.
  • the optical encoder 32 is used to control the timing and number of reflectance samples that are obtained from the outputs of the photodetectors 26 as the bill 17 moves across the scanheads.
  • FIG. 2b illustrates a modified cunency scanning and counting device similar to that of FIG. 2a but having a scanhead on only a single side of the transport path.
  • FIG. 2c illustrates anodier modified cunency scanning and counting device similar to that of FIG. 2b but illustrating feeding and scanning of bills along their wide direction.
  • FIGs. 2b-2c the transport mechanism 16 moves cunency bills with a preselected one of their two dimensions (na ⁇ ow or wide) being parallel to the transport path and the scan direction.
  • FIGs. 2b and 4a illustrate bills oriented with their na ⁇ ow dimension "W" parallel to the direction of movement and scanning
  • FIGs. 2c and 4b illustrate bills oriented with their wide dimension "L” parallel to the direction of movement and scanning.
  • FIG. 2d there is shown a functional block diagram illustrating a prefe ⁇ ed embodiment of a cunency discriminating and authenticating system. The operation of the system of FIG. 2d is the same as that of FIG. 2a except as modified below.
  • the system includes a bill accepting station 12 where stacks of cunency bills that need to be identified, authenticated, and counted are positioned. Accepted bills are acted upon by a bill separating station 14 which functions to pick out or separate one bill at a time for transfer to a bill transport mechanism 16 which transports each bill along a precisely predetermined transport path, across two scanheads 18 and 39 where the cunency denomination of the bill is identified and the genuineness of the bill is authenticated.
  • scanhead 18 is an optical scanhead that scans for a first type of characteristic information from a scanned bill 17 which is used to identify the bill's denomination.
  • a second scanhead 39 scans for a second type of characteristic information from the scanned bill 17.
  • While the illustrated scanheads 18 and 39 are separate and distinct, they may be incorporated into a single scanhead.
  • the first characteristic sensed is intensity of reflected light and the second characteristic sensed is color
  • a single optical scanhead having a plurality of detectors, one or more without filters and one or more with colored filters, may be employed (U.S. Pat. No. 4,992, 860 incorporated herein by reference).
  • the scanned bill is then transported to a bill stacking station 20 where bills so processed are stacked for subsequent removal.
  • the optical scanhead 18 of the embodiment depicted in FIG. 2d comprises at least one light source 22 directing a beam of coherent light downwardly onto the bill transport path so as to illuminate a substantially rectangular light strip 24 upon a cunency bill 17 positioned on the transport path below the scanhead 18.
  • Light reflected off die iUurninated strip 24 is sensed by a photodetector 26 positioned directly above the strip.
  • the analog output of photodetector 26 is converted into a digital signal by means of an analog-to-digital (ADC) convertor unit 28 whose output is fed as a digital input to a central processing unit (CPU) 30.
  • ADC analog-to-digital
  • the second scanhead 39 comprises at least one detector 41 for sensing a second type of characteristic info ⁇ nation from a bill.
  • the analog output of the detector 41 is converted into a digital signal by means of a second analog-to-digital converter 43 whose output is also fed as a digital input to the central processing unit (CPU) 30.
  • CPU central processing unit
  • the scanhead 18 in the embodiment of FIG. 2d is an optical scanhead
  • the first and second scanheads 18 and 39 may be designed to detect a variety of characteristic information from cunency bills. Additionally these scanheads may employ a variety of detection means such as magnetic or optical sensors. For example, a variety of cunency characteristics can be measured using magnetic sensing. These include detection of pattems of changes in magnetic flux (U.S. Pat. No. 3,280,974), pattems of vertical grid lines in the portrait area of bills (U.S. Pat. No. 3,870,629), the presence of a security thread (U.S. Pat. No. 5,151,607), total amount of magnetizable material of a bill (U.S. Pat. No.
  • Color detection techniques may employ color filters, colored lamps, and/or dichroic beamsplitters (U.S. Pat. Nos. 4,841,358; 4,658,289; 4,716,456; 4,825,246, 4,992,860 and EP 325,364). Prescribed hues or intensities of a given color may be detected.
  • Reflection and/or fluorescence of ultraviolet light may also be used, as described in detail below. Absorption of infrared light may also be used as an authenticating technique.
  • the modified pattem generation method discussed below is useful in discrimination systems designed to accommodate bills other than U.S. cunency because many non-U.S. bills lack a borderline around the printed indicia on their bills.
  • the modified patte generation method may be important in discrimination systems designed to accommodate bills other than U.S.
  • the printed indicia of many non- U.S. bills lack sharply defined edges which in turns inhibits using the edge of the printed indicia of a bill as a trigger for the initiation of the scanning process and instead promotes reliance on using the edge of the bill itself as the trigger for the initiation of the scanning- process.
  • optical encoder 32 for controlling the sampling process relative to the physical movement of a bill 17 across the scanheads 18a, 18b is also advantageous in that the encoder 32 can be used to provide a predetermined delay following detection of the borderline 17a prior to initiation of samples.
  • the encoder delay can be adjusted in such a way that the bill 17 is scanned only across those segments which contain the most distinguishable printed indicia relative to the different cunency denominations.
  • the optical encoder can be used to control the scanning process so that reflectance samples are taken for a set period of time and only after a certain period of time has elapsed after the borderline 17a is detected, thereby restricting the scanning to the desired central portion of the na ⁇ ow dimension of the bill.
  • FIGs. 3-5b illustrate the scanning process in more detail.
  • FIG. 4a as a bill 17 is advanced in a direction parallel to the na ⁇ ow edges of the bill, scanning via a slit in the scanhead 18a or 18b is effected along a segment S of the central portion of the bill 17.
  • This segment S begins a fixed distance D inboard of the borderline 17a.
  • the photodetector 26 produces a continuous output signal which is proportional to the intensity of the light reflected from the illuminated strip s at any given instant.
  • This output is sampled at intervals controlled by the encoder, so that the sampling intervals are precisely synchronized with the movement of the bill across the scanhead.
  • FIG. 4b is similar to FIG. 4a but illustrates scanning along the wide dimension of the bill 17.
  • the sampling intervals be selected so that the strips s that are illuminated for successive samples overlap one another.
  • the odd-numbered and even-numbered sample strips have been separated in FIGs. 3, 5a, and 5b to more clearly illustrate this overlap.
  • the first and second strips si and s2 overlap each other
  • the second and third strips s2 and s3 overlap each other, and so on.
  • Each adjacent pair of strips overlap each other. In the illustrative example, this is accomplished by sampling strips that are 0.050 inch (0J27 cm) wide at 0.029 inch (0.074 cm) intervals, along a segment S that is 1.83 inch (4.65 cm) long (64 samples).
  • FIGs. 6a and 6b illustrate two opposing surfaces of U.S. bills.
  • the printed pattems on the black and green surfaces of the bill are each enclosed by respective thin borderlines B j and B 9 .
  • scanning via the wide slit of one of the scanheads is effected along a segment S A of the central portion of die black surface of the bill (FIG. 6a).
  • the orientation of the bill along the transport path determines whether the upper or lower scanhead scans the black surface of the bill.
  • This segment S A begins a fixed distance D ⁇ inboard of die borderline B ⁇ , which is located a distance W j from the edge of the bill.
  • the scanning along segment S A is as described in connection with FIGs. 3, 4a, and 5a.
  • the other of the two scanheads scans a segment Sg of die central portion of the green surface of the bill (FIG. 6b).
  • the orientation of the bill along the transport path determines whether the upper or lower scanhead scans the green surface of the bill.
  • This segment Sg begins a fixed distance D 9 inboard of the border line B2, which is located a distance W2 from the edge of the bill. For U.S. cunency, the distance W2 on die green surface is greater than the distance W j on the black surface. It is this feature of U.S.
  • FIGs. 6c and 6d are side elevations of FIG. 2a.
  • FIG. 6c shows the first surface of a bill scanned by an upper scanhead and the second surface of the bill scanned by a lower scanhead
  • FIG. 6d shows the first surface of a bill scanned by a lower scanhead and the second surface of the bill scanned by an upper scanhead.
  • FIGs. 6c and 6d illustrate the pair of optical scanheads 18a, 18b disposed on opposite sides of the transport path to permit optical scanning of both surfaces of a bill. With respect to United States cunency, these opposing surfaces co ⁇ espond to the black and green surfaces of a bill.
  • One of the optical scanheads 18 (the "upper" scanhead 18a in FIGs. 6c-6d) is positioned above die transport path and illuminates a light strip upon a first surface of the bill, while the other of the optical scanheads 18 (the “lower” scanhead 18b in FIGs. 6c-6d) is positioned below the transport path and illuminates a light strip upon the second surface of the bill.
  • the surface of die bill scanned by each scanhead 18 is determined by the orientation of the bill relative to the scanheads 18.
  • the upper scanhead 18a is located slightiy upstream relative to the lower scanhead 18b.
  • the photodetector of the upper scanhead 18a produces a first analog output conesponding to the first surface of the bill, while die photodetector of the lower scanhead 18b produces a second analog output conesponding to the second surface of the bill.
  • the first and second analog outputs are converted into respective first and second digital outputs by means of respective analog-to-digital (ADC) convertor units 28 whose outputs are fed as digital inputs to a central processing unit (CPU) 30.
  • ADC analog-to-digital
  • CPU 30 uses the sequence of operations illustrated in FIG.
  • the master characteristic pattems are preferably generated by performing scans on the green surfaces, not black surfaces, of bills of different denominations. According to a prefe ⁇ ed embodiment, fhe analog output co ⁇ esponding to die black surface of the bill is not used for subsequent conelation.
  • the optical sensing and conelation technique is based upon using the above process to generate a series of stored intensity signal pattems using genuine bills for each denomination of cunency mat is to be detected.
  • two or four sets of master intensity signal samples are generated and stored within the system memory, preferably in the form of an EPROM 34 (see FIG. 2a), for each detectable cunency denomination.
  • the sets of master intensity signal samples for each bill are generated from optical scans, performed on the green surface of the bill and taken along both the "forward" and "reverse" directions relative to the pattem printed on the bill. Altematively, the optical scanning may be performed on the black side of U.S. cunency bills or on either surface of foreign bills. Additionally, me optical scanning may be performed on both sides of a bill. In adapting this technique to U.S.
  • cunency for example, sets of stored intensity signal samples are generated and stored for seven different denominations of U.S. cunency, i.e., $1, $2, $5, $10, $20, $50 and $100.
  • U.S. cunency For bills which produce significant pattem changes when shifted slightiy to the left or right, such as the $2, the $10 and/or the $100 bills in U.S. cunency, it is prefened to store two green-side pattems for each of the "forward" and "reverse" directions, each pair of pattems for the same direction represent two scan areas that are slightly displaced from each other along the long dimension of the bill.
  • a set of 16 [or 18] different green-side master characteristic pattems are stored within the EPROM for subsequent conelation purposes (four master pattems for the $10 bill [or four master pattems for the $10 bill and the $2 bill and/or the $100 bill] and two master pattems for each of the other denominations).
  • the generation of the master pattems is discussed in more detail below.
  • the pattem generated by scanning a bill under test is compared by fhe CPU 30 with each of the 16 [or 18] master pattems of stored intensity signal samples to generate, for each comparison, a conelation number representing the extent of conelation, i.e., similarity between conesponding ones of the plurality of data samples, for the sets of data being compared.
  • the CPU 30 is programmed to identify the denomination of the scanned bill as conesponding to e set of stored intensity signal samples for which me conelation number resulting from pattem comparison is found to be the highest.
  • a bi-level threshold of conelation is used as me basis for making a "positive" call. If a "positive" call can not be made for a scanned bill, an e ⁇ or signal is generated.
  • master pattems are also stored for selected denominations co ⁇ esponding to scans along the black side of U.S. bills. More particularly, according to a prefened embodiment, multiple black-side master pattems are stored for $20, $50 and $100 bills. For each of these denominations, three master pattems are stored for scans in the forward and reverse directions for a total of six pattems for each denomination.
  • black-side master pattems are generated by scanning a conesponding denominated bill along a segment located about the center of the na ⁇ ow dimension of the bill, a segment slightly displaced (0.2 inches) to the left of center, and a segment slightiy displaced (0.2 inches) to the right of center.
  • me scanned pattem generated from me green side of a test bill fails to sufficientiy conelate with one of the green-side master pattems
  • me scanned pattem generated from the black side of a test bill is then compared to black-side master pattems in some situations as described in more detail below in conjunction with FIGs. 19a-19c.
  • the CPU 30 is programmed to count the number of bills belonging to a particular cunency denomination as part of a given set of bills that have been scanned for a given scan batch, and to determine d e aggregate total of the cunency amount represented by the bills scanned during a scan batch.
  • the CPU 30 is also linked to an output unit 36 (FIGs. 2a and FIG. 2b) which is adapted to provide a display of the number of bills counted, die breakdown of die bills in terms of cunency denomination, and die aggregate total of the cunency value represented by counted bills.
  • the output unit 36 can also be adapted to provide a print-out of the displayed information in a desired format.
  • the CPU 30 will have either determined die denomination of fhe scanned bill 17 or determined that die first scanned signal samples fail to sufficiently conelate with any of the sets of stored intensity signal samples in which case an enor is generated. Provided diat an enor has not been generated as a result of diis first comparison based on reflected light intensity characteristics, a second comparison is performed.
  • This second comparison is performed based on a second type of characteristic information, such as alternate reflected light properties, similar reflected light properties at alternate locations of a bill, light transmissivity properties, various magnetic properties of a bill, the presence of a security thread embedded within a bill, the color of a bill, the thickness or other dimension of a bill, etc.
  • the second type of characteristic info ⁇ nation is retrieved from a scanned bill by the second scanhead 39.
  • the scanning and processing by scanhead 39 may be controlled in a manner similar to that described above with regard to scanhead 18.
  • the EPROM 34 stores sets of stored second characteristic information for genuine bills of me different denominations which the system 10 is capable of handling. Based on the denomination indicated by die first comparison, the CPU 30 retrieves the set or sets of stored second characteristic data for a genuine bill of the denomination so indicated and compares the retrieved information with die scanned second characteristic information. If sufficient conelation exists between the retrieved information and the scanned information, the CPU 30 verifies the genuineness of the scanned bill 17. Otherwise, the CPU generates an e ⁇ or. While the prefe ⁇ ed embodiment illustrated in FIG.
  • the CPU 30 is programmed to count die number of bills belonging to a particular cunency denomination whose genuineness has been verified as part of a given set of bills that have been scanned for a given scan batch, and to determine the aggregate total of the cunency amount represented by the bills scanned during a scan batch.
  • FIGs. 7a and 7b there is shown a representation, in block diagram form, of a prefened circuit arrangement for processing and co ⁇ elating reflectance data according to the system of this invention.
  • the CPU 30 accepts and processes a variety of input signals including those from the optical encoder 32, the sensor 26 and the erasable programmable read only memory (EPROM) 60.
  • the EPROM 60 has stored within it die conelation program on the basis of which pattems are generated and test pattems compared with stored master programs in order to identify the denomination of test cunency.
  • a crystal 40 serves as the time base for the CPU 30, which is also provided with an extemal reference voltage " VREF 42 on the basis of which peak detection of sensed reflectance data is performed.
  • the CPU 30 also accepts a timer reset signal from a reset unit 44 which, as shown in FIG. 7b, accepts the output voltage from the photodetector 26 and compares it, by means of a threshold detector 44a, relative to a pre-set voltage threshold, typically 5.0 volts, to provide a reset signal which goes "high” when a reflectance value conesponding to the presence of paper is sensed. More specifically, reflectance sampling is based on the premise that no portion of the illuminated light strip (24 in FIG. 2a) is reflected to the photodetector in the absence of a bill positioned below the scanhead. Under these conditions, fhe output of the photodetector represents a "dark" or "zero" level reading.
  • the photodetector output changes to a "white” reading, typically set to have a value of about 5.0 volts, when the edge of a bill first becomes positioned below die scanhead and falls under the light strip 24.
  • the reset unit 44 provides a "high" signal to die CPU 30 and marks the initiation of the scanning procedure.
  • the machine-direction dimension that is, the dimension parallel to fhe direction of bill movement, of the illuminated strip of light produced by the light sources within me scanhead is set to be relatively small for the initial stage of the scan when the thin borderline is being detected, according to a prefened embodiment.
  • the use of the nanow slit increases the sensitivity with which the reflected light is detected and allows minute variations in the "gray" level reflected off the bill surface to be sensed. This ensures mat the thin borderline of fhe pattem, i.e., fhe starting point of fhe printed pattem on the bill, is accurately detected.
  • the CPU 30 processes die output of fhe sensor 26 through a peak detector 50 which essentially functions to sample the sensor output voltage and hold the highest, i.e., peak, voltage value encountered after the detector has been enabled.
  • the peak detector is also adapted to define a scaled voltage on the basis of which the printed borderline on the cunency bills is detected.
  • the output of the peak detector 50 is fed to a voltage divider 54 which lowers the peak voltage down to a scaled voltage Vg representing a predefined percentage of this peak value.
  • the voltage Vg is based upon die percentage drop in output voltage of the peak detector as it reflects the transition from the "high" reflectance value resulting from fhe scanning of the unprinted edge portions of a cunency bill to the relatively lower "gray” reflectance value resulting when the thin borderline is encountered.
  • the scaled voltage Vg is set to be about 70 - 80 percent of the peak voltage.
  • the scaled voltage Vg is supplied to a line detector 56 which is also provided with the incoming instantaneous output of die sensor 26.
  • the line detector 56 compares the two voltages at its input side and generates a signal L D £ T which normally stays “low” and goes “high” when the edge of the bill is scanned.
  • the signal L DET goes “low” when the incoming sensor output reaches the pre-defined percentage of the peak output up to that point, as represented by the voltage Vg.
  • die signal goes "low" it is an indication that the borderline of fhe bill pattem has been detected.
  • the CPU 30 initiates me actual reflectance sampling under control of the encoder 32, and the desired fixed number of reflectance samples are obtained as fhe cunency bill moves across the iUuminated light strip and is scanned along die central section of its na ⁇ ow dimension.
  • the reflectance samples resulting from fhe scanning of one or more genuine bills for each denomination are loaded into conesponding designated sections within a system memory 60, which is preferably an EPROM.
  • a system memory 60 which is preferably an EPROM.
  • the reflectance values resulting from fhe scanning of a test bill are sequentially compared, under control of die conelation program stored within the EPROM 60, with die conesponding master characteristic pattems stored within the EPROM 60.
  • a pattem averaging procedure for scanning bills and generating characteristic pattems is described below in connection with FIGs. 15a-15e. The intenelation between d e use of the first and second type of characteristic information can be seen by considering FIGs.
  • step 1748 reflected light intensity information is retrieved from a bill being scanned (step 1750).
  • second characteristic information is also retrieved from the bill being scanned (step 1752). Denomination e ⁇ or and second characteristic enor flags are cleared (steps 1753 and 1754).
  • the scanned intensity information is compared to each set of stored intensity information co ⁇ esponding to genuine bills of all denominations e system is programmed to accommodate (step 1758). For each denomination, a conelation number is calculated.
  • the system determines either the denomination of the scanned bill or generates a denomination enor by setting me denomination enor flag steps 1760 and 1762). In die case where the denomination enor flag is set (step 1762), the process is ended (step 1772). Altematively, if based on this first comparison, die system is able to determine die denomination of the scanned bill, the system proceeds to compare the scanned second characteristic information with the stored second characteristic information conesponding to die denomination determined by the first comparison (step 1764).
  • the system need not make comparisons with stored second characteristic information for die other denominations the system is programmed to accommodate. If based on this second comparison (step 1764) it is determined that the scanned second characteristic information does not sufficientiy match diat of die stored second characteristic information (step 1766), then a second characteristic e ⁇ or is generated by setting the second characteristic e ⁇ or flag (step 1768) and the process is ended (step 1772). If the second comparison results in a sufficient match between the scanned and stored second characteristic information (step 1766), then the denomination of the scanned bill is indicated (step 1770) and the process is ended (step 1772).
  • Table 1 depicts relative total magnetic content thresholds for various denominations of genuine bills. Columns 1-5 represent varying degrees of sensitivity. The values in Table 1 are set based on fhe scanning of genuine bills of varying denominations for total magnetic content and setting required thresholds based on die degree of sensitivity selected. The information in Table 1 is based on the total magnetic content of a genuine $1 being 1000. The following discussion is based on a sensitivity setting of 4. In this example it is assumed that magnetic content represents the second characteristic tested.
  • first characteristic information such as reflected light intensity
  • second characteristic information such as reflected light intensity
  • the total magnetic content of die scanned bill is compared to die total magnetic content threshold of a genuine $10 bill, i.e., 200. If the magnetic content of the scanned bill is less than 200, the bill is rejected. Otherwise it is accepted as a $10 bill.
  • FIGs. 9-1 lb there are shown flow charts illustrating the sequence of operations involved in implementing the above-described optical sensing and conelation technique.
  • FIGs. 9 and 10 illustrate fhe sequences involved in detecting the presence of a bill adjacent die scanheads and die borderlines on each side of fhe bill.
  • the lower scanhead fine line interrupt is initiated upon die detection of die fine line by the lower scanhead.
  • An encoder counter is maintained that is incremented for each encoder pulse. The encoder counter scrolls from 0 - 65,535 and then starts at 0 again.
  • die value of the encoder counter is stored in memory upon die detection of the fine line by the lower scanhead.
  • the lower scanhead fine line interrupt is disabled so that it will not be triggered again during the interrupt period.
  • a lower scanhead bit in die trigger flag is set. This bit is used to indicate that die lower scanhead has detected the fine line.
  • the magnetic sampler is initialized at step 77, and the magnetic sampling interrupt is enabled at step 78.
  • a density sampler is initialized at step 79, and a density sampling interrupt is enabled at step 80.
  • the lower read data sampler is initialized at step 81, and a lower scanhead data sampling interrupt is enabled at step 82.
  • the lower scanhead fine line interrupt flag is reset, and at step 84 the program returns from the interrupt.
  • the upper scanhead fine line interrupt is initiated upon the detection of the fine line by the upper scanhead.
  • die value of the encoder counter is stored in memory upon die detection of the fine line by the upper scanhead.
  • This information in connection with the encoder counter value associated with die detection of the fine line by the lower scanhead may then be used to determine the face orientation of a bill, mat is whether a bill is fed green side up or green side down in the case of U.S. bills, as is described in more detail below in connection with FIG. 12.
  • die upper scanhead fine line interrupt is disabled so diat it will not be triggered again during the interrupt period.
  • die upper scanhead bit in the trigger flag is set. This bit is used to indicate that the upper scanhead has detected the fine line. By checking the lower and upper scanhead bits in the trigger flag, it can be determined whether each side has detected a respective fine line.
  • FIG. Ila is a flow chart illustrating the sequential procedure involved in the analog-to-digital conversion routine associated with the lower scanhead.
  • the routine is started at step 93a.
  • the sample pointer is decremented at step 94a so as to maintain an indication of the number of samples remaining to be obtained.
  • the sample pointer provides an indication of the sample being obtained and digitized at a given time.
  • step 95a the digital data conesponding to die output of the photodetector associated with the lower scanhead for the cunent sample is read.
  • the data is converted to its final form at step 96a and stored widiin a pre-defined memory segment as Xfl v f-L at ste P 97a.
  • step 98a a check is made to see if the desired fixed number of samples "N" has been taken. If the answer is found to be negative, step 99a is accessed where the interrupt authorizing the digitization of die succeeding sample is enabled, and die program retums from interrupt at step 100a for completing fhe rest of the digitizing process.
  • a flag namely the lower scanhead done flag bit, indicating die same is set at step 101a, and the program returns from interrupt at step 102a.
  • FIG. lib is a flow chart illustrating the sequential procedure involved in the analog-to-digital conversion routine associated with the upper scanhead.
  • the routine is started at step 93b.
  • the sample pointer is decremented at step 94b so as to maintain an indication of the number of samples remaining to be obtained.
  • the sample pointer provides an indication of fhe sample being obtained and digitized at a given time.
  • the digital data conesponding to die output of the photodetector associated with the upper scanhead for the cunent sample is read.
  • the data is converted to its final form at step 96b and stored within a pre-defined memory segment as jjs f.T j at step 97b.
  • step 98b a check is made to see if die desired fixed number of samples "N" has been taken. If the answer is found to be negative, step 99b is accessed where the interrupt authorizing the digitization of die succeeding sample is enabled and the program retums from interrupt at step 100b for completing the rest of the digitizing process. However, if the answer at step 98b is found to be positive, i.e., die desired number of samples have already been obtained, a flag, namely the upper scanhead done flag bit, indicating the same is set at step 101b, and the program retums from interrupt at step 102b.
  • the CPU 30 is programmed with fhe sequence of operations in FIG. 12 to conelate at least initially only the test pattem conesponding to the green surface of a scanned bill.
  • the upper scanhead 18a is located slighdy upstream adjacent the bill transport path relative to the lower scanhead 18b.
  • the distance between the scanheads 18a, 18b in a direction parallel to the transport path co ⁇ esponds to a predetermined number of encoder counts.
  • diat die encoder 32 produces a repetitive tracking signal synchronized with incremental movements of the bill transport mechanism, and diis repetitive tracking signal has a repetitive sequence of counts (e.g., 65,535 counts) associated therewith.
  • the CPU 30 monitors the output of the upper scanhead 18a to detect die borderline of a first bill surface facing the upper scanhead 18a. Once this borderline of the first surface is detected, the CPU 30 retrieves and stores a first encoder count in memory.
  • the CPU 30 monitors the output of the lower scanhead 18b to detect the borderline of a second bill surface facing the lower scanhead 18b. Once die borderline of the second surface is detected, the CPU 30 retrieves and stores a second encoder count in memory. Referring to FIG. 12, the CPU 30 is programmed to calculate the difference between the first and second encoder counts (step 105a). If this difference is greater than the predetermined number of encoder counts co ⁇ esponding to the distance between the scanheads 18a, 18b plus some safety factor number "X", e.g., 20 (step 106), the bill is oriented with its black surface facing the upper scanhead 18a and its green surface facing the lower scanhead 18b. This can best be understood by reference to FIG.
  • the CPU 30 sets a flag to indicate that the test pattem produced by fhe lower scanhead 18b should be conelated (step 107).
  • this test pattem is conelated with the green-side master characteristic pattems stored in memory (step 109).
  • step 106 die CPU 30 is programmed to determine whedier the difference between the first and second encoder counts is less than the predetermined number minus some safety number "X", e.g., 20 (step 108). If the answer is negative, the orientation of the bill relative to fhe scanheads 18a, 18b is uncertain, so the CPU 30 is programmed to conelate die test pattems produced by both die upper and lower scanheads 18a, 18b with die green-side master characteristic pattems stored in memory (steps 109, 110, and 111).
  • the bill is oriented with its green surface facing the upper scanhead 18a and its black surface facing the lower scanhead 18b.
  • FIG. 6d shows a bill with the foregoing orientation.
  • die borderline B j must travel for a distance less than die distance between the upper and lower scanheads 18a, 18b in order to pass over the lower scanhead 18b.
  • the difference between fhe second encoder count associated with die borderline B ⁇ and the first encoder count associated with die borderline Bo should be less dian the predetermined number of encoder counts co ⁇ esponding to die distance between the scanheads 18a, 18b.
  • the CPU 30 is programmed to conelate die test pattem produced by die upper scanhead 18a widi die green-side master characteristic pattems stored in memory (step 111).
  • the CPU 30 is programmed to perform the bi-level threshold check (step 112).
  • a simple conelation procedure is utilized for processing digitized reflectance values into a form which is conveniently and accurately compared to conesponding values pre-stored in an identical format. More specifically, as a first step, the mean value X for the set of digitized reflectance samples (comparing "n" samples) obtained for a bill scan run is first obtained as below:
  • each reflectance sample is normalized by obtaining the difference between the sample and the above-calculated mean value and dividing it by the square root of die normalizing factor ⁇ as defined by die following equation:
  • die fixed number of reflectance samples which are digitized and normalized for a bill scan is selected to be 64. It has experimentally been found diat the use of higher binary orders of samples (such as 128, 256, etc.) does not provide a co ⁇ espondingly increased discrimination efficiency relative to the increased processing time involved in implementing die above-described conelation procedure. It has also been found diat fhe use of a binary order of samples lower than 64, such as 32, produces a substantial drop in discrimination efficiency.
  • the conelation factor can be represented conveniently in binary terms for ease of conelation.
  • me factor of unity which results when a hundred percent conelation exists is represented in terms of fhe binary number 2 , which is equal to a decimal value of 1024.
  • die normalized samples within a test pattem are compared to tiie master characteristic pattems stored within the system memory in order to determine me particular stored pattem to which the test pattem conesponds most closely by identifying the comparison which yields a conelation number closest to 1024.
  • a bi-level threshold of conelation is required to be satisfied before a particular call is made, for at least certain denominations of bills.
  • the conelation procedure is adapted to identify fhe two highest conelation numbers resulting from the comparison of the test pattem to one of the stored pattems.
  • a minimum threshold of conelation is required to be satisfied by these two conelation numbers. It has experimentally been found diat a conelation number of about 850 serves as a good cut-off threshold above which positive calls may be made with a high degree of confidence and below which die designation of a test pattem as co ⁇ esponding to any of the stored pattems is uncertain.
  • a second threshold level a minimum separation is prescribed between the two highest conelation numbers before making a call.
  • me minimum separation between conelation numbers is set to be 150 when the highest conelation number is between 800 and 850. When the highest conelation number is below 800, no call is made.
  • test pattems The procedure involved in comparing test pattems to master pattems is discussed below in connection with FIG. 18a.
  • Step 115 determines whether the bill has been identified as a $2 bill, and, if the answer is negative, step 116 determines whether the best conelation number ("call #1") is greater than 799. If the answer is negative, the conelation number is too low to identify fhe denomination of the bill with certainty, and thus step 117 generates a "no call” code. A "no call previous bill” flag is then set at step 118, and the routine retums to the main program at step 119.
  • step 116 advances the system to step 120, which determines whether the sample data passes an ink stain test (described below). If the answer is negative, a "no call" code is generated at step 117. If die answer is affirmative, the system advances to step 121 which determines whether the best conelation number is greater than 849. An affirmative answer at step 121 indicates diat die conelation number is sufficiently high that the denomination of the scanned bill can be identified with certainty without any further checking. Consequently, a "denomination" code identifying the denomination represented by the stored pattem resulting in the highest conelation number is generated at step 122, and me system retums to the main program at step 119.
  • a negative answer at step 121 indicates that the conelation number is between 800 and 850. It has been found diat conelation numbers within this range are sufficient to identify all bills except the $2 bill. Accordingly, a negative response at step 121 advances die system to step 123 which determines whether the difference between the two highest conelation numbers ("call #1" and "call #2") is greater dian
  • step 123 produces a negative response which advances the system to step 117 to generate a "no call" code.
  • an affirmative response at this step indicates diat the initial call is a $2 bill.
  • This affirmative response initiates a series of steps 124-127 which are identical to steps 116, 120, 121 and 123 described above, except that die numbers 799 and 849 used in steps 116 and 121 are changed to 849 and 899, respectively, in steps 124 and 126.
  • the result is either the generation of a "no call" code at step 117 or the generation of a $2 "denomination” code at step 122.
  • One problem encountered in cunency recognition and counting systems is the difficulty involved in interrupting (for a variety of reasons) and resuming the scanning and counting procedure as a stack of bills is being scanned.
  • CRU cunency recognition unit
  • the CRU has to be halted due to a "minor” e ⁇ or, such as the identification of a scanned bill as being a counterfeit (based on a variety of monitored parameters) or a "no call" (a bill which is not identifiable as belonging to a specific cunency denomination based on the plurality of stored master pattems and/or other criteria)
  • a "minor” e ⁇ or such as the identification of a scanned bill as being a counterfeit (based on a variety of monitored parameters) or a "no call" (a bill which is not identifiable as belonging to a specific cunency denomination based on the plurality of stored master pattems and/or other criteria)
  • this bill B j be transported directly to a retum conveyor or to the system stacker, and die CRU brought to a halt, while at the same time ensuring that the following bills are maintained in positions along the bill transport padi whereby CRU operation can be conveniently resumed without any disruption of the recognition/counting process.
  • the basic problem is that if the CRU is halted with bill B2 only partially scanned, it is difficult to reference the data reflectance samples extracted therefrom in such a way that the scanning may be later continued (when die CRU is restarted) from exactly fhe same point where the sample extraction process was interrupted when die CRU was stopped.
  • the CRU when the CRU is restarted, the overall bill recognition and counting procedure can be resumed without any disruption as if the CRU had never been halted at all.
  • the CRU is subjected to a controlled deceleration process whereby the speed at which bills are moved across the scanhead is reduced from the normal operating speed.
  • die "no call" bill (B j ) is transported to the retum conveyor, at the same time, the following bill B2 is subjected to die standard scanning procedure in order to identify fhe denomination.
  • the rate of deceleration is such that optical scanning of bill B2 is completed by the time the CRU operating speed is reduced to a predefined operating speed. While the exact operating speed at die end of the scanning of bill B2 is not critical, the objective is to permit complete scanning of bill B2 without subjecting it to backlash effects that would result if the ramping were too fast, while at the same time ensuring that bill B ⁇ has in fact been transported to the return conveyor.
  • die deceleration is preferably such that die CRU operating speed is reduced to about one-fifth of its normal operating speed at the end of e deceleration phase, i.e., by the time optical scanning of bill B2 has been completed. It has been determined diat at these speed levels, positive calls can be made as to the denomination of bill B2 based on reflectance samples gathered during the deceleration phase widi a relatively high degree of certainty (i.e., with a conelation number exceeding about 850).
  • the reduced operating speed of the machine at the end of fhe deceleration phase is such diat the CRU can be brought to a total halt before the next following bill B3 has been transported over die optical scanhead.
  • bill B ⁇ is in die retum conveyor
  • bill B2 is maintained in transit between d e optical scanhead and die stacking station after it has been subjected to scanning, and the following bill B3 is stopped short of die optical scanhead.
  • the overall scanning operation can be resumed in an uninterrupted fashion by using the stored call results for bill B2 as the basis for updating die system count appropriately, moving bill B2 from its earlier transitional position along the transport path into the stacking station, and moving bill B3 along the transport path into the optical scanhead area where it can be subjected to normal scanning and processing.
  • a routine for executing the deceleration/ stopping procedure described above is illustrated by the flow chart in FIG. 14. This routine is initiated at step 170 with die CRU in its normal operating mode. At step 171 , a test bill B j is scanned and the data reflectance samples resulting therefrom are processed.
  • step 172 a determination is made as to whether or not test bill B ⁇ is a "no call" using predefined criteria in combination with the overall bill recognition procedure, such as the routine of FIG. 13. If the answer at step 172 is negative, i.e. , the test bill B j can be identified, step 173 is accessed where normal bill processing is continued in accordance with the procedures described above. If, however, die test bill B s found to be a "no call" at step 172, step 174 is accessed where CRU deceleration is initiated, e.g., die transport drive motor speed is reduced to about one-fifth its normal speed.
  • CRU deceleration is initiated, e.g., die transport drive motor speed is reduced to about one-fifth its normal speed.
  • Step 175 determines whether the scanning of bill B 2 is complete.
  • step 176 determines whether a preselected "bill timeout" period has expired so that the system does not wait for the scanning of a bill that is not present.
  • An affirmative answer at step 176 results in the transport drive motor being stopped at step 179 while a negative answer at step 176 causes steps 175 and 176 to be reiterated until one of them produces an affirmative response.
  • step 178 determines whedier eidier of die sensors SI or S2 (described below) is covered by a bill.
  • a negative answer at step 178 indicates diat die bill has cleared botii sensors SI and S2, and tiius die transport drive motor is stopped at step 179. This signifies the end of die deceleration/ stopping process.
  • bill B2 remains in transit while the following bill B3 is stopped on die transport path just short of the optical scanhead.
  • step 179 conective action responsive to me identification of a "no call" bill is conveniently undertaken, and d e CRU is then in condition for resuming fhe scanning process. Accordingly, the CRU can be restarted and die stored results conesponding to bill B2, are used to appropriately update the system count.
  • die identified bill B2 is guided along the transport path to the stacking station, and die CRU continues with its normal processing routine. While the above deceleration process has been described in the context of a "no call" enor, other minor enors (e.g., suspect bills, stranger bills in stranger mode, etc.) are handled in die same manner.
  • a master pattem for a given denomination is generated by averaging a plurality of component pattems.
  • Each component patte is generated by scanning a genuine bill of fhe given denomination.
  • master pattems are generated by scanning a standard bill a pluraUty of times, typically three (3) times, and obtaining the average of conesponding data samples before storing the average as representing a master pattem.
  • a master pattem for a given denomination is generated by averaging a plurality of component pattems, wherein all of the component pattems are generated by scanning a single genuine bill of "standard" quality of fhe given denomination.
  • the "standard" bill is a slightiy used bill, as opposed to a crisp new bill or one which has been subject to a high degree of usage. Rather, the standard bill is a bill of good to average quality.
  • Component pattems generated according to this first methods are illustrated in FIGs. 15a-15c. More specifically, FIGs.
  • test pattems in FIGS. 15a-15c show three test pattems generated, respectively, for the forward scanning of a $1 bill along its green side, die reverse scanning of a $2 bill on its green side, and die forward scanning of a $100 bill on its green side. It should be noted diat, for purposes of clarity fhe test pattems in FIGS. 15a-15c were generated by using 128 reflectance samples per bill scan, as opposed to the prefe ⁇ ed use of only 64 samples. The marked difference existing among co ⁇ esponding samples for these three test pattems is indicative of fhe high degree of confidence widi which cunency denominations may be called using the foregoing optical sensing and conelation procedure.
  • a master pattem for a given denomination is generated by scanning two or more standard bills of standard quality and obtaining a plurality of component pattems. These component pattems are then averaged in deriving a master pattem. For example, it has been found tiiat some genuine $5 bills have dark stairs on the Lincoln Memorial while omer genuine $5 bills have light stairs. To compensate for this variation, standard bills for which component pattems are derived may be chosen with at least one standard bill scanned having dark stairs and with at least one standard bill having light stairs. It has been found that an alternate method can lead to improved performance in a discrimination systems, especially with regards to certain denominations.
  • a master pattem for a given denomination is generated by averaging a plurality of component pattems, wherein some of the component pattems are generated by scanning one or more new bills of me given denomination, and some of the component pattems are generated by scanning one or more old bills of the given denomination.
  • New bills are bills of good quality which have been printed in recent years and have a security thread incorporated therein (for those denominations in which security threads are placed). New bills are preferably relatively crisp.
  • a new $10 bill is preferably a 1990 series or later bill of very high quality, meaning that fhe bill is in near mint condition. Old bills are bills exhibiting some shrinkage and often some discoloration.
  • Shrinkage may result from a bill having been subjected to a relatively high degree of use.
  • a new bill utilized in diis third method is of higher quality than a standard bill of die previous methods, while an old bill in diis third mediod is of lower quality than a standard bill.
  • the third method can be understood by considering Table 2 which summarizes the manner in which component pattems are generated for a variety of denominations.
  • Table 2 summarizes the position of the scanhead relative to the center of the green surface of United States currency as well as the type of bill to be scanned for generating component patterns for various denominations.
  • the three component patterns ("CP") for a given denomination and for a given scan direction are averaged to yield a co ⁇ esponding master pattern.
  • the eighteen (18) rows co ⁇ espond to the preferred method of storing eighteen (18) master patterns.
  • the scanhead position is indicated relative to the center of the borderlined area of die bdi. Thus a position of "0.0" indicates that die scanhead is centered over me center of die borderlined area of the bill. Displacements to the left of center are indicated by negative numbers, while displacements to tiie right are indicated by positive numbers.
  • a position of "-0.2" indicates a displacement of 2/lOth of an inch to the left of die center of a bill
  • a position of "+0J” indicates a displacement of 1/lOti ⁇ s of an inch to the right of the center of a bill.
  • Table 2 indicates that component pattems for a $20 bill scanned in the forward direction are obtained by scanning an old $20 bill 2/10ths of a inch to the right and to the left of the center of the bill and by scanning a new $20 bill directly down the center of the bill.
  • FIG. 15d is a graph illustrating these three pattems. These three pattems are then averaged to obtain the master pattem for a $20 bill scanned in the forward direction.
  • FIG. 15e is a graph illustrating a pattem for a $20 bill scanned in the forward direction derived by averaging the pattems of FIG. 15d. This pattem becomes the co ⁇ esponding $20 master pattem after undergoing normalization.
  • a scanning device in which a bill to be scanned is held stationary and a scanhead is moved over the bill.
  • a scanning device permits the scanhead to be moved laterally, left and right, over a bill to be scanned and tiius permits the scanhead to be positioned over the area of the bill which one wishes to scan, for example, 2/lOths of inch to the left of the center of the borderlined area.
  • the third method of using old and new bills is not used; rather, a standard (“std") bill is used for generating all three component pattems as with the first method.
  • the master pattem for a $1 bill scanned in die forward direction is obtained by averaging three component pattems generated by scanning a standard bill three times, once 2/10ths of an inch to the left, once down die center, and once 2/10ths of an inch to the right.
  • a discrimination system may employ a combmation of methods wherein, for example, some master patte s are generated according die first method and some master pattems are generated according to the diird memod. Likewise, a discrimination system may combine the scanning of new, standard, and old bills to generate component pattems to be averaged in obtaining a master pattem. Additionally, a discrimination system may generate master patte s by scanning bills of various qualities and/or having various characteristics and then averaging the resultant pattems.
  • a discrimination system may scan multiple bills of a given quality for a given denomination, e.g., three new $50 bills, while scanning one or more bills of a different quality for a different denomination, e.g. , tiiree old and worn $1 bills, to generate component pattems to be averaged in obtaining master pattems.
  • the above-described conelation technique can be modified by use of a progressive shifting approach whereby a test pattem which does not co ⁇ espond to any of the master pattems is partitioned into predefined sections, and samples in successive sections are progressively shifted and compared again to the stored pattems in order to identify the denomination. It has experimentally been determined that such progressive shifting effectively counteracts any sample displacement resulting from shrinkage of a bill along the preselected dimension.
  • FIG. 16a shows the conelation between a test pattem (represented by a heavy line) and a conesponding master pattem (represented by a thin line). It is clear from FIG. 16a that the degree of conelation between the two pattems is relatively low and exhibits a conelation factor of 606.
  • the manner in which the conelation between these pattems is increased by employing progressive shifting is best illustrated by considering the conelation at the reference points designated as A-E along the axis defining the number of samples.
  • FIG. 16b shows “single” shifting of the test pattem of FIG. 16a. This is effected by dividing the test patte into two equal segments each comprising 64 samples. The first segment is retained without any shift, whereas the second segment is shifted by a factor of one data sample. Under these conditions, it is found tiiat the conelation factor at the reference points located in the shifted section, particularly at point E, is improved.
  • FIG. 16c shows die effect produced by "double" progressive shifting whereby sections of the test pattem are shifted in three stages. This is accomplished by dividing the overall pattem into three approximately equal sized sections. Section one is not shifted, section two is shifted by one data sample (as in FIG. 16b), and section three is shifted by a factor of two data samples. With “double" shifting, it can be seen that the conelation factor at point E is further increased.
  • FIG. 16d shows the effect on conelation produced by "triple" progressive shifting where the overall pattem is first divided into four approximately equal sized sections. Subsequently, section one is retained without any shift, section two is shifted by one data sample, section three is shifted by two data samples, and section four is shifted by three data samples. Under these conditions, the conelation factor at point E is seen to have increased again.
  • FIG. 16e shows the effect on conelation produced by "quadruple” shifting, where the pattem is first divided into five approximately equal sized sections. The first four sections are shifted in accordance with the "triple" shifting approach of FIG. 16d, whereas the fifth section is shifted by a factor of four data samples. From FIG. 16e it is clear mat the conelation at point E is increased almost to me point of superimposition of the compared data samples.
  • the degree of shrinkage of a scanned bill is determined by comparing the length of the scanned bill, as measured by the scanhead, with die length of an "unshrunk” bill. This "unshrunk” length is pre-stored in the system memory.
  • the type of progressive shifting, e.g., "single", “double”, “triple”, etc., applied to the test pattem is then directly based upon the measured degree of shrinkage. The greater the degree of shrinkage, the greater the number of sections into which the test pattem is divided.
  • An advantage of this approach is mat only one conelation factor is calculated, as opposed to potentially calculating several conelation factors for different types of progressive shifting.
  • progressive shifting is applied to each of the master pattems.
  • the master pattems in the system memory are partitioned into predefined sections, and samples in successive sections are progressively shifted and compared again to the scanned test pattem in order to identify the denomination.
  • the degree of progressive shifting which should be applied to die master pattems may be determined by first measuring the degree of shrinkage of the scanned bill. By first measuring the degree of shrinkage, only one type of progressive shifting is applied to the stored master patte s.
  • the system memory may contain pre-stored pattems co ⁇ esponding to various types of progressive shifting.
  • the scanned test pattem is then compared to all of these stored pattems in the system memory.
  • this approach may be modified to first measure the degree of shrinkage and to dien select only tiiose stored pattems from the system memory which conespond to the measured degree of shrinkage for comparison with the scanned test patte .
  • the progressive shifting realizes substantial increases in the overall conelation factor resulting from pattem comparison.
  • the original conelation factor of 606 (FIG. 16a) is increased to 681 by the "single" shifting shown in FIG. 16b.
  • the "double” shifting shown in FIG. 16c increases the conelation number to 793
  • the "triple” shifting of FIG. 16d increases the conelation number to 906
  • the "quadruple” shifting shown in FIG. 16e increases the overall conelation number to 960.
  • the degree of conelation between a scanned pattem and a master pattem may be negatively impacted if the two pattems are not properly aligned with each odier. Such misalignment between pattems may in turn negatively impact upon the performance of a cunency identification system. Misalignment between patte s may result from a number of factors. For example, if a system is designed so that the scanning process is initiated in response to the detection of the thin borderline sunounding U.S. cunency or the detection of some other printed indicia such as die edge of printed indicia on a bill, stray marks may cause initiation of the scanning process at an improper time.
  • location of the leading edge of printed indicia on Canadian cunency relative to the edge of Canadian cunency may vary up to approximately 0.2 inches (approximately 0.5 cm).
  • the problems associated with misaligned pattems may be overcome by removing data samples from one end of a pattem to be modified and adding data values on the opposite end equal to die data values contained in the conesponding sequence positions of the pattem to which the modified pattem is to be compared. This process may be repeated, up to a predetermined number of times, until a sufficientiy high conelation is obtained between the two pattems so as to permit the identity of a bill under test to be called.
  • a prefened embodiment of the technique can be further understood by considering Table 3.
  • Table 3 contains data samples generated by scanning the nanow dimension of Canadian $2 bills along a segment positioned about the center of the bill on the side opposite the portrait side. More specifically, the second column of Table 3 represents a scanned pattem generated by scanning a test Canadian $2 bill. The scanned pattem comprises 64 data samples a ⁇ anged in a sequence. Each data sample has a sequence position, 1-64, associated therewith. The fifth column represents a master pattem associated with a Canadian $2 bill. The master pattem likewise comprises a sequence of 64 data samples. The third and fourth columns represent the scanned patte after it has been modified in the forward direction one and two times, respectively. In the embodiment depicted in Table 3, one data sample is removed from the beginning of the preceding patte during each modification.
  • the modified pattem represented in the third column is generated by adding an additional data value to the end of the original scanned pattern sequence which effectively removes the first data sample of the original pattern, e.g., 93, from the modified patte .
  • the added data value in the last sequence position, 64 is set equal to the data value contained in the 64th sequence position of the master pattern, e.g., 2210. This copying of the 64th data sample is indicated by an asterisk in the third column.
  • the second modified pattern represented in the fourth column is generated by adding two additional data values to the end of the original scanned pattem which effectively removes the first two data samples of the original scanned, e.g., 93 and 50, from the second modified pattem.
  • the last two sequence positions, 63 and 64 are filled with the data values contained in the 63rd and 64th sequence positions of the master pattem, e.g., 2240 and 2210, respectively.
  • the copying of the 63rd and 64th data samples is indicated by asterisks in the fourth column.
  • the printed area of the bill under test from which the scanned patte was generated was farther away from the leading edge of the bill than was the printed area of die bill from which the master pattem was generated. As a result, fhe scanned pattem trailed the master pattem.
  • the prefened embodiment of the pattem generation method described in conjunction with Table 3 compensates for the variance of the distance between the edge of the bill and e edge of me printed indicia by modifying the scanned pattem in the forward direction.
  • the conelation between die original and modified versions of the scanned pattem and die master pattem increased from 705 for the original, unmodified scanned pattem to 855 for the first modified patte and to 988 for the second modified pattem. Accordingly, the bill under test which would otherwise have been rejected may now be properly called as a genuine $2 Canadian bill tiirough the employment of the pattem generation method discussed above.
  • FIGs. 17a-17c Another modified discrimination technique can be understood with reference to the flowchart of FIGs. 17a-17c.
  • the process of FIGs. 17a-17c involves a method of identifying a bill under test by comparing a scanned pattem retrieved from a bill under test with one or more master pattems associated with one or more genuine bills. After the process begins at step 128a, the scanned pattem is compared wim one or more master pattems associated with genuine bills (step 128b). At step 129 it is determined whether the bill under test can be identified based on the comparison at step 128b. This may be accomplished by evaluating the conelation between the scanned pattem and each of the master pattems. If the bill can be identified, the process is ended at step 130.
  • one or more of the master pattems are designated for further processing at step 131.
  • all of the master pattems may be designated for further processing.
  • less than all of the master pattems may be designated based on a preliminary assessment about the identity of the bill under test. For example, only the master pattems which had the four highest conelation values with respect to the scanned pattem at step 128b might be chosen for further processing. In any case, the number of master pattems designated for further processing is Ml.
  • either the scanned pattem is designated for modification or the
  • Ml master pattems designated at step 131 are designated for modification.
  • the scanned pattem is designated for modification and the master pattems remain unmodified.
  • the iteration counter, I is set equal to one. The iteration counter is used to keep track of how many times the working pattems have been modified.
  • the number of incremental data samples, R, to be removed during each iteration is set.
  • step 136 it is determined whether the scanned pattem has been designated for modification. If it has, then the scanned pattem is replicated Ml times and the Ml replicated pattems, one for each of the Ml master pattems, are designated as working pattems at step 137. If the scanned pattem has not been designated for modification, then the Ml master pattems have been so designated, and die Ml master pattems are replicated and designated as working pattems at step 138.
  • step 139 it is determined whether forward or reverse modification is to be performed on the working pattems.
  • the first R x I data samples from each working pattem are removed at step 140.
  • the first R x I data samples may either be explicitly removed from the working patte s or be removed as a result of adding additional data samples (step 141) to the end of the patte and designating the beginning of the modified pattem to be the (R x I) + 1 sequence position of die original pattem.
  • the data sample which was in the 64th sequence position in the original working pattem will be in the 64 - (R x I) sequence position.
  • the added data values in the last R x I sequence positions of a working pattem are copied from the data samples in tiie last R x I sequence positions of a co ⁇ esponding non-designated pattem at step 141.
  • the working pattems are compared with either respective ones of the non-designated pattems (scanned pattem modified/Ml master pattems not designated for modification) or the non-designated pattem (Ml master pattems designated for modification/scanned pattem not designated for modification) at step 142. Altematively, if reverse modification is to be performed, the last R x I data samples from each working pattem are removed at step 143.
  • the last R x I data samples may either be explicitly removed from me working pattems or be removed as a result of adding additional data samples (step 144) to the beginning of the pattem and designating the beginning of the modified pattem to start with the added data samples.
  • the data sample which was in the 1st sequence position in die original working patte will be in the (R x I) + 1 sequence position.
  • the added data samples in die first R x I sequence positions of a working pattem are copied from the data samples in the first R x I sequence positions of a co ⁇ esponding non-designated pattem at step 144.
  • the working pattems are compared with either respective ones of the non-designated pattems (scanned pattem modified/Ml master pattems not designated for modification) or the non-designated pattem (Ml master pattems designated for modification/ scanned pattem not designated for modification) at step 142. For example, if the scanned pattem is designated for forward modification and four master pattems are designated for further processing, four working pattems are generated from the scanned pattem at step 137, one for each of the four master patte s.
  • R is set to two at step 135, during the first iteration the last two data samples from each of the Ml master pattems are copied and added to die end of fhe Ml working pattems so as to become the last two sequence positions of the Ml working patte s, one working pattem being associated with each of fhe Ml master pattems.
  • four different working pattems are generated witii each working patte co ⁇ esponding to a modified version of the scanned pattem but with each having data values in its last two sequence positions copied from the last two sequence positions of a respective one of the Ml master patte s.
  • the last four sequence positions of each of the Ml master pattems are copied and added to die end of the Ml working pattems so as to become the last four sequence positions of a respective one of the Ml working pattems.
  • four master pattems are designated for further processing and the four designated master pattems are designated for forward modification, four working pattems are generated at step 138, one from each of the four designated master pattems.
  • R is set to two at step 135, during the first iteration the last two data samples of the scanned pattem are copied and added to the end of the Ml working pattems so as to become the last two sequence positions of the Ml working pattems, one working pattem being associated with each of the Ml master pattems.
  • each working pattem co ⁇ esponding to a modified version of a conesponding master pattem but widi each having data values in its last two sequence position copied from the last two sequence positions of the scanned pattem.
  • the last four sequence positions of the scanned pattem are copied and added to the end of the Ml working pattems so as to become the last four sequence positions of the Ml working pattems.
  • step 145 it is determined whether die bill under test can be identified at step 145. If the bill can be identified the process is ended at step 146. Otherwise, fhe iteration counter, I, is incremented by one (step 147), and die incremented iteration counter is compared to a maximum iteration number, T (step 148). If the iteration counter, I, is greater than the maximum iteration number, T, then a no call is issued (step 149a), meaning that a match sufficient to identify the bill under test was not obtained, and the process is ended (step 149b). Otherwise, if the iteration is not greater than the maximum iteration number, the modification process is repeated beginning with step 136.
  • FIGs. 17a- 17c The flowchart of FIGs. 17a- 17c is intended to illustrate one prefened embodiment of the above technique. However, it is recognized diat there are numerous ways in which the steps of the flowchart of FIGs. 17a- 17c may be rearranged or altered and yet still result in the comparison of the same pattems as would be compared if the steps of FIGs. 17a-17c were followed exactly. For example, instead of generating multiple working pattems, a single working pattem may be generated and the leading or trailing sequence positions successively altered before comparisons to conesponding non-designated pattems. Likewise, instead of generating multiple modified pattems directly from unmodified pattems, multiple modified pattems may be generated from the preceding modified pattems.
  • die first data sample of die single forward modified scanned pattem may be removed and one data sample added to the end of the single modified scanned pattem, and then die data samples in the last two sequence positions may be set equal to the data samples in the last 2R sequence positions of a co ⁇ esponding master pattem.
  • modified master patte s are stored, for example in EPROM 60 of FIG. 7a, before a bill under test is scanned.
  • a scanned pattem retrieved from a bill under test is compared to die modified master pattems stored in memory.
  • Modified master pattems are generated by modifying a co ⁇ esponding master pattem in either the forward or backward direction, or both, and filling in any trailing or leading sequence positions widi zeros.
  • step 151a the best and second best conelation results (refened to in FIG. 18a as the "#1 and #2 answers") are initialized to zero and, at step 152a, the test pattem is compared with each of the sixteen or eighteen original master pattems stored in the memory.
  • step 153a the calls conesponding to the two highest conelation numbers obtained up to that point are determined and saved.
  • step 154a a post ⁇ processing flag is set.
  • step 155a the test pattem is compared with each of a second set of 16 or 18 master pattems stored in the memory.
  • This second set of master pattems is the same as the 16 or 18 original master pattems except mat the last sample is dropped and a zero is inserted in front of the first sample. If any of the resulting conelation numbers is higher than the two highest numbers previously saved, the #1 and #2 answers are updated at step 156. Steps 155a and 156a are repeated at steps 157a and 158a, using a tiiird set of master pattems formed by dropping the last two samples from each of the 16 original master pattems and inserting two zeros in front of me first sample.
  • steps 159a and 160a the same steps are repeated again, but using only $50 and $100 master pattems formed by dropping die last tiiree samples from the original master pattems and adding three zeros in front of the first sample.
  • Steps 161a and 162a repeat the procedure once again, using only $1, $5, $10 and $20 master patte s formed by dropping the 33rd sample, whereby original samples 34-64 become samples 33-63, and inserting a 0 as the new last sample.
  • steps 163a and 164a repeat the same procedure, using master pattems for $10 and $50 bills printed in 1950, which differ significantly from bills of the same denominations printed in later years. This routine then retums to die main program at step 165a.
  • the above multiple sets of master pattems may be pre-stored in EPROM 60.
  • FIG. 18b shows the routine as starting at step 150b.
  • the best and second best conelation results (refe ⁇ ed to in FIG. 18b as the "#1 and #2 answers") are initialized to zero and, at step 152b, die test pattem is compared with each of the eighteen original green-side master pattems stored in the memory.
  • the calls co ⁇ esponding to the two highest conelation numbers obtained up to that point are determined and saved.
  • a post ⁇ processing flag is set.
  • the test pattem is compared with each of a second set of 18 green-side master pattems stored in the memory.
  • This second set of master pattems is the same as the 18 original green-side master pattems except that the last sample is dropped and a zero is inserted in front of the first sample. If any of the resulting conelation numbers is higher than the two highest numbers previously saved, the #1 and #2 answers are updated at step 156b.
  • Steps 155b and 156b are repeated at steps 157b and 158b, using a third set of green-side master pattems formed by dropping the last two samples from each of the 18 original master pattems and inserting two zeros in front of die first sample.
  • steps 159b and 160b the same steps are repeated again, but using only $50 and $100 master pattems (two pattems for the $50 and four pattems for the $100) formed by dropping the last three samples from the original master pattems and adding tiiree zeros in front of the first sample.
  • Steps 161b and 162b repeat the procedure once again, using only $1, $5, $10, $20 and $50 master pattems (four pattems for the $10 and two pattems for the otiier denominations) formed by dropping the 33rd sample whereby original samples 34-64 become samples 33-63, and inserting a 0 as the new last sample.
  • steps 163b and 164b repeat die same procedure, using master pattems for $10 and $50 bills printed in 1950 (two pattems scanned along a center segment for each denomination), which differ significantly from bills of the same denominations printed in later years. This routine then retums to the main program at step 165b.
  • the above multiple sets of master pattems may be pre-stored in EPROM 60.
  • a modified version of the routine designated as "CORRES" is initiated.
  • the procedure involved in executing the modified version of CORRES is illustrated at FIG. 19a, which shows the routine as starting at step 180.
  • Step 181 determines whether the bill has been identified as a $2 bill, and, if the answer is negative, step 182 determines whether die best conelation number ("call #1") is greater than 799. If the answer is negative, the conelation number is too low to identify the denomination of die bill with certainty, and at step 183b a black side conelation routine is called (described in more detail below in conjunction with FIGs. 19b-19c).
  • step 182 advances the system to step 186, which determines whether the sample data passes an ink stain test (described below). If the answer is negative, a "no call” bit is set in a conelation result flag at step 183a. A "no call previous bill” flag is then set at step 184, and the routine retums to the main program at step 185. If die answer at step 186 is affirmative, the system advances to step 187 which determines whether the best conelation number is greater than 849. An affirmative answer at step 187 indicates tiiat the conelation number is sufficientiy high that the denomination of the scanned bill can be identified widi certainty without any further checking.
  • a separate register associated with the best conelation number (#1) may men be used to identify the denomination represented by the stored pattem resulting in the highest conelation number.
  • the system retums to the main program at step 185.
  • a negative answer at step 187 indicates that the conelation number is between 800 and 850. It has been found that conelation numbers within this range are sufficient to identify all bills except the $2 bill. Accordingly, a negative response at step 187 advances fhe system to step 189 which determines whether the difference between the two highest conelation numbers ("call #1" and "call #2”) is greater man 149.
  • step 188 If die answer is affirmative, the denomination identified by the highest conelation number is acceptable, and tiius the "good call" bit is set in die conelation result flag at step 188. If the difference between the two highest conelation numbers is less than 150, step 189 produces a negative response which advances the system to step 183b where the black side conelation routine is called.
  • an affirmative response at this step indicates diat die initial call is a $2 bill.
  • This affirmative response initiates a series of steps 190-193 which are similar to steps 182, 186, 187 and 189 described above, except that die numbers 799 and 849 used in steps 182 and 187 are changed to 849 and 899, respectively, in steps 190 and 192.
  • the result is either the setting of a "no call” bit in a conelation result flag at step 183a, the setting of the "good call” bit in the conelation result flag at step 188, or the calling of the black side conelation routine at step 183b.
  • FIGs. 19b and 19c there is shown a flowchart illustrating the steps of the black side conelation routine called at step 183b of FIG. 19a.
  • the black side conelation routine is initiated at step 600, it is determined at step 602 whether the lower read head was the read head that scanned the black side of the test bill. If it was, the lower read head data is normalized at step 604. Otherwise, it is determined at step 606 whedier the upper read head was the read head that scanned the black side of the test bill. If it was, the upper read head data is normalized at step 608.
  • the pattems generated from both sides of the test bill are conelated against the green-side master pattems (see, e.g., step 110 of FIG. 12).
  • the "no call” bit in the conelation result flag is set at step 610
  • the "no call previous bill” flag is set at step 611
  • the program retums to the calling point at step 612.
  • a negative response at step 614 results in the "no call” bit in the conelation result flag being set at step 610, the "no call previous bill” flag being set at step 611, and the program returning to the calling point at step 612.
  • An affirmative response at step 614 results in a determination being made as to whether the best call from the green side conelation conesponds to a $20, $50, or $100 bill at step 616.
  • a negative response at step 616 results in the "no call” bit in the conelation result flag being set at step 610, the "no call previous bill” flag being set at step 611, and the program retuming to the calling point at step 612.
  • the scanned pattem from the black side is conelated against the black-side master pattems associated with the specific denomination and scan direction associated with the best call from tiie green side.
  • multiple black-side master pattems are stored for $20, $50 and $100 bills.
  • three master pattems are stored for scans in the forward direction, and tiiree master pattems are stored for scans in die reverse direction, for a total of six pattems for each denomination.
  • black-side master patte s are generated by scanning a co ⁇ esponding denominated bill along a segment located about the center of the na ⁇ ow dimension of the bill, a segment slightiy displaced (0.2 inches) to the left of center, and a segment slightly displaced (0.2 inches) to die right of center.
  • step 618 it is determined whether the best call from the green side is associated with a forward scan of a $20 bill and, if it is, die normalized data from the black side of the test bill is co ⁇ elated against die black-side master pattems associated with a forward scan of a $20 bill at step 620.
  • step 622 it is determined whether the black-side conelation number is greater than 900 at step 622. If it is, the "good call" bit in the conelation result flag is set at step 648, and the program returns to the calling point at step 646.
  • the program branches accordingly at steps 624 - 640 so diat the normalized data from the black side of the test bill is conelated against die appropriate black-side master pattems.
  • the mechanical portions of the cunency scanning and counting module are shown in FIGS. 20a - 22.
  • the bills are moved in seriatim from the bottom of the stack along a curved guideway 211 which receives bills moving downwardly and rearwardly and changes die direction of travel to a forward direction.
  • the curvature of the guideway 211 conesponds substantially to the curved periphery of the drive roll 223 so as to form a nanow passageway for the bills along the rear side of the drive roll.
  • the exit end of the guideway 211 directs die bills onto a linear path where the bills are scanned.
  • the bills are transported with the nanow dimension of the bills maintained parallel to the transport path and the direction of movement at all times.
  • Bills that are stacked on the bottom wall 205 of the input receptacle are stripped, one at a time, from the bottom of the stack.
  • the bills are stripped by a pair of stripping wheels 220 mounted on a drive shaft 221 which, in turn, is supported across side plates 201, 202.
  • the stripping wheels 220 project through a pair of slots formed in a cover 207.
  • Part of the periphery of each wheel 220 is provided with a raised high-friction, serrated surface 222 which engages the bottom bill of die input stack as the wheels 220 rotate, to initiate feeding movement of the bottom bill from the stack.
  • the serrated surfaces 222 project radially beyond the rest of the wheel peripheries so that the wheels "jog" the bill stack during each revolution so as to agitate and loosen die bottom cunency bill within the stack, diereby facilitating the stripping of the bottom bill from the stack.
  • the stripping wheels 220 feed each stripped bill B (FIG. 21a) onto a drive roll 223 mounted on a driven shaft 224 supported across the side plates 201 and 202.
  • the drive roll 223 includes a central smooth friction surface 225 formed of a material such as mbber or hard plastic. This smooth friction surface 225 is sandwiched between a pair of grooved surfaces 226 and 227 having serrated portions 228 and 229 formed from a high-friction material.
  • the serrated surfaces 228, 229 engage each bill after it is fed onto the drive roll 223 by the stripping wheels 220, to frictionally advance the bill into the na ⁇ ow arcuate passageway formed by die curved guideway 211 adjacent the rear side of the drive roll 223.
  • the rotational movement of the drive roll 223 and the stripping wheels 220 is synchronized so that die serrated surfaces on the drive roll and the stripping wheels maintain a constant relationship to each other.
  • the drive roll 223 is dimensioned so diat the circumference of the outermost portions of the grooved surfaces is greater than the width W of a bill, so that the bills advanced by the drive roll 223 are spaced apart from each other. That is, each bill fed to the drive roll 223 is advanced by hat roll only when the serrated surfaces 228, 229 come into engagement with the bill, so that the circumference of the drive roll 223 determines die spacing between the leading edges of successive bills.
  • the stripping wheels 220 are always stopped with the raised, serrated portions 222 positioned below the bottom wall 205 of the input receptacle. This is accomplished by continuously monitoring the angular position of the serrated portions of the stripping wheels 220 via the encoder 32, and then controlling the stopping time of the drive motor so that the motor always stops the stripping wheels in a position where the serrated portions 222 are located beneath the bottom wall 205 of the input receptacle. Thus, each time a new stack of bills is loaded into the machine, those bills will rest on the smootii portions of the stripping wheels.
  • an idler roll 230 urges each incoming bill against the smooth central surface 225 of tiie drive roll 223.
  • the idler roll 230 is journalled on a pair of arms 231 which are pivotally mounted on a support shaft 232.
  • the wheels 233, 234 are locked to the shaft 232, which in turn is locked against movement in the direction of the bill movement (clockwise as viewed in FIG. 20a) by a one-way spring clutch 235.
  • the clutch 235 is energized to turn the shaft 232 just a few degrees in a direction opposite the direction of bill movement.
  • This pressure roll 236 is journalled on a pair of arms 237 pivoted on a stationary shaft 238.
  • a spring 239 attached to die lower ends of die arms 237 urges the roll 236 against die drive roll 233, through an aperture in the curved guideway 211.
  • the bill being transported by the drive roll 223 engages a flat guide plate 240 which carries a lower scan head 18.
  • Cunency bills are positively driven along the flat plate 240 by means of a transport roll arrangement which includes the drive roll 223 at one end of die plate and a smaller driven roll 241 at the other end of the plate.
  • Both the driver roll 223 and the smaller roll 241 include pairs of smooth raised cylindrical surfaces 242 and 243 which hold the bill flat against the plate 240.
  • a pair of O rings 244 and 245 fit into grooves formed in both the roll 241 and the roll 223 to engage the bill continuously between the two rolls 223 and 241 to transport the bill while helping to hold die bill flat against the guide plate 240.
  • the flat guide plate 240 is provided with openings through which the raised surfaces 242 and 243 of both the drive roll 223 and the smaller driven roll 241 are subjected to counter-rotating contact with conesponding pairs of passive transport rolls 250 and 251 having high-friction mbber surfaces.
  • the passive rolls 250, 251 are mounted on the underside of the flat plate 240 in such a manner as to be freewheeling about their axes 254 and 255 and biased into counter-rotating contact with die conesponding upper rolls 223 and 241.
  • the passive rolls 250 and 251 are biased into contact with the driven rolls 223 and 241 by means of a pair of H-shaped leaf springs 252 and 253 (see FIGs. 23 and 24).
  • Each of the four roUs 250, 251 is cradled between a pair of parallel arms of one of the H-shaped leaf springs 252 and 253.
  • the central portion of each leaf spring is fastened to the plate 240, which is fastened rigidly to the machine frame, so that the relatively stiff arms of the H- shaped springs exert a constant biasing pressure against the rolls and push diem against the upper rolls 223 and 241.
  • the points of contact between the driven and passive transport rolls are preferably coplanar with the flat upper surface of the plate 240 so that cunency bills can be positively driven along die top surface of the plate in a flat manner.
  • the distance between the axes of the two driven transport rolls, and the conesponding counter- rotating passive rolls, is selected to be just short of the length of the most na ⁇ ow dimension of the cunency bills. Accordingly, the bills are firmly gripped under uniform pressure between the upper and lower transport rolls within the scanhead area, thereby minimizing the possibility of bill skew and enhancing the reliability of the overall scanning and recognition process.
  • the positive guiding anangement described above is advantageous in that uniform guiding pressure is maintained on the bills as they are transported tiirough the optical scanhead area, and twisting or skewing of the bills is substantially reduced.
  • This positive action is supplemented by die use of die H-springs 252, 253 for uniformly biasing the passive rollers into contact with the active rollers so that bill twisting or skew resulting from differential pressure applied to the bills along the transport path is avoided.
  • the O-rings 244, 245 function as simple, yet extremely effective means for ensuring that the central portions of the bills are held flat.
  • the adjustment screw 257 adjusts the proximity of the magnetic head 256 relative to a passing bill and thereby adjusts the strength of the magnetic field in the vicinity of the bill.
  • FIG. 22 shows die mechanical a ⁇ angement for driving the various means for transporting cunency bills through the machine.
  • a motor 260 drives a shaft 261 carrying a pair of pulleys 262 and 263.
  • the pulley 262 drives the roll 241 through a belt 264 and pulley 265, and the pulley 263 drives the roll 223 through a belt 266 and pulley 267.
  • Both pulleys 265 and 267 are larger than pulleys 262 and 263 in order to achieve the desired speed reduction from the typically high speed at which the motor 260 operates.
  • the shaft 221 of the stripping wheels 220 is driven by means of a pulley 268 provided thereon and linked to a co ⁇ esponding pulley 269 on the shaft 224 through a belt 270.
  • the pulleys 268 and 269 are of the same diameter so that the shafts 221 and 224 rotate in unison.
  • the optical encoder 32 is mounted on the shaft of the roller 241 for precisely tracking the position of each bill as it is transported through the machine, as discussed in detail above in connection with the optical sensing and conelation technique.
  • the upper and lower scanhead assemblies are shown most clearly in FIGs. 25- 28. It can be seen that the housing for each scanhead is formed as an integral part of a unitary molded plastic support member 280 or 281 that also forms the housings for the light sources and photodetectors of the photosensors PSI and PS2.
  • the lower member 281 also forms the flat guide plate 240 that receives the bills from the drive roll 223 and supports the bills as they are driven past the scanheads 18a and 18b.
  • the two support members 280 and 281 are mounted facing each other so that the lenses 282 and 283 of the two scanheads 18a, 18b define a nanow gap through which each bill is transported. Similar, but slightiy larger, gaps are formed by die opposed lenses of the light sources and photodetectors of the photosensors PSI and PS2.
  • the upper support member 280 includes a tapered entry guide 280a which guides an incoming bill into the gaps between the various pairs of opposed lenses.
  • the lower support member 281 is attached rigidly to the machine frame.
  • the upper support member 280 is mounted for limited vertical movement when it is lifted manually by a handle 284, to facilitate the clearing of any paper jams that occur beneath the member 280.
  • the member 280 is slidably mounted on a pair of posts 285 and 286 on the machine frame, with a pair of springs 287 and 288 biasing the member 280 to its lowermost position.
  • the upper scanhead 18a includes a pair of LEDs 22a, directing light downwardly through an optical mask on top of the lens 282 onto a bill traversing the flat guide plate 240 beneath the scanhead.
  • the LEDs 22a are angularly disposed relative to the vertical axis of the scanhead so that their respective light beams combine to illuminate the desired light strip defined by an aperture in the mask.
  • the scanhead 18a also includes a photodetector 26a mounted directly over the center of the illuminated strip for sensing the light reflected off die strip.
  • the photodetector 26a is linked to the CPU 30 through the ADC 28 for processing the sensed data as described above.
  • the illumination by the LED's as a function of the distance from the central point "0" along the X axis should optimally approximate a step function as illustrated by the curve A in FIG. 29.
  • the variation in illumination by an LED typically approximates a Gaussian function, as illustrated by die curve B in FIG. 29.
  • the two LEDs 22a are angularly disposed relative to the vertical axis by angles ⁇ and ⁇ , respectively.
  • the angles a and ⁇ are selected to be such that the resultant strip illumination by the LED's is as close as possible to the optimum distribution curve A in FIG. 29.
  • the LED illumination distribution realized by this arrangement is illustrated by the curve designated as "C" in FIG. 29 which effectively merges the individual Gaussian distributions of each light source to yield a composite distribution which sufficiently approximates the optimum curve A.
  • each scanhead includes two pairs of LEDs and two photodetectors for illuminating, and detecting light reflected from, strips of two different sizes.
  • each mask also includes :two slits which are formed to allow light from the LEDs to pass through and illuminate light strips of the desired dimensions. More specifically, one slit illuminates a relatively wide strip used for obtaining the reflectance samples which co ⁇ espond to the characteristic pattem for a test bill. In a prefe ⁇ ed embodiment, the wide slit has a length of about 0.500" and a width of about 0.050".
  • the second slit forms a relatively na ⁇ ow illuminated strip used for detecting the thin borderline sunounding the printed indicia on cunency bills, as described above in detail.
  • the nanow slit 283 has a length of about 0.300" and a width of about 0.010".
  • each scanhead includes tiiree resilient seals or gaskets 290, 291, and 292.
  • the two side seals 290 and 291 seal the outer ends of die LEDs 22, while the center seal 292 seals the outer end of the photodetector 26.
  • the photosensors PSI and PS2 include photodetectors 293 and 294 mounted within the lower support member 281 in immediate opposition to co ⁇ esponding light sources 295 and 296 mounted in the upper support member 280.
  • the photodetectors 293, 294 detect beams of light directed downwardly onto the bill transport path from the light sources 295, 296 and generate analog outputs which co ⁇ espond to the sensed light passing tiirough the bill. Each such output is converted into a digital signal by a conventional ADC convertor unit (not shown) whose output is fed as a digital input to and processed by the system CPU.
  • the presence of a bill adjacent the photosensors PSI and PS2 causes a change in the intensity of the detected light, and the conesponding changes in the analog outputs of the photodetectors 293 and 294 serve as a convenient means for density- based measurements for detecting the presence of "doubles" (two or more overlaid or overlapped bills) during die cunency scanning process.
  • the photosensors may be used to collect a predefined number of density measurements on a test bill, and die average density value for a bill may be compared to predetermined density thresholds (based, for instance, on standardized density readings for master bills) to determine the presence of overlaid bills or doubles.
  • both the light sources and the photodetectors are enclosed by lenses mounted so close to the bill path that they are continually wiped by the bills. This provides a self-cleaning action which reduces maintenance problems and improves the reliability of the outputs from the photosensors over long periods of operation.
  • the CPU 30, under control of software stored in d e EPROM 34, monitors and controls the speed at which the bill transport mechanism 16 transports bills from the bill separating station 14 to the bill stacking unit. Flowcharts of the speed control routines stored in the EPROM 34 are depicted in FIGs. 31-35. To execute more than the first step in any given routine, the cunency discriminating system 10 must be operating in a mode requiring the execution of the routine.
  • the transport speed of the bill transport mechamsm 16 when a user places a stack of bills in the bill accepting station 12 for counting, the transport speed of the bill transport mechamsm 16 must accelerate or "ramp up" from zero to top speed. Therefore, in response to receiving the stack of bills in the bill accepting station 12, the CPU 30 sets a ramp-up bit in a motor flag stored in die memory unit 38. Setting the ramp-up bit causes the CPU 30 to proceed beyond step 300b of the ramp-up routine. If die ramp-up bit is set, the CPU 30 utilizes a ramp-up counter and a fixed parameter "ramp-up step" to incrementally increase the transport speed of the bill transport mechanism 16 until the bill transport mechanism 16 reaches its top speed.
  • the "ramp-up step” is equal to the incremental increase in the transport speed of the bill transport mechamsm 16, and fhe ramp-up counter determines the amount of time between incremental increases in the bill transport speed.
  • the greater the value of the "ramp-up step” the greater the increase in the transport speed of the bill transport mechanism 16 at each increment.
  • the greater the value of the "ramp-up step” and the lesser the maximum value of the ramp-up counter the lesser the time it takes the bill transport mechanism 16 to reach its top speed.
  • the ramp-up routine in FIG. 31 employs a variable parameter "new speed”, a fixed parameter “full speed”, and the variable parameter "transport speed”.
  • the “full speed” represents the top speed of die bill transport mechanism 16, while the “new speed” and “transport speed” represent the desired cu ⁇ ent speed of the bill transport mechanism 16.
  • the "transport speed” of the bill transport mechanism 16 actually differs from the "new speed” by a "speed offset value”. Outputting the "transport speed" to the bill transport mechanism 16 causes the bill transport mechanism 16 to operate at the transport speed.
  • the variable parameter "transport speed” represents the top speed of die bill transport mechanism 16
  • the "new speed" and “transport speed” represent the desired cu ⁇ ent speed of the bill transport mechanism 16.
  • step 301 first decrements the ramp-up counter from its maximum value (step 301). If the maximum value of the ramp-up counter is greater than one at step 302, the CPU 30 exits the speed control software in FIGs. 31-35 and repeats steps 300b, 301, and 302 during subsequent iterations of the ramp-up routine until the ramp-up counter is equal to zero. When the ramp-up counter is equal to zero, the CPU 30 resets the ramp-up counter to its maximum value (step 303). Next, the CPU 30 increases the "new speed” by the "ramp-up step” (step 304). If the "new speed” is not yet equal to the "full speed” at step 305, die “transport speed” is set equal to the "new speed” plus the "speed offset value” (step 306).
  • the "transport speed” is output to the bill transport mechanism 16 at step 307 of the routine in FIG. 31 to change the speed of the bill transport mechanism 16 to the "transport speed”.
  • the CPU 30 repeats steps 300b-306 until the "new speed" is greater than or equal to the "full speed”.
  • step 306 the ramp-up bit in the motor flag is cleared (step 308), a pause-after-ramp bit in the motor flag is set (step 309), a pause-after-ramp counter is set to its maximum value (step 310), and the parameter "new speed” is set equal to the "full speed” (step 311).
  • die “transport speed” is set equal to the “new speed” plus the "speed offset value” (step 306). Since fhe "new speed” is equal to the "full speed”, outputting the "transport speed” to die bill transport mechanism 16 causes the bill transport mechanism 16 to operate at its top speed.
  • the CPU 30 sets a ramp-to-slow-speed bit in the motor flag. Setting the ramp-to-slow- spe ⁇ d bit causes die CPU 30 to proceed beyond step 312 of the ramp-to-slow-speed routine in FIG. 32 on the next iteration of the software in FIGs. 31-35.
  • the CPU 30 causes the bill transport mechanism 16 to controllably decelerate or "ramp down" from its top speed to a slow speed.
  • the ramp-to-slow speed routine in FIG. 32 is similar to the ramp-up routine in FIG. 31, it is not described in detail herein.
  • the CPU 30 decrements a ramp-down counter (step 313) and determines whether or not the ramp-down counter is equal to zero (step 314). If the ramp-down counter is not equal to zero, the CPU 30 exits the speed control software in FIGs. 31-35 and repeats steps 312, 313, and 314 of the ramp-to-slow-speed routine in FIG. 32 during subsequent iterations of the speed control software until the ramp-down counter is equal to zero. Once the ramp-down counter is equal to zero, the CPU 30 resets the ramp-down counter to its maximum value (step 315) and subtracts a "ramp-down step" from the variable parameter "new speed” (step 316). The "new speed” is equal to the fixed parameter "full speed" prior to initiating the ramp-to-slow-speed routine in FIG. 32.
  • die “new speed” is compared to a fixed parameter "slow speed” (step 317). If die “new speed” is greater than the “slow speed”, the "transport speed” is set equal to the "new speed” plus the "speed offset value” (step 318) and this "transport speed” is output to the bill transport mechanism 16 (step 307 of FIG. 31). During subsequent iterations of the ramp-to-slow-speed routine, the CPU 30 continues to decrement the "new speed” by die “ramp-down step” until the "new speed" is less than or equal to the
  • FIG. 33 depicts a ramp-to-zero-speed routine in which the CPU 30 ramps down the transport speed of the bill transport mechanism 16 to zero either from its top speed or its slow speed.
  • the CPU 30 In response to completion of counting of a stack of bills, the CPU 30 enters this routine to ramp down the transport speed of the bill transport mechanism 16 from its top speed to zero.
  • die CPU 30 In response to the optical scanhead 18 detecting a stranger, suspect, or no call bill and the ramp-to- slow-speed routine in FIG. 32 causing the transport speed to be equal to a slow speed, die CPU 30 enters the ramp-to-zero-spe ⁇ d routine to ramp down die transport speed from the slow speed to zero.
  • the CPU 30 determines whether or not an initial-braking bit is set in the motor flag (step 324). Prior to ramping down die transport speed of die bill transport mechanism 16, the initial- braking bit is clear. Therefore, flow proceeds to fhe left branch of the ramp-to-zero- speed routine in FIG. 33. In this left branch, the CPU 30 sets the initial-braking bit in the motor flag (step 325), resets the ramp-down counter to its maximum value (step 326), and subtracts an "initial-braking step" from the variable parameter "new speed" (step 327). Next, the CPU 30 determines whether or not the "new speed" is greater than zero (step 328).
  • variable parameter "transport speed” is set equal to die “new speed” plus the "speed offset value” (step 329) and this "transport speed” is output to the bill transport mechanism 16 at step 307 in FIG. 31.
  • the CPU 30 enters the right branch of the routine at step 324 because the initial-braking bit was set during die previous iteration of the ramp-to-zero-speed routine. With the initial-braking bit set, the CPU 30 decrements the ramp-down counter from its maximum value (step 330) and determines whether or not the ramp-down counter is equal to zero (step 331). If the ramp-down counter is not equal to zero, the CPU 30 immediately exits tiie speed control software in FIGs. 31-35 and repeats steps 323, 324, 330, and 331 of the ramp-to-slow-speed routine during subsequent iterations of the speed control software until the ramp-down counter is equal to zero.
  • the CPU 30 resets the ramp-down counter to its maximum value (step 332) and subtracts a "ramp-down step" from the variable parameter "new speed” (step 333).
  • This "ramp-down step” is smaller than the "initial-braking step” so that the “initial-braking step” causes a larger decremental change in the transport speed of the bill transport mechanism 16 than that caused by the "ramp-down step”.
  • the CPU 30 determines whether or not the "new speed” is greater than zero (step 328). If the "new speed” is greater than zero, the "transport speed” is set equal to the “new speed” plus the “speed offset value” (step 329) and this "transport speed” is outputted to the bill transport mechanism 16 (step 307 in FIG. 31). During subsequent iterations of the speed control software, the CPU 30 continues to decrement the "new speed” by the "ramp-down step” at step 333 until the "new speed” is less than or equal to zero at step 328.
  • the CPU 30 clears the ramp- to-zero-speed bit and the initial-braking bit in the motor flag (step 334), sets a motor-at-rest bit in the motor flag (step 335), and sets the "new speed” equal to zero (step 336). Finally, the "transport speed” is set equal to the “new speed” plus the "speed offset value” (step 329). Since the "new speed” is equal to zero, outputting the "transport speed” to the bill transport mechanism 16 at step 307 in FIG. 31 halts the bill transport mechanism 16.
  • the CPU 30 monitors and stabilizes the transport speed of the bill transport mechanism 16 when the bill transport mechanism 16 is operating at its top speed or at slow speed.
  • the CPU 30 monitors the optical encoder 32. While monitoring the optical encoder 32, it is important to synchronize the feedback loop routine with any transport speed changes of die bill transport mechanism 16.
  • the CPU 30 enters a pause-after- ramp routine in FIG. 34 prior to entering the feedback loop routine in FIG. 35 if the bill transport mechanism 16 completed ramping up to its top speed or ramping down to slow speed during the previous iteration of the speed control software in FIGs. 31- 35.
  • the pause-after-ramp routine in FIG. 34 allows the bill transport mechanism 16 to "catch up" to the CPU 30 so that the CPU 30 does not enter the feedback loop routine in FIG. 35 prior to the bill transport mechanism 16 changing speeds.
  • the CPU 30 sets a pause-after-ramp bit during step 309 of the ramp-up routine in FIG. 31 or step 320 of the ramp-to-slow-speed routine in FIG. 32. With the pause-after-ramp bit set, flow proceeds from step 337 of the pause- after-ramp routine to step 338, where the CPU 30 decrements a pause-after-ramp counter from its maximum value.
  • the CPU 30 exits the pause-after-ramp routine in FIG. 34 and repeats steps 337, 338, and 339 of the pause-after-ramp routine during subsequent iterations of the speed control software until die pause-after-ramp counter is equal to zero.
  • the CPU 30 clears the pause- after-ramp bit in the motor flag (step 340) and sets the feedback loop counter to its maximum value (step 341).
  • the maximum value of the pause-after-ramp counter is selected to delay the CPU 30 by an amount of time sufficient to permit the bill transport mechanism 16 to adjust to a new transport speed prior to the CPU 30 monitoring the new transport speed with the feedback loop routine in FIG. 35.
  • the CPU 30 decrements a feedback loop counter from its maximum value (step 343). If the feedback loop counter is not equal to zero at step 344, the CPU 30 immediately exits the feedback loop routine in FIG. 35 and repeats steps 342, 343, and 344 of the feedback loop routine during subsequent iterations of the speed control software in FIGs. 31-36 until the feedback loop counter is equal to zero.
  • the CPU 30 resets : the feedback loop counter to its maximum value (step 345), stores the present count of the optical encoder 32 (step 346), and calculates a variable parameter "actual difference” between the present count and a previous count of the optical encoder 32 (step 347).
  • the "actual difference” between the present and previous encoder counts represents the transport speed of the bill transport mechanism 16. The larger the “actual difference” between the present and previous encoder counts, die greater the transport speed of the bill transport mechanism.
  • the CPU 30 subtracts the "actual difference” from a fixed parameter "requested difference” to obtain a variable parameter "speed difference” (step 348).
  • die CPU 30 multiplies the "speed difference” by a "gain constant” (step 354) and sets the variable parameter "transport speed” equal to the multiplied difference from step 354 plus the "speed offset value” plus a fixed parameter "target speed” (step 355).
  • the "target speed” is a value tiiat, when added to the "speed offset value", produces the ideal transport speed.
  • the calculated "transport speed” is greater than this ideal transport speed by the amount of the multiplied difference.
  • the calculated "transport speed” is output to the bill transport mechanism 16 at step 307 so that the bill transport mechanism 16 operates at the calculated "transport speed”. If, however, die calculated "transport speed” is greater than the "maximum allowable speed” at step 356, the parameter "transport speed” is set equal to the "maximum allowable speed” (step 357) and is output to the bill transport mechanism 16 (step 307). If the "speed difference" is less than or equal to zero at step 349, the bill transport speed of the bill transport mechanism 16 is too fast or is ideal.
  • the CPU 30 multiplies the "speed difference" by a "gain constant” (step 350) and sets the variable parameter "transport speed” equal to the multiplied difference from step 350 plus the "speed offset value” plus a fixed parameter “target speed” (step 351).
  • the calculated “transport speed” is less than this ideal transport speed by the amount of the multiplied difference. If the calculated "transport speed” is nonetheless greater than or equal to a fixed parameter "minimum allowable speed” at step 352, the calculated “transport speed” is output to the bill transport mechanism 16 at step 307 so that the bill transport mechanism 16 operates at the calculated "transport speed". If, however, the calculated "transport speed” is less than die “minimum allowable speed” at step 352, the parameter “transport speed” is set equal to the "minimum allowable speed” (step 353) and is output to the bill transport mechanism 16 (step 307).
  • Step 403 determines whether the cunent bill is a "no call,” and if the answer is negative, the denomination determined for the new bill is retrieved at step 404.
  • a "density setting" is retrieved from memory at step 405. If the "density setting" has been turned off, this condition is sensed at step 406, and the system retums to the main program at step 413. If the "density setting" is not turned off, a denominational density comparison value is retrieved from memory at step 407.
  • the memory preferably contains five different density values (for five different density settings, i.e., degrees of sensitivity) for each denomination. Thus, for a cunency set containing seven different denominations, the memory contains 35 different values.
  • the denomination retrieved at step 404 (or step 402 in the event of a "no call"), and the density setting retrieved st step 405, determine which of the 35 stored values is retrieved at step 407 for use in the comparison steps described below.
  • the density comparison value retrieved at step 407 is compared to d e average density represented by the output of the photosensor PSI. The result of this comparison is evaluated at step 409 to determine whether the output of sensor SI identifies a doubling of bills for the particular denomination of bill determined at step 402 or 404. If the answer is negative, the system retums to the main program at step 413.
  • step 410 compares the retrieved density comparison value to die average density represented by the output of the second sensor PS2. The result of this comparison is evaluated at step 411 to determine whether the output of the photosensor PS2 identifies a doubling of bills. Affirmative answers at both step 409 and step 411 result in the setting of a "doubles e ⁇ or" flag at step 412, and die system then retums to the main program at step 413.
  • the "doubles e ⁇ or" flag can, of course, be used to stop die bill transport motor.
  • FIG. 37 illustrates a routine that enables the system to detect bills which have been badly defaced by dark marks such as ink blotches, felt-tip pen marks and die like. Such severe defacing of a bill can result in such distorted scan data that the data can be interpreted to indicate the wrong denomination for the bill. Consequently, it is desirable to detect such severely defaced bills and dien stop die bill transport mechanism so that the bill in question can be examined by die operator.
  • the routine of FIG. 37 retrieves each successive data sample at step 450b and then advances to step 451 to determine whether that sample is too dark. As described above, the output voltage from the photodetector 26 decreases as the darkness of the scanned area increases.
  • a preselected threshold level for the photodetector output voltage such as a threshold level of about 1 volt, is used to designate a sample that is "too dark.”
  • step 451 An affirmative answer at step 451 advances the system to step 452 where a "bad sample” count is incremented by one. A single sample that is too dark is not enough to designate the bill as seriously defaced. Thus, the "bad sample” count is used to determine when : a preselected number of consecutive samples, e.g., ten consecutive samples, are determined to be too dark. From step 452, the system advances to step 453 to determine whether ten consecutive bad samples have been received. If fhe answer is affirmative, the system advances to step 454 where an e ⁇ or flag is set. This represents a "no call" condition, which causes die bill transport system to be stopped in the same manner discussed above.
  • step 451 When a negative response is obtained at step 451, the system advances to step 455 where the "bad sample” count is reset to zero, so that diis count always represents the number of consecutive bad samples received. From step 455 the system advances to step 456 which determines when all the samples for a given bill have been checked. As long as step 456 yields a negative answer, the system continues to retrieve successive samples at step 450b. When an affirmative answer is produced at step 456, the system retums to the main program at step 457.
  • a routine for automatically monitoring and making any necessary co ⁇ ections in various line voltages is illustrated in FIG. 38. This routine is useful in automatically compensating for voltage drifts due to temperature changes, aging of components and the like.
  • the routine starts at step 550 and reads the output of a line sensor which is monitoring a selected voltage at step 550b.
  • Step 551 determines whether the reading is below 0.60, and if the answer is affirmative, step 552 determines whether the reading is above 0.40. If step 552 also produces an affirmative response, the voltage is witiiin the required range and us the system returns to the main program step 553. If step 551 produces a negative response, an incremental co ⁇ ection is made at step 554 to reduce the voltage in an attempt to retum it to the desired range. Similarly, if a negative response is obtained at step 552, an incremental co ⁇ ection is made at step 555 to increase the voltage toward the desired range.
  • sensors may be added to deterrnine fhe size of a bill to be scanned. These sensors are placed upstream of the scanheads.
  • a prefened embodiment of size determining sensors is illustrated in FIG. 39.
  • Two leading/trailing edge sensors 1062 detect the leading and trailing edges of a bill 1064 as it passes along the transport path. These sensors in conjunction with the encoder 32 (FIG. 2a-2b) may be used to determine the dimension of die bill along a direction parallel to the scan direction which in FIG. 39 is the nanow dimension (or width) of the bill 1064.
  • two side edge sensors 1066 are used to detect the dimension of a bill 1064 transverse to the scan direction which in FIG.
  • leading edge detectors 1068 are shown.
  • the detection of the leading edge 1069 of a bill 1070 by leading edge sensors 1068 triggers scanning in an area a given distance away from the leading edge of die bill 1070, e.g. , D ⁇ or D2, which may vary depending upon the preliminary indication of the identity of a bill based on the dimensions of a bill.
  • the leading edge 1069 of a bill may be detected by one or more of the scanheads (to be described below) in a similar manner as that described with respect to FIGs. 7a and 7b.
  • the beginning of scanning may be triggered by positional information provided by the encoder 32 of FIG. 2a-2b, for example, in conjunction with the signals provided by sensors 1062 of FIG. 39, thus eliminating the need for leading edge sensors 1068.
  • a scanned pattem can be offset from a co ⁇ esponding master pattem by one or more samples. Such offsets can lead to e ⁇ oneous rejections of genuine bills due to poor conelation between scanned and master pattems.
  • overall scanned patte s and master pattems can be shifted relative to each other as illustrated in FIGs. 41a and 41b. More particularly, FIG. 41a illustrates a scanned pattem which is offset from a co ⁇ esponding master pattem.
  • FIG. 41b illustrates the same pattems after the scanned pattem is shifted relative to the master pattem, thereby increasing the conelation between the two pattems.
  • master pattems may be stored in memory conesponding to different offset amounts.
  • the central area may not be suitable for bills originating in other countries. For example, for bills originating from Country 1, it may be determined that segment S j (FIG. 40) provides a more preferable area to be scanned, while segment S2 (FIG. 40) is more preferable for bills originating from Country 2.
  • multiple scanheads may be positioned next to each otiier.
  • a prefe ⁇ ed embodiment of such a multiple scanhead system is depicted in FIG. 42.
  • Multiple scanheads 1072a-c and 1072d-f are positioned next to each other along a direction lateral to die direction of bill movement.
  • Such a system permits a bill 1074 to be scanned along different segments.
  • Multiple scanheads 1072a-f are arranged on each side of the transport path, thus permitting both sides of a bill 1074 to be scanned.
  • Two-sided scanning may be used to permit bills to be fed into a cunency discrimination system according to die present invention with either side face up.
  • An example of a two-sided scanhead arrangement is described above in connection with FIGs. 2a, 6c, and 6d.
  • Master pattems generated by scanning genuine bills may be stored for segments on one or both sides.
  • the pattems retrieved by scanning both sides of a bill under test may be compared to a master set of single- sided master pattems.
  • a pattem retrieved from one side of a bill under test should match one of the stored master pattems, while a pattem retrieved from the other side of the bill under test should not match one of the master pattems.
  • master pattems may be stored for both sides of genuine bills. In such a two-sided system, a pattem retrieved by scanning one side of a bill under test should match with one of the master pattems of one side (Match 1) and a pattem retrieved from scanning the opposite side of a bill under test should match the master pattem associated widi tiie opposite side of a genuine bill identified by Match 1.
  • the number of comparisons may be reduced by limiting comparisons to pattems conesponding to the same side of a bill. That is, for example, when it is known that a bill is "face up”, scanned patte s associated with scanheads above the transport path need only be compared to master pattems generated by scanning the "face" of genuine bills.
  • face of a bill it is meant a side which is designated as d e front surface of the bill. For example, the front or "face” of a U.S.
  • the face orientation may be determinable in some situations by sensing the color of the surfaces of a bill.
  • An altemative method of determining the face orientation of U.S. bills by detecting the borderline on each side of a bill is described above in connection with FIGs. 6c, 6d, and 12. The implementation of color sensing is discussed in more detailed below.
  • the bill transport mechanism operates in such a fashion that the central area C of a bill 1074 is transported between central scanheads 1072b and 1072e.
  • Scanheads 1072a and 1072c and likewise scanheads 1072d and 1072f are displaced the same distance from central scanheads 1072b and 1072e, respectively.
  • a bill may be scanned in either direction, e.g., top edge first (forward direction) or bottom edge first (reverse direction).
  • master pattems are stored from the scanning of genuine bills in both the forward and reverse directions.
  • FIG. 42 illustrates a system having three scanheads per side, any number of scanheads per side may be utilized. Likewise, it is not necessary that there be a scanhead positioned over die central region of a bill.
  • FIG. 43 illustrates another prefe ⁇ ed embodiment of the present invention capable of scanning the segments S 1 and S 2 of FIG. 40.
  • Scanheads 1076a, 1076d, 1076e, and 1076h scan a bill 1078 along segment S l while scanheads 1076b, 1076c, 1076f, and 1076g scan segment S2.
  • FIG. 44 depicts another prefe ⁇ ed embodiment of a scanning system according to the present invention having laterally moveable scanheads 1080a-b. Similar scanheads may be positioned on the opposite side of the transport path. Moveable scanheads 1080a-b may provide more flexibility that may be desirable in certain scanning situations. Upon the determination of the dimensions of a bill as described in connection with FIG. 39, a preliminary determination of the identity of a bill may be made. Based on this preliminary determination, the moveable scanheads 1080a-b may be positioned over die area of the bill which is most appropriate for retrieving discrimination information.
  • scanheads 1080a and 1080b may be appropriately positioned for scanning such a segment, e.g., scanhead 1080a positioned 2.0 cm left of center and scanhead 1080b positioned 2.0 cm right of center. Such positioning permits proper discrimination regardless of the whether the scanned bill is being fed in the forward or reverse direction.
  • FIG. 44 depicts a system in which the transport mechanism is designed to deliver a bill 1082 to be scanned centered witiiin the area in which scanheads 1080a-b are located. Accordingly, scanheads 1080a-b are designed to move relative to the center of the transport path with scanhead 1080a being moveable within the range R j and scanhead 1080b being moveable witiiin range R2.
  • FIG. 45 depicts anodier prefe ⁇ ed embodiment of a scanning system according to the present invention wherein bills to be scanned are transported in a left justified manner along the transport path, that is wherein the left edge L of a bill 1084 is positioned in the same lateral location relative to the transport path. Based on the dimensions of the bill, the position of the center of the bill may be determined and the scanheads 1086a-b may in rum be positioned accordingly. As depicted in FIG. 45, scanhead 1086a has a range of motion R3 and scanhead 1086b has a range of motion R 4 . The ranges of motion of scanheads 1086a-b may be influenced by the range of dimensions of bills which the discrimination system is designed to accommodate. Similar scanheads may be positioned on the opposite side of the transport path.
  • the transport mechanism may be designed such that scanned bills are not necessarily centered or justified along die lateral dimension of the transport path. Rather the design of die transport mechanism may permit the position of bills to vary left and right within the lateral dimension of the transport path.
  • the edge sensors 1066 of FIG. 39 may be used to locate die edges and center of a bill, and tiius provide positional information in a moveable scanhead system and selection criteria in a stationary scanhead system.
  • FIG. 46 is a top view of a staggered scanhead arrangement according to a prefened embodiment of the present invention. As illustrated in FIG. 46, a bill 1130 is transported in a centered manner along the transport path 1132 so that the center 1134 of the bill 1130 is aligned widi die center 1136 of me transport path 1132.
  • Scanheads 1140a-h are arranged in a staggered manner so as to permit scanning of the entire widdi of the transport path 1132.
  • the areas illuminated by each scanhead are illustrated by strips 1142a, 1142b, 1142e, and 1142f for scanheads 1140a, 1140b, 1140e, and 1140f, respectively.
  • scanheads 1140a and 1140h may either not be activated or their ou ⁇ ut ignored.
  • preliminary infonnation about a document can be obtained, such as its size or color, appropriately positioned stationary scanheads may be activated or laterally moveable scanheads may be appropriately positioned provided die preliminary information provides some indication as to die potential identity of the document.
  • die preliminary information provides some indication as to die potential identity of the document.
  • many or all of die scanheads of a system may be activated to scan a document.
  • only the output or derivations thereof of appropriately located scanheads may be used to generate scanned pattems.
  • Derivations of output signals include, for example, data samples stored in memory generated by sampling output signals.
  • information enabling a pretiminary determination as to a document's identity may be obtained by analyzing information either from sensors separate from the scanheads or from one or more of the scanheads themselves.
  • scanheads 1140a-h of FIG. 46 are arranged in a non-overlapping manner, they may altematively be arranged in an overlapping manner.
  • an overlapping scanhead arrangement may provide greater selectivity in the segments to be scanned. This increase in scanable segments may be beneficial in compensating for cunency manufacturing tolerances which result in positional variances of fhe printed indicia on bills relative to their edges.
  • scanheads positioned above fhe transport path are positioned upstream relative to their conesponding scanheads positioned below die transport path.
  • FIG. 47a and 47b illustrate another embodiment wherein a plurality of analog sensors 1150 such as photodetectors are laterally displaced from each other and are a ⁇ anged in a linear array within a single scanhead 1152.
  • FIG. 47a is a top view while FIG. 47b is a side elevation view of such a linear array embodiment.
  • the output of individual sensors 1150 are connected to photodetectors (not shown) through die use of graded index fibers, such as a "lens array" manufactured by MSG America, Inc., part number SLA20A1675702A3, and subsequently to analog-to- digital converters and a CPU (not shown) in a manner similar to that depicted in FIGs. 1 and 6a.
  • a bill 1154 is transported past the linear array scanhead 1152 in a centered fashion.
  • a prefened length for the linear array scanhead is about 6-7 inches (15 cm - 17 cm).
  • a discriminating system incorporating a linear array scanhead according the present invention would be capable of accommodating a wide variety of bill- types. Additionally, a linear array scanhead provides a great deal of flexibility in how information may be read and processed with respect to various bills.
  • scanned pattems may be "generated” or approximated in a direction perpendicular to the direction of bill movement.
  • the linear array scanhead 1152 comprises one hundred and sixty (160) sensors 1150 over a length of 7 inches (17.78 cm) instead of taking samples for 64 encoder pulses from say 30 sensors, samples may be taken for 5 encoder pulses from all 160 cells (or all those positioned over the bill 1154).
  • 160 scanned pattems (or selected ones thereof) of 5 data samples each may be used for pattem comparisons.
  • the data acquisition time is significantly reduced from 64 encoder pulses to only 5 encoder pulses.
  • the time saved in acquiring data can be used to permit more time to be spent processing data and/or to reduce the total scanning time per bill thus enabling increased diroughput of the identification system.
  • the linear array scanhead permits a great deal of flexibility in tailoring the areas to be scanned. For example, it has been found diat the leading edges of Canadian bills contain valuable graphic information. Accordingly, when it is determined that a test bill may be a Canadian bill (or when the identification system is set to a Canadian cunency setting), the scanning area can be limited to the leading edge area of bills, for example, by activating many laterally displaced sensors for a relatively small number of encoder pulses.
  • FIG. 48 is a top view of anodier prefened embodiment of a linear array scanhead 1170 having a plurality of analog sensors 1172 such as photodetectors wherein a bill 1174 is transported past the scanhead 1170 in a non-centered manner.
  • positional information from size-determining sensors may be used to select appropriate sensors.
  • the linear array scanhead itself may be employed to determine the size of a bill, thus eliminating the need for separate si-ze-determining sensors. For example, all sensors may be activated, data samples derived from sensors located on the ends of the linear array scanhead may be pretiminarily processed to determine the lateral position and die lengtii of a bill.
  • FIG. 49 is a top view of another embodiment of a linear scanhead 1180 having the ability to compensate for skewing of bills.
  • Scanhead 1180 has a plurality of analog sensors 1182 and a bill 1184 is transported past scanhead 1180 in a skewed manner. Once the skew of a bill has been determined, for example through the use of leading edge sensors, readings from sensors 1182 along the scanhead 1180 may be appropriately delayed.
  • the bill may be treated as being fed in a non-skewed manner since the amount of lateral deviation between a scan along a skewed angle and a scan along a non-skewed angle is minimal for a scan of only a few encoder pulses.
  • a single scanned pattem may be generated from me outputs of more than one sensor.
  • a scanned pattem may be generated by taking data samples from sensor 1186a for a given number of encoder pulses, then taking data samples from sensor 1186b for a next given number of encoder pulses, and then taking data samples from sensor 1186c for a next given number of encoder pulses.
  • the number of given encoder pulses for which data samples may be taken from the same sensor is influenced by die degree of skew: die greater the degree of skew of the bill, the fewer the number of data samples which may be obtained before switching to the next sensor.
  • master pattems may be generated and stored for various degrees of skew, thus permitting a single sensor to generate a scanned patte from a bill under test.
  • FIGs. 47-49 while only a single linear array scanhead is shown, another linear array scanhead may be positioned on the opposite side of the transport path to permit scanning of either or both sides of a bill.
  • the benefits of using a linear array scanhead may also be obtainable using a multiple scanhead arrangement which is configured appropriately, such as depicted in FIG. 46 or a linear anangement of multiple scanheads.
  • color may also be used to discriminate bills. For example, while all U.S. bills are printed in the same colors, e.g., a green side and a black side, bills from other countries often vary in color with the denomination of die bill. For example, a German 50 deutsche mark bill is brown in color while a German 100 deutsche mark bill is blue in color. Altematively, color detection may be used to determine die face orientation of a bill, such as where the color of each side of a bill varies. For example, color detection may be used to determine the face orientation of U.S. bills by detecting whether or not the "green" side of a U.S. bill is facing upwards.
  • color information may be used in addition to size information to preliminarily identify a bill.
  • color information may be used to dete ⁇ nine the face orientation of a bill, which determination may be used to select upper or lower scanheads for scanning a bill, or to compare scanned pattems retrieved from upper scanheads with a set of master pattems generated by scanning a co ⁇ esponding face while the scanned pattems retrieved from fhe lower scanheads are compared with a set of master pattems generated by scanning an opposing face.
  • color sensing may be incorporated into the scanheads described above.
  • Such color sensing may be achieved by, for example, incorporating color filters, colored light sources, and/or dichroic beamsplitters into the cunency discrimination system of the present invention.
  • Color information acquisition is described in more detail in co ⁇ pending U.S. application Serial No. 08/219,093 filed March 29, 1994, for a "Cunency Discriminator and Authenticator", incorporated herein by reference.
  • Various color information acquisition techniques are described in U.S. Patent Nos. 4,841,358; 4,658,289; 4,716,456; 4,825,246; and 4,992,860.
  • a cunency discriminator may be further understood by referring to the flowchart of FIGs. 50a and 50b.
  • a bill is fed along a transport path (step 1102) past sensors which measure die length and width of the bill (step 1104). These size determining sensors may be, for example, those illustrated in FIG. 39.
  • step 1106 it is deterrnined whether die measured dimensions of the bill match me dimensions of at least one bill stored in memory, such as EPROM 60 of FIG. 7a. If no match is found, an appropriate e ⁇ or is generated at step 1108. If a match is found, the color of the bill is scanned at step 1110.
  • step 1112 it is deterrnined whether the color of the bill matches a color associated with a genuine bill having the dimensions measured at step 1104. An enor is generated at step 1114 if no such match is found. However, if a match is found, a pretiminary set of potentially matching bills is generated at step 1116. Often, only one possible identity will exist for a bill having a given color and dimensions. However, the preUr nary set of step 1116 is not limited to tiie identification of a single bill-type, that is, a specific denomination of a specific cunency system; but rather, the preUrninary set may comprise a number of potential bill-types. For example, all U.S. bills have the same size and color. Therefore, the preliminary set generated by scanning a U.S. $5 bill would include U.S. bills of all denominations.
  • selected scanheads in a stationary scanhead system may be activated (step 1118). For example, if the preliminary identification indicates that a bill being scanned has the color and dimensions of a German 100 deutsche mark bill, the scanheads over regions associated widi die scanning of an appropriate segment for a German 100 deutsche mark bill may be activated. Then upon detection of the leading edge of die bill by sensors 1068 of FIG. 40, the appropriate segment may be scanned. Altematively, all scanheads may be active widi only the scanning information from selected scanheads being processed. Altematively, based on the preliminary identification of a bill (step 1116), moveable scanheads may be appropriately positioned (step 1118).
  • the bill is scanned for a characteristic pattem (step 1120).
  • the scanned pattems produced by the scanheads are compared with the stored master pattems associated with genuine bills as dictated by the prehminary set.
  • processing time may be reduced.
  • the preliminary set indicated that the scanned bill could only possibly be a German 100 deutsche mark bill
  • the master patte or pattems associated with a German 100 deutsche mark bill need be compared to the scanned pattems. If no match is found, an appropriate enor is generated (step 1124).
  • a scanned pattem does match an appropriate master pattem, the identity of the bill is accordingly indicated (step 1126) and the process is ended (step 1128).
  • the system may be adapted to identify a bill under test as either belonging to a specific bill-type or not.
  • the system may be adapted to store master information associated with only a single bill-type such as a United Kingdom 5 pound bill. Such a system would identify bills under test which were United Kingdom 5 pound bills and would reject all other bill- types.
  • the scanheads described above may be incorporated into a cunency identification system capable of identifying a variety of cunencies.
  • the system may be designed to accommodate a number of cunencies from different countries.
  • Such a system may be designed to permit operation in a number of modes.
  • the system may be designed to permit an operator to select one or more of a plurality of bill-types which the system is designed to accommodate. Such a selection may be used to limit the number of master pattems with which scanned pattems are to be compared.
  • the operator may be permitted to select the manner in which bills will be fed, such as all bills face up, all bills top edge first, random face orientation, and/or random top edge orientation.
  • the system may be designed to permit output information to be displayed in a variety of formats to a variety of output devices, such as a monitor, LCD display, or printer.
  • the system may be designed to count the number of each specific bill-type identified and to tabulate the total amount of cunency counted for each of a plurality of cunency systems.
  • a stack of bills could be placed in the bill accepting station 12 of FIG. 2a-2b, and the output unit 36 of FIG. 2a-2b may indicate that a total of 370 British pounds and 650
  • German marks were counted. Altematively, the output from scanning the same batch of bills may provide more detailed information about the specific denominations counted, for example, one 100 pound bill, five 50 pound bills, and one 20 pound bill and thirteen 50 deutsche mark bills.
  • a manual selection device such as a switch or a scrolling selection display, may be provided so that the customer may designate what type of cunency is to be discriminated. For example, in a system designed to accommodate botii Canadian and German cunency, the customer could turn a dial to the Canadian bill setting or scroll through a displayed menu and designate Canadian bills.
  • scanned pattems need only be compared to master pattems co ⁇ esponding to the indicated type of cunency, e.g., Canadian bills.
  • the indicated type of cunency e.g., Canadian bills.
  • die processing time can be reduced.
  • a system may be designed to compare scanned pattems to all stored master pattems. In such a system, the customer need not pre-declare what type of cunency is to be scanned. This reduces die demands on the customer.
  • such a system would permit the inputting of a mixture of bills from a number of countries. The system would scan each bill and automatically determine the issuing country and the denomination.
  • an alternate system employs a semi-automatic bill-type discriminating method.
  • a semi-automatic bill-type discriminating method operates in a manner similar to the stranger mode described above.
  • a stack of bills is placed in the input hopper.
  • the first bill is scanned and the generated scanned patte is compared with the master pattems associated with bills from a number of different countries.
  • the discriminator identifies the country- type and the denomination of the bill. Then the discriminator compares all subsequent bills in the stack to the master pattems associated with bills only from the same country as the first bill. For example, if a stack of U.S. bills were placed in the input hopper and die first bill was a $5 bill, the first bill would be scanned.
  • the scanned pattem would be compared to master pattems associated with bills from a number of countries, e.g. , U.S. , Canadian, and German bills.
  • scanned patte s from the remaining bills in the stack are compared only to master patte s associated with U.S. bills, e.g., $1, $2, $5, $10, $20, $50, and $100 bills.
  • the bill may be flagged as described above such as by stopping the transport mechanism while the flagged bill is retumed to the customer.
  • a cunency discriminating device designed to accommodate both Canadian and German cunency bills will now be described.
  • a cunency discriminating device similar to that described above in connection with scanning U.S. cunency (see, e.g. , FIGs. 1-38 and accompanying description) is modified so as to be able to accept both Canadian and German cunency bills.
  • the discriminator is designed to accept either stacks of Canadian bills or stacks of German bills, the bills in the stacks being faced so that the picture side of all fhe bills will be scanned by a triple scanhead arrangement to be described in connection with FIG. 51.
  • this triple scanhead replaces the single scanhead anangement housed in the unitary molded plastic support member 280 (see, e.g. , FIGs. 25 and 26).
  • the triple scanhead arrangement 1200 comprises a center scanhead 1202, a left scanhead 1204, and a right scanhead 1206 housed in a unitary molded plastic support member 1208.
  • a bill 1210 passes under the anangement 1200 in the direction shown.
  • O-rings are positioned near each scanhead, preferably two O-rings per scanhead, one on each side of a respective scanhead, to engage the bill continuously while transporting the bill between rolls 223 and 241 (FIG. 20a) and to help hold the bill flat against the guide plate 240 (FIG. 20a).
  • the left 1204 and right 1206 scanhead are placed slightly upstream of the center scanhead 1202 by a distance D .
  • D is 0.083 inches (0.21 cm).
  • the center scanhead 1202 is centered over the center C of the transport path 1216.
  • the center LQ of the left scanhead 1204 and the center RQ of the right scanhead 1206 are displaced laterally from center C of the transport path in a symmetrical fashion by a distance D 4 .
  • D4 is 1.625 inches (4.128 cm).
  • the scanheads 1202, 1204, and 1206 are each similar to the scanheads described above connection with FIGs. 1-38, except only a wide slit having a length of about 0.500 inch and a width of about 0.050 inch is utilized.
  • each scanhead The wide slit of each scanhead is used both to detect the leading edge of a bill and to scan a bill after the leading edge has been detected.
  • Two photosensors 1212 and 1214 are located along the lateral axis of the left and right scanheads 1204 and 1206, one on either side of the center scanhead 1202. Photosensors 1212 and 1214 are same as the photosensors PSI and PS2 described above (see, e.g. , FIGs. 26 and 30). Photosensors 1212 and 1214 are used to detect doubles and also to measure the dimensions of bills in the direction of bill movement which in the prefened embodiment depicted in FIG. 51 is the nanow dimension of bills.
  • Photosensors 1212 and 1214 are used to measure the nanow dimension of a bill by indicating when the leading and trailing edges of a bill passes by the photosensors 1212 and 1214. This information in combination with the encoder information permits the nanow dimension of a bill to be measured. All Canadian bills are 6 inches (15.24 cm) in their long dimension and 2.75 inches (6.985 cm) in their nanow dimension. German bills vary in size according to denomination. In a prefe ⁇ ed embodiment of the cunency discriminating system, the discriminating device is designed to accept and discriminate $2, $5, $10, $20, $50. and $100 Canadian bills and 10 DM, 20 DM, 50 DM, and 100 DM German bills.
  • German bills vary in size from 13.0 cm (5J2 inches) in the long dimension by 6.0 cm (2.36 inches) in the nanow dimension for 10 DM bills to 16.0 cm (6.30 inches) in the long dimension by 8.0 cm (3.15 inches) in the nanow dimension for 100 DM bills.
  • the input hopper of the discriminating device is made sufficientiy wide to accommodate all the above listed Canadian and German bills, e.g., 6.3 inches (16.0 cm) wide.
  • FIG. 52 is a top view of a Canadian bill illustrating the areas scanned by the triple scanhead anangement of FIG. 51.
  • segments SL j , SC ⁇ , and SR ⁇ are scanned by the left 1204, center 1202, and right 1206 scanheads, respectively, on the picture side of the bill 1300. These segments are similar to segment S in FIG. 4. Each segment begins a predetermined distance D ⁇ inboard of the leading edge of the bill. In a prefened embodiment D ⁇ is 0.5" (1.27 cm).
  • Segments SL j , SC j , and S j each comprise 64 samples as shown in FIGs. 3 and 5.
  • Canadian bills are scanned at a rate of 1000 bills per minute.
  • the lateral location of segments SL j , SC j , and SR ] _ is fixed relative to the transport path 1301 but may vary left to right relative to bill 1300 since the lateral position of bill 1300 may vary left to right within the transport path 1301.
  • a set of eighteen master Canadian pattems are stored for each type of Canadian bill that the system is designed to discriminate, three for each scanhead in both the forward and reverse directions. For example, three pattems are generated by scanning a given genuine Canadian bill in the forward direction with the center scanhead. One pattem is generated by scanning down the center of the bill along segment SC ⁇ , a second is generated by scanning along a segment SC2 initiated 1.5 samples before the beginning of SC ] _, and a third is generated by scanning along a segment SC3 initiated 1.5 samples after the beginning of SC j . The second and third pattems are generated to compensate for the problems associated with triggering off the edge of a bill as discussed above.
  • each of the above master pattems is generated after considering the conelation results achieved when a bill is displaced slightly to the left or to the right of the center of each scanhead, i.e., lines LQ, SQ, and RQ.
  • a scan of a genuine bill may be taken down the center of a bill
  • a second scan may be taken along a segment 0J5 inch to the right of center (+0J5 inch)
  • a third scan may be taken along a segment 0J5 inch to the left of center (-0J5 inch).
  • the actual scan location may be adjusted slightly to the right or left so the effect of the lateral displacement of a bill on the conelation results is minimized.
  • the master pattem associated with a forward scan of a Canadian $2 bill using the center scanhead 1202 may be taken along a line 0.05 inch to the right of the center of the bill.
  • the above stored master pattems are generated either by scanning both a relatively new crisp genuine bill and an older yellowed genuine bill and averaging the patte s generated from each or, altematively, by scanning an average looking bill.
  • Master pattems are stored for nine types of Canadian bills, namely, the newer series $2, $5, $10, $20, $50, and $100 bills and the older series $20, $50, and $100 bills. Accordingly, a total of 162 Canadian master pattems are stored (9 types x 18 per type).
  • FIG. 53 is a flowchart of the threshold test utilized in calling the denomination of a Canadian bill.
  • the flowchart of FIG. 53 replaces the flowchart of FIG. 13.
  • the conelation results associated with conelating a scanned pattem to a master pattem of a given type of Canadian bill in a given scan direction and a given offset in the direction of bill movement from each of fhe three scanheads are summed.
  • the highest of the resulting 54 summations is designated the #1 conelation and the second highest is preliminarily designated the #2 conelation.
  • the #1 and #2 co ⁇ elations each have a given bill type associated with them.
  • the preliminarily designated #2 conelation is substituted with the next highest conelation where the bill denomination is different from the denomination of the bill type associated with the #1 conelation.
  • Step 1304 checks the denomination associated with the #1 conelation. If the denomination associated with the #1 conelation is not a $50 or $100, the #1 conelation is compared to a threshold of 1900 at step 1306. If the #1 conelation is less than or equal to 1900, the conelation number is too low to identify the denomination of the bill with certainty. Therefore, step 1308 sets a "no call" bit in a conelation result flag and the system retums to the main program at step 1310. If, however, the #1 conelation is greater than 1900 at step 1306, the system advances to step 1312 which determines whether the #1 conelation is greater than 2000.
  • the conelation number is sufficiently high that the denomination of the scanned bill can be identified with certainty without any further checking. Consequently, a "good call" bit is set in the conelation result flag at step 1314 and the system retums to the main program at step 1310.
  • step 1316 checks the denomination associated with the #2 conelation. If the denomination associated with the #2 conelation is not a $50 or $100, the #2 conelation is compared to a threshold of 1900 at step 1318. If the #2 conelation is less than or equal to 1900, the denomination identified by the #1 conelation is acceptable, and thus the "good call” bit is set in the conelation result flag at step 1314 and the system retums to the main program at step 1310. If, however, the #2 conelation is greater than 1900 at step 1318, the denomination of the scanned bill cannot be identified with certainty because the #1 and #2 conelations are both above 1900 and, yet, are associated with different denominations. Accordingly, the "no call" bit is set in the conelation result flag at step 1308.
  • the #2 conelation is compared to a threshold of 1500 at step 1320. If the #2 conelation is less than or equal to 1500, the denomination identified by the #1 conelation is acceptable, and thus the "good call” bit is set in the conelation result flag at step 1314 and the system retums to the main program at step 1310. If, however, the #2 conelation is greater than 1500 at step 1320, the denomination of the scanned bill cannot be identified with certainty. As a result, the "no call" bit is set in the conelation result flag at step 1308.
  • the #1 conelation is compared to a threshold of 1500 at step 1322. If the #1 conelation is less than or equal to 1500, the denomination of the scanned bill cannot be identified with certainty and, therefore, the "no call” bit is set in the conelation result flag at step 1308. If, however, the #1 conelation at step 1322 is greater than 1500, the system advances to step 1312 which determines whether the #1 conelation is greater than 2000. If the #1 conelation is greater than 2000, the conelation number is sufficiently high that the denomination of the scanned bill can be identified with certainty without any further checking. Consequently, a "good call" bit is set in the conelation result flag at step 1314 and the system retums to the main program at step 1310.
  • step 1316 checks the denomination associated with the #2 conelation. If the denomination associated with the #2 conelation is not a $50 or $100, the #2 conelation is compared to a threshold of 1900 at step 1318. If the #2 conelation is less than or equal to 1900, the denomination identified by the #1 conelation is acceptable, and thus the "good call” bit is set in the conelation result flag at step 1314 and the system retums to the main program at step 1310. If, however, the #2 conelation is greater than 1900 at step 1318, the denomination of the scanned bill cannot be identified with certainty. Accordingly, the "no call" bit is set in the conelation result flag at step 1308.
  • the #2 conelation is compared to a threshold of 1500 at step 1320. If the #2 conelation is less than or equal to 1500, the denomination identified by the #1 conelation is acceptable, and thus the "good call” bit is set in the conelation result flag at step 1314 and the system retums to the main program at step 1310. If, however, the #2 conelation is greater than 1500 at step 1320, the denomination of the scanned bill cannot be identified with certainty. As a result, the "no call" bit is set in the conelation result flag at step 1308 and the system returns to the main program at step 1310. Now the use of the triple scanhead anangement 1200 in scanning and discriminating German cunency will be described.
  • a segment similar to segment S of FIG. 4 is scanned over the center of the transport path at a predetermined distance Dg inboard after the leading edge of a bill is detected.
  • Dg is 0.25" (0.635 cm).
  • the scanned segment comprises 64 samples as shown in FIGs. 3 and 5.
  • German bills are scanned at a rate of 1000 bills per minute.
  • the lateral location of the scanned segment is fixed relative to the transport path 1216 but may vary left to right relative to bill 1210 since the lateral position of bill 1210 may vary left to right within the transport path 1216.
  • FIG. 54a illustrates the general areas scanned in generating master 10 DM German pattems. Due to the short length of 10 DM bills in their long dimension relative to the width of the transport path, thirty 10 DM master pattems are stored. A first set of five pattems are generated by scanning a genuine 10 DM bill 1400 in the forward direction along laterally displaced segments all beginning a predetermined distance Dg inboard of the leading edge of the bill 1400. Each of these five laterally displaced segments is centered about a respective one of lines Li -Lc. One such segment S10 j centered about line L ⁇ is illustrated in FIG. 54a. Line L ⁇ is disposed down the center C of fhe bill 1400.
  • lines I- ⁇ -Lg are disposed in a symmetrical fashion about the center C of the bill 1400.
  • lines L2 and L3 are laterally displaced from L j by a distance Dy where D7 is 0.24" (0.61 cm) and lines L and L5 are laterally displaced from L ⁇ by a distance Dg where Dg is 0.48" (1.22 cm).
  • a second set of five pattems are generated by scanning a genuine 10 DM bill
  • the second predetermined distance is such that scanning begins one sample earlier than Dg, that is about 30 mils before the initiation of the pattems in the first set of five pattems.
  • a third set of five pattems are generated by scanning a genuine 10 DM bill 1400 in the forward direction along laterally displaced segments along lines j -L ⁇ all beginning at a third predetermined distance inboard of the leading edge of the bill 1400, the third predetermined distance being greater than the predetermined distance Dg.
  • One such segment SIO3 centered about line L j is illustrated in FIG. 54a.
  • the third predetermined distance is such that scanning begins one sample later than Dg, that is about 30 mils after the initiation of the pattems in the first set of five pattems.
  • FIG. 54b illustrates the general areas scanned in generating master 20 DM, 50 DM, and 100 DM German pattems. Due to the lengths of 20 DM, 50 DM, and 100 DM bills in their long dimension being shorter than the width of the transport path, eighteen 20 DM master pattems, eighteen 50 DM master pattems, and eighteen 100 DM master pattems are stored. The 50 DM master patte s and the 100 DM master pattems are taken in the same manner as the 20 DM master pattems except mat the 50 DM master pattems and 100 DM master pattems are generated from respective genuine 50 DM bills and 100 DM bills while the 20 DM master pattems are generated from genuine 20 DM bills. Therefore, only the generation of the 20 DM master pattems will be described in detail.
  • a first set of three pattems are generated by scanning a genuine 20 DM bill 1402 in the forward direction along laterally displaced segments all beginning a predetermined distance Dg inboard of the leading edge of the bill 1402. Each of these three laterally displaced segments is centered about a respective one of lines Lg- Lg. One such segment S20 j centered about line Lg is illustrated in FIG. 54b.
  • Line Lg is disposed down the center C of the bill 1402.
  • lines L ⁇ -Lg are disposed in a symmetrical fashion about the center C of the bill 1402.
  • lines L ⁇ and Lg are laterally displaced from Lg by a distance D 9 where D 9 is 0.30" (0.76 cm) for the 20 DM bill.
  • D Q is 0.20" (0.51 cm) for the 50 DM bill and 0.10" (0.25 cm) for the 100 DM bill.
  • a second set of three pattems are generated by scanning a genuine 20 DM bill 1402 in the forward direction along laterally displaced segments along lines Lg-Lg all beginning at a second predetermined distance inboard of the leading edge of the bill 1402, the second predetermined distance being less than the predetermined distance Dg.
  • One such segment S20- ? centered about line Lg is illustrated in FIG. 54b.
  • the second predetermined distance is such that scanning begins one sample earlier than Dg, that is about 30 mils before the initiation of the pattems in the first set of three pattems.
  • a third set of three pattems are generated by scanning a genuine 20 DM bill 1402 in the forward direction along laterally displaced segments along lines Lg-Lg all beginning at a third predetermined distance inboard of the leading edge of the bill 1402, the third predetermined distance being greater than the predetermined distance Dg.
  • One such segment S2O centered about line Lg is illustrated in FIG. 54b.
  • the third predetermined distance is such that scanning begins one sample later than Dg, that is about 30 mils after the initiation of the pattems in the first set of three patte s.
  • the above three sets of three pattems yield nine pattems in the forward direction.
  • Nine additional 20 DM master patte s taken in the manner described above but in the reverse direction are also stored.
  • the above stored master pattems are generated either by scanning both a relatively new crisp genuine bill and an older yellowed genuine bill and averaging the pattems generated from each or, altematively, by scanning an average looking bill.
  • the na ⁇ ow dimension of each bill is measured independently by photosensors 1212 and 1214 and then averaged to indicate the length of the na ⁇ ow dimension of a bill.
  • a first number of encoder pulses occur between the detection of the leading and trailing edges of a bill by the photosensor 1212.
  • a second number of encoder pulses occur between the detection of the leading and trailing edges of the bill by the photosensor 1214.
  • the photosensors 1212 and 1214 can also determine the degree of skew of a bill as it passes by the triple scanhead anangement 1200. By counting the number of encoder pulses between the time when photosensors 1212 and 1214 detect the leading edge of a bill, the degree of skew can be determined in terms of encoder pulses. If no or little skew is measured, a generated scanned pattern is only compared to master patterns associated with genuine bills having the same narrow dimension length. If a relatively large degree of skew is detected, a scanned pattern will be compared with master patte s associated with genuine bills having the next smaller denominational amount than would be indicated by the measured nanow dimension length.
  • Table 4 indicates which denominational set of master patterns are chosen for comparison to the scanned pattern based on the measured narrow dimension length in terms of encoder pulses and the measured degree of skew in terms of encoder pulses:
  • FIG. 55 is a flowchart of the threshold test utilized in calling the denomination of a German bill. It should be understood that this threshold test compares the scanned bill patte only to the set of master patterns selected in accordance with Table 4. Therefore, the selection made in accordance with Table 4 provides a preliminary indication as to the denomination of the scanned bill.
  • the threshold test in FIG. 55 serves to confirm or overturn the preliminary indication given by Table 4.
  • Step 1326 checks the nanow dimension length of the scanned bill in terms of encoder pulses. If the nanow dimension length is less than 1515 at step 1326, the preliminary indication is that the denomination of the scanned bill is a 10 DM bill. In order to confirm this preliminary indication, the #1 correlation is compared to 550 at step 1328. If the #1 conelation is greater than 550, the correlation number is sufficiently high to identify the denomination of the bill as a 10 DM bill. Accordingly, a "good call" bit is set in a correlation result flag at step 1330, and the system returns to the main program at step 1332.
  • step 1334 sets a "no call” bit in the correlation result flag. If step 1326 determines that the nanow dimension length is greater than or equal to 1515, a conelation threshold of 800 is required to confirm the preliminary denominational indication provided by Table 4. Therefore, if the #1 conelation is greater than 800 at step 1336, the preliminary indication provided by Table 4 is confirmed. To confirm the preliminary indication, the "good call” bit is set in the conelation result flag. If, however, the #1 correlation is less than or equal to 800 at step 1336, the preliminary indication is rejected and the "no call” bit in the conelation result flag is set at step 1334. The system then retums to the main program at step 1332.
  • FIG. 56 is a functional block diagram illustrating another embodiment of a cunency discriminator system 1662.
  • the discriminator system 1662 comprises an input receptacle 1664 for receiving a stack of cunency bills.
  • a transport mechanism (as represented by a ⁇ ows A and B) transports the bills in the input receptacle past an authenticating and discriminating unit 1666 to a canister 1668 where the bills are re- stacked.
  • the authenticating and discriminating unit 1666 may additionally include various authenticating tests such as the ultraviolet authentication test described below.
  • Signals from the authenticating and discriminating unit 1666 are sent to a signal processor such as a central processor unit (“CPU") 1670.
  • the CPU 1670 records the results of the authenticating and discriminating tests in a memory 1672.
  • the value of the bill is added to a total value counter in memory 1672 that keeps track of the total value of the stack of bills that was inserted in the input receptacle 1664 and scanned by the authenticating and discriminating unit 1666. Additionally, depending on the mode of operation of the discriminator system 1662, counters associated with one or more denominations are maintained in the memory 1672.
  • a $1 counter may be maintained to record how many $1 bills were scanned by the authenticating and discriminating unit 1666.
  • a $5 counter may be maintained to record how many $5 bills were scanned, and so on.
  • the total value of the scanned bills may be determined without maintaining a separate total value counter.
  • the total value of the scanned bills and/or the number of each individual denomination may be displayed on a display 1674 such as a monitor or LCD display.
  • a discriminating unit such as the authenticating and discriminating unit 1666 may not be able to identify the denomination of one or more bills in the stack of bills loaded into the input receptacle 1664. For example, if a bill is excessively worn or soiled or if the bill is torn, a discriminating unit may not be able to identify the bill. Furthermore, some known discrimination methods do not have a high discrimination efficiency and thus are unable to identify bills which vary even somewhat from an "ideal" bill condition or which are even somewhat displaced by the transport mechanism relative to the scanning mechanism used to discriminate bills. Accordingly, such poorer performing discriminating units may yield a relatively large number of bills which are not identified.
  • the discriminator system 1662 may be designed so that when the authenticating and discriminating unit is unable to identify a bill, the transport mechanism is altered to divert the unidentified bill to a separate storage canister. Such bills may be "flagged” or "marked” to indicate that the bill is a no call or suspect bill. Altematively, the unidentified bill may be retumed to the customer.
  • the discriminator system 1662 may be designed to continue operation automatically when a bill is diverted from the normal transport path because the bill is a "no call" or a counterfeit suspect, or the system may be designed to require a selection element to be depressed. For example, upon examination of a returned bill the customer may conclude that the retumed bill is genuine even though it was not identified by the discriminating unit.
  • the discriminator system 1680 comprises an input receptacle 1682 for receiving a stack of cunency bills.
  • a transport mechanism (as represented by a ⁇ ow C) transports the bills from the input receptacle, one at a time, past an authenticating and discriminating unit 1684.
  • a bill is either transported to a verified- deposit canister 1686 (a ⁇ ow D), to an escrow canister 1688 (a ⁇ ow E), or to a retum station 1690 (a ⁇ ow F).
  • a ⁇ ow D verified- deposit canister
  • escrow canister 1688 escrow canister
  • a ⁇ ow F retum station
  • the bill is transported to the verified-deposit canister 1686.
  • the authenticating and discriminating unit determines that a bill is a fake, the bill is immediately routed (a ⁇ ow E) to the escrow canister 1688.
  • the flagged bill is retumed (a ⁇ ow F) to the customer at station 1690. If the customer concludes that the bill is genuine, the customer may deposit the retumed bill or bills in an envelope for later verification by the bank and crediting to the customer's account. The discriminator system 1680 then resumes operation, and the suspect bills in the deposit envelope are held for manual pick-up without incrementing the counters associated with the various denomination and/or the total value counters.
  • a UV light source 2102 illuminates a document 2104. Depending upon the characteristics of the document, ultraviolet light mav be reflected off the document and/or fluorescent light mav be emitted from the document.
  • a detection system 2106 is positioned so as to receive any light reflected or emitted toward it but not to receive any UV light directly from the light source 2102.
  • the detection system 2106 comprises a UV sensor 2108, a fluorescence sensor 2110, filters, and a plastic housing.
  • the light source 2102 and the detection system 2106 are both mounted to a printed circuit board 2112.
  • the document 2104 is transported in the direction indicated by a ⁇ ow A by a transport system (not shown).
  • the document is transported over a transport plate 2114 which has a rectangular opening 2116 in it to permit passage of light to and from the document.
  • the rectangular opening 2116 is 1.375 inches (3.493 cm) by 0.375 inches (0.953 cm).
  • the opening 2116 is covered with a transparent UV-transmitting acrylic window 2118.
  • the UV light source 2102 and the detection system 2106 are completely enclosed within a housing (not shown) comprising the transport plate 2114.
  • FIG. 59 there is shown a functional block diagram illustrating a prefe ⁇ ed embodiment of a UV authenticating system.
  • FIG. 59 shows a UV sensor 2202, a fluorescence sensor 2204, and filters 2206, 2208 of a detection system such as the detection system 2106 of FIG. 59.
  • Light from the document passes through the filters 2206, 2208 before striking the sensors 2202, 2204, respectively.
  • An ultraviolet filter 2206 filters out visible light and permits UV light to be transmitted and hence to strike the UV sensor 2202.
  • Similarlv. a visible light filter 2208 filters out UV light and permits visible light to be transmitted and hence to strike fluorescence sensor 2204.
  • UV light which has a wavelength below 400 nm
  • visible light which has a wavelength greater than 400 nm
  • the UV filter 2206 transmits light having a wavelength between about 260 nm and about 380 nm and has a peak transmittance at 360 nm.
  • the visible light filter 2208 is a blue filter and preferably transmits light having a wavelength between about 415 nm and about 620 nm and has a peak transmittance at 450 nm.
  • the prefe ⁇ ed blue filter comprises a combination of a blue component filter and a yellow component filter.
  • the blue component filter transmits light having a wavelength between about 320 nm and about 620 nm and has a peak transmittance at 450 nm.
  • the yellow component filter transmits light having a wavelength between about 415 nm and about 2800 nm.
  • suitable filters are UG1 (UV filter), BG23 (blue bandpass filter), and GG420 (yellow longpass filter), all manufactured by Schott.
  • the UV sensor 2202 outputs an analog signal proportional to the amount of light incident thereon, and this signal is amplified by amplifier 2210 and fed to a microcontroller 2212.
  • the fluorescence sensor 2204 outputs an analog signal proportional to the amount of light incident thereon and this signal is amplified by amplifier 2214 and fed to a microcontroller 2212.
  • Analog-to-digital converters 2216 within the microcontroller 2212 convert the signals from the amplifiers 2210, 2214 to digital and these digital signals are processed by the software of the microcontroller 2212.
  • the UV sensor 2202 may be, for example, an ultraviolet enhanced photodiode sensitive to light having a wavelength of about 360 nm and the fluorescence sensor 2204 may be a blue enhanced photodiode sensitive to light having a wavelength of about 450 nm.
  • Such photodiodes are available from, for example, Advanced Photonix, Inc., Massachusetts.
  • the microcontroller 2212 may be, for example, a Motorola 68HC16.
  • the exact characteristics of the sensors 2202, 2204 and the filters 2206, 2208 including the wavelength transmittance ranges of the above filters are not as critical as the prevention of the fluorescence sensor from generating an output signal in response to ultraviolet light, and the prevention of the ultraviolet sensor from generating an output signal in response to visible light.
  • the authentication system may employ an ultraviolet sensor which is not responsive to light having a wavelength longer than 400 nm and/or a fluorescence sensor which is not responsive to light having a wavelength shorter than 400 nm.
  • Calibration potentiometers 2218, 2220 permit the gains of amplifiers 2210, 2214 to be adjusted to appropriate levels. Calibration may be performed by positioning a piece of white fluorescent paper on the transport plate 2114 so that it completely covers the rectangular opening 2116. The potentiometers 2218, 2220 may then be adjusted so that the output of the amplifiers 2210, 2214 is 5 volts.
  • Counterfeit bills in categories (1) and (2) may be detected by a cunency authenticator employing an ultraviolet light reflection test.
  • Counterfeit bills in category (3) may be detected by a cunency authenticator employing both an ultraviolet reflection test and a fluorescence test.
  • Only counterfeits in category (4) are not detected by the authenticating methods of the present invention. Fluorescence is determined by any signal that is above the noise floor. Thus, the amplified fluorescent sensor signal 2222 will be approximately 0 volts for genuine U.S. cunency and will vary between approximately 0 and 5 volts for counterfeit bills, depending upon their fluorescence characteristics. Accordingly, an authenticating system will reject bills when signal 2222 exceeds approximately 0 volts.
  • a high level of reflected UV light (“high UV”) is indicated when the amplified UV sensor signal 2224 is above a predetermined threshold.
  • the high/low UV threshold is a function of lamp intensity and reflectance. Lamp intensity can degrade by as much as 50% over the life of the lamp and can be further attenuated by dust accumulation on the lamp and the sensors. The problem of dust accumulation is mitigated by enclosing the lamp and sensors in a housing as discussed above.
  • the authenticating system tracks the intensity of the UV light source and readjusts the high/low threshold accordingly. The degradation of the UV light source may be compensated for by periodically feeding a genuine bill into the system, sampling the output of the UV sensor, and adjusting the threshold accordingly.
  • degradation may be compensated for by periodically sampling the output of the UV sensor when no bill is present in the rectangular opening 2116 of the transport plate 2114. It is noted that a certain amount of UV light is always reflected off the acrylic window 2118.
  • the system can compensate for light source degradation.
  • sampling can also be used to indicate when the ultraviolet light source has burned out or otherwise requires replacement. This may be accomplished, for example, by means of a display reading or an illuminated light emitting diode (“LED").
  • the amplified ultraviolet sensor signal 2224 will initially vary between 1.0 and 5.0 volts depending upon the UV reflectance characteristics of the document being scanned and will slowly drift downward as the light source degrades.
  • the sampling of the UV sensor output may be used to adjust the gain of the amplifier 2210, thereby maintaining the output of the amplifier 2210 at its initial levels.
  • a 2-to-l ratio is used to discriminate between genuine and counterfeit bills. For example, if a genuine U.S. bill generates an amplified UV output sensor signal 2224 of 4.0 volts, documents generating an amplified UV output sensor signal 2224 of 2.0 volts or less will be rejected as counterfeit. As described above, this threshold of 2.0 volts may either be lowered as the light source degrades or the gain of the amplifier 2210 may be adjusted so that 2.0 volts remains an appropriate threshold value.
  • the determination of whether the level of UV reflected off a document is high or low is made by sampling the output of the UV sensor at a number of intervals, averaging the readings, and comparing the average level with the predetermined high/low threshold. Altematively, a comparison may be made by measuring the amount of UV light reflected at a number of locations on the bill and comparing these measurements with those obtained from genuine bills. Altematively, the output of one or more UV sensors may be processed to generate one or more patte s of reflected UV light and these patte s may be compared to the pattems generated by genuine bills.
  • FIGS. 60-63 illustrate a disc-type coin sorter that uses a coin-driving member having a resilient surface for moving coins along a metal coin-guiding surface of a stationary coin-guiding member.
  • the coin-driving member is a rotating disc
  • the coin-guiding member is a stationary sorting head.
  • a hopper 1510 receives coins of mixed denominations and feeds them through central openings in a housing 1511 and a coin-guiding member in the form of an annular sorting head or guide plate 1512 inside or underneath the housing. As the coins pass through these openings, they are deposited on the top surface of a coin-driving member in the form of a rotatable disc 1513.
  • This disc 1513 is mounted for rotation on a stub shaft (not shown) and driven by an electric motor 1514 mounted to a base plate 1515.
  • the disc 1513 comprises a resilient pad 1516 bonded to the top surface of a solid metal disc 1517.
  • the top surface of the resilient pad 1516 is preferably spaced from the lower surface of the sorting head 1512 by a gap of about 0.005 inches (0J3 mm).
  • the gap is set around the circumference of the sorting head 1512 by a three point mounting anangement including a pair of rear pivots 1518, 1519 loaded by respective torsion springs 1520 which tend to elevate the forward portion of the sorting head.
  • the forward portion of the sorting head 1512 is held in position by a latch 1522 which is pivotally mounted to the frame 1515 by a bolt 1523.
  • the latch 1522 engages a pin 1524 secured to the sorting head.
  • the latch For gaining access to the opposing surfaces of the resilient pad 1516 and the sorting head, the latch is pivoted to disengage the pin 1524, and the forward portion of the sorting head is raised to an upward position (not shown) by the torsion springs 1520.
  • the coins 1525 deposited on the top surface thereof tend to slide outwardly over the surface of the pad due to centrifugal force.
  • the coins 1525 for example, are initially displaced from the center of the disc 1513 by a cone 1526, and therefore are subjected to sufficient centrifugal force to overcome their static friction with the upper surface of the disc.
  • the coins are sorted into their respective denominations, and the coins for each denomination issue from a respective exit slot, such as the slots 1527, 1528, 1529, 1530, 1531 and 1532 (see FIGS. 60 and 61) for dimes, pennies, nickels, quarters, dollars, and half-dollars, respectively.
  • the coins for any given cunency are sorted by the variation in diameter for the various denominations.
  • the aligning, referencing, sorting, and ejecting operations are performed when the coins are pressed into engagement with the lower surface of the sorting head 1512.
  • the distance between the lower surfaces of the sorting head 1512 with the passages conveying the coins and the upper surface of the rotating disc 1513 is less than the thickness of the coins being conveyed.
  • positive control permits the coin sorter to be quickly stopped by braking the rotation of the disc 1513 when a preselected number of coins of a selected denomination have been ejected from the sorter.
  • Positive control also permits the sorter to be relatively compact yet operate at high speed.
  • the positive control permits the single file stream of coins to be relatively dense, and ensures that each coin in this stream can be directed to a respective exit slot.
  • FIG. 61 there is shown a bottom view of the prefened sorting head 1512 including various channels and other means especially designed for high-speed sorting with positive control of the coins, yet avoiding the galling problem. It should be kept in mind that the circulation of the coins, which is clockwise in FIG. 60, appears counterclockwise in FIG. 61 because FIG. 61 is a bottom view.
  • the various means operating upon the circulating coins include an entrance region 1540, means 1541 for stripping "shingled" coins, means 1542 for selecting thick coins, first means 1544 for recirculating coins, first referencing means 1545 including means 1546 for recirculating coins, second referencing means 1547, and the exit means 1527, 1528, 1529, 1530, 1531 and 1532 for six different coin denominations, such as dimes, pennies, nickels, quarters, dollars and half-dollars.
  • the lowermost surface of the sorting head 1512 is indicated by the reference numeral 1550.
  • the outwardly moving coins initially enter under a semi-annular region underneath a planar surface 1561 formed in the underside of the guide plate or sorting head 1512.
  • Coin Cl superimposed on the bottom plan view of the guide plate in FIG. 61 is an example of a coin which has entered the entrance region 1540. Free radial movement of the coins within the entrance region 1540 is terminated when they engage a wall 1562, though the coins continue to move circumferentially along the wall 1562 by the rotational movement of the pad 1516, as indicated by the central a ⁇ ow in the counterclockwise direction in FIG. 61.
  • the planar region 1561 is provided with an inclined surface 1541 forming a wall or step 1563 for engaging the upper most coin in a shingled pair.
  • an upper coin C2 is shingled over a lower coin C3.
  • movement of the upper coin C2 is limited by the wall 1563 so that the upper coin C2 is forced off of the lower coin C3 as the lower coin is moved by the rotating disc 1513.
  • the circulating coins in the entrance region 1540 are next directed to the means 1542 for selecting thick coins.
  • This means 1542 includes a surface 1564 recessed into the sorting head 1512 at a depth of 0.070 inches (1.78 mm) from the lowermost surface 1550 of the sorting head.
  • a step or wall 1565 is formed between the surface 1561 of the entrance region 1540 and the surface 1564.
  • the distance between the surface 1564 and the upper surface of the disc 1513 is therefore about 0.075 inches so that relatively thick coins between the surface 1564 and the disc 1513 are held by pad pressure.
  • an initial portion of the surface 1564 is formed with a ramp 1566 located adjacent to the wall 1562. Therefore, as the disc 1513 rotates, thick coins in the entrance region that are next to the wall 1562 are engaged by the ramp 1566 and thereafter their radial position is fixed by pressure between the disc and the surface 1564.
  • Thick coins which fail to initially engage the ramp 1566, however, engage the wall 1565 and are therefore recirculated back within the central region of the sorting head. This is illustrated, for example, in ?IG. 63 for the coin C4. This initial selecting and positioning of the thick coins prevents misaligned thick coins from hindering the flow of coins to the first referencing means 1545.
  • the ramp 1566 in the means 1542 for selecting the thick coins can also engage a pair or stack of thin coins.
  • a stack or pair of thin coins will be carried under pad pressure between the surface 1564 and the rotating disc 1513.
  • a pair of stacked coins will have its radial position fixed and will be carried toward the first referencing means 1545.
  • the first means 1545 for referencing the coins obtains a single-file stream of coins directed against the outer wall 1562 and leading up to a ramp 1573.
  • Coins are introduced into the referencing means 1545 by the thinner coins moving radially outward via centrifugal force, or by the thicker coin(s) C52a following concentricity via pad pressure.
  • the stacked coins C58a and C50a are separated at the inner wall 1582 such that the lower coin C58a is canied against surface 1572a.
  • the progression of the lower coin C58a is depicted by its positions at C58b, C58c, C58d, and C58e. More specifically, the lower coin C58 becomes engaged between the rotating disc 1513 and the surface 1572 in order to carry the lower coin to the first recirculating means 1544, where it is recirculated by the wall 1575 at positions C58d and C58e.
  • a ramp 1590 is used to recycle coins not fully between the outer and inner walls 1562 and 1582 and under the sorting head 1512. As shown in FIG. 61, no other means is needed to provide a proper introduction of the coins into the referencing means 1545.
  • the referencing means 1545 is further recessed over a region 1591 of sufficient length to allow the coins C54 of the widest denomination to move to the outer wall 1562 by centrifugal force. This allows coins C54 of the widest denomination to move freely into the referencing means 1545 toward its outer wall 1562 without being pressed between the resilient pad 1516 and the sorting head 1512 at the ramp 1590.
  • the inner wall 1582 is preferably constructed to follow the contour of the recess ceiling.
  • the region 1591 of the referencing recess 1545 is raised into the head 1512 by ramps 1593 and 1594, and the consistent contour at the inner wall 1582 is provided by a ramp 1595.
  • the first referencing means 1545 is sufficiently deep to allow coins C50 having a lesser thickness to be guided along the outer wall 1562 by centrifugal force, but sufficiently shallow to permit coins C52, C54 having a greater thickness to be pressed between the pad 1516 and the sorting head 1512, so that they are guided along the inner wall 1582 as they move through the referencing means 1545.
  • the referencing recess 1545 includes a section 1596 which bends such that coins C52, which are sufficiently thick to be guided by the inner wall 1582 but have a width which is less than the width of the referencing recess 1545, are carried away from the inner wall 1582 from a maximum radial location 1583 on the inner wall toward the ramp 1573.
  • This configuration in the sorting head 1512 allows the coins of all denominations to converge at a nanow ramped finger 1573a on the ramp 1573, with coins C54 having the largest width being carried between the inner and outer walls via the surface 1596 to the ramped finger 1573a so as to bring the outer edges of all coins to a generally common radial location.
  • the coins C50 radially inward along the latter portion of the outer wall 1562, the probability of coins being offset from the outer wall 1562 by adjacent coins and being led onto the ramped finger 1573a is significantly reduced.
  • any coins C50 which are slightiy offset from the outer wall 1562 while being led onto the ramp finger 1573a may be accommodated by moving the edge 1551 of exit slot 1527 radially inward, enough to increase the width of the slot 1527 to capture offset coins C50 but to prevent the capture of coins of the larger denominations.
  • the width of the ramp finger 1573a may be about 0J40 inch.
  • the coins become firmly pressed into the pad 16 and are carried forward to the second referencing means 1547.
  • a coin such as the coin C50c will be carried forward to the seconci referencing means 1547 so long as a portion of the coin is engaged by the nanow ramped finger 1573a on the ramp 1573. If a coin is not sufficiently close to the wall 1562 so as to be engaged by this ramped finger 1573a, then the coin strikes a wall 1574 defined by the second recirculating means 1546, and that coin is recirculated back to the entrance region 1540.
  • the first recirculating means 1544, the second recirculating means 1546 and the second referencing means 1547 are defined at successive positions in the sorting head 1512. It should be apparent that the first recirculating means 1544, as well as the second recirculating means 1546, recirculate the coins under positive control of pad pressure.
  • the second referencing means 1547 also uses positive control of the coins to align the outer most edge of the coins with a gaging wall 1577.
  • the second referencing means 1547 includes a surface 1576, for example, at 0J 10 inches (1.27 mm) from the bottom surface of the sorting head 1512, and a ramp 1578 which engages the inner edge portions of the coins, such as the coin C50d.
  • the initial portion of the gaging wall 1577 is along a spiral path with respect to the center of the sorting head 1512 and the sorting disc 1513, so that as the coins are positively driven in the circumferential direction by the rotating disc 1513, the outer edges of the coins engage the gaging wall 1577 and are forced slightly radially inward to a precise gaging radius, as shown for the coin C16 in FIG. 62.
  • FIG. 62 further shows a coin C17 having been ejected from the second recirculating means 1546.
  • the second referencing means 1547 terminates with a slight ramp 1580 causing the coins to be firmly pressed into the pad 1516 on the rotating disc with their outer most edges aligned with the gaging radius provided by the gaging wall 1577.
  • the coins are gripped between the guide plate 1512 and the resilient pad 1516 with the maximum compressive force. This ensures that the coins are held securely in the new radial position determined by the wall 1577 of the second referencing means 1547.
  • the sorting head 1512 further includes sorting means comprising a series of ejection recesses 1527, 1528, 1529, 1530, 1531 and 1532 spaced circumferentially around the outer periphery of the plate, with the innermost edges of successive slots located progressively farther away from the common radial location of the outer edges o: all the coins for receiving and ejecting coins in order of increasing diameter.
  • each ejection recess is slightly larger than the diameter of the coin to be received and ejected by that particular recess, and the surface of the guide plate adjacent the radially outer edge of each ejection recess presses the outer portions of the coins received by that recess into the resilient pad so that the inner edges of those coins are tilted upwardly into the recess.
  • the ejection recesses extend outwardly to the periphery of the guide plate so that the inner edges of these recesses guide the tilted coins outwardly and eventually eject those coins from between the guide plate 1512 and the resilient pad 1516.
  • the innermost edges of the ejection recesses are positioned so that the inner edge of a coin of only one particular denomination can enter each recess; the coins of all other remaining denominations extend inwardly beyond the innermost edge of that particular recess so that the inner edges of those coins cannot enter the recess.
  • the first ejection recess 1527 is intended to discharge only dimes, and thus the innermost edge 1551 of this recess is located at a radius that is spaced inwardly from the radius of the gaging wall 1577 by a distance that is only slightly greater than the diameter of a dime. Consequently, only dimes can enter the recess 1527.
  • the inner edges of the pennies, nickels, quarters, dollars and half dollars all extend inwardly beyond the innermost edge of the recess 1527, thereby preventing these coins from entering that particular recess.
  • the inner edges of only pennies are located close enough to the periphery of the sorting head 1512 to enter the recess.
  • the inner edges of all the larger coins extend inwardly beyond the innermost edge 1552 of the recess 1528 so that they remain gripped between the guide plate and the resilient pad. Consequently, all the coins except the pennies continue to be rotated past the recess 1528.
  • One of six proximity sensors S S 6 is mounted along the outboard edge of each of the six exit channels 1527-1532 in the sorting head for sensing and counting coins passing through the respective exit channels.
  • each sensor is dedicated to one particular denomination of coin, and thus it is not necessary to process the sensor output signals to determine the coin denomination.
  • the effective fields of the sensors S,-S 6 are all located just outboard of the radius at which the outer edges of all coin denominations are gaged before they reach the exit channels 1527-1532, so that each sensor detects only the coins which enter its exit channel and does not detect the coins which bypass that exit channel. Only the largest coin denomination (e.g., U.S.
  • each of the exit channels 1527-1532 also includes one of six coin discrimination sensors D1-D6. These sensors D1-D6 are the eddy cunent sensors, and will be described in more detail below in connection with FIGS. 64-67 of the drawings.
  • the disc When one of the discrimination sensors detects a coin material that is not the proper material for coins in that exit channel, the disc may be stopped by de-energizing or disengaging the drive motor and energizing a brake. The suspect coin may then be discharged by jogging the drive motor with one or more electrical pulses until the trailing edge of the suspect coin clears the exit edge of its exit channel.
  • the exact disc movement required to move the trailing edge of a coin from its sensor to the exit edge of its exit channel can be empirically determined for each coin denomination and then stored in the memory of the control system.
  • An encoder on the sorter disc can then be used to measure the actual disc movement following the sensmg of the suspect coin, so that the disc can be stopped at the precise position where the suspect coin clears the exit edge of its exit channel, thereby ensuring that no coins following the suspect coin are discharged.
  • one embodiment of the present invention employs an eddy cunent sensor 1710 to perform as the coin handling system's coin discrimination sensors D1-D6.
  • the eddy cunent sensor 1710 includes an excitation coil 1712 for generating an alternating magnetic field used to induce eddy cunents in a coin 1714.
  • the excitation coil 1712 has a start end 1716 and a finish end 1718.
  • An embodiment an a-c. excitation coil voltage V e ⁇ e.g. , a sinusoidal signal of 250 KHz and 10 volts peak-to-peak, is applied across the start end 1716 and the finish end 1718 of the excitation coil 1712.
  • the alternating voltage V £ ⁇ produces a conesponding cunent in the excitation coil 1712 which in turn produces a conesponding alternating magnetic field.
  • the alternating magnetic field exists within and around the excitation coil 1712 and extends outwardly to the coin 1714.
  • the magnetic field penetrates the coin 1714 as the coin is moving in close proximity to the excitation coil 1712, and eddy cunents are induced in the coin 1714 as the coin moves through the alternating magnetic field.
  • the eddy cunents themselves also produce a conesponding magnetic field.
  • a proximal detector coil 1722 and a distal coil 1724 are disposed above the coin 1714 so that the eddy cunent-generated magnetic field induces voltages upon the coils 1722, 1724.
  • the distal detector coil 1724 is positioned above the coin 1714, and the proximal detector coil 1722 is positioned between the distal detector coil 1724 and the passing coin 1714.
  • the excitation coil 1712, the proximal detector coil 1722 and the distal detector coil 1724 are all wound in the same direction (either clockwise or counterclockwise).
  • the proximal detection coil 1722 and the distal detector coil 1724 are wound in the same direction so that the voltages induced on these coils by the eddy cunents are properly oriented.
  • the proximal detection coil 1722 has a starting end 1726 and a finish end 1728.
  • the distal coil 1724 has a starting end 1730 and a finish end 1632.
  • the detector coils 1722, 1724 are positioned as follows: finish end 1728 of the proximal detector coil 1722, start end 1726 of the proximal detector coil 1722, finish end 1732 of the distal detector coil 1724 and start end 1730 of the distal detector coil 1724.
  • the finish end 1728 of the proximal detection coil 1722 is connected to the finish end 1732 of the distal detector coil 1724 via a conductive wire 1734.
  • the proximal detection coil 1722 is wound in the opposite direction of the distal detection coil 1724.
  • the start end 1726 of the proximal coil 1722 is connected to the finish end 1732 of the distal coil 1724.
  • Eddy cunents in fhe coin 1714 induce voltages V_ ro ⁇ and V ⁇ j . respectively on the detector coils 1722, 1724.
  • the excitation coil 1712 also induces a common-mode voltage V com on each of the detector coils 1722, 1724.
  • the common-mode voltage V com is effectively the same on each detector coil due to the symmetry of the detector coils' physical a ⁇ angement within the excitation coil 1712.
  • the common-mode voltage V com induced by the excitation coil 1712 is subtracted out, leaving only a difference voltage V jjj f co ⁇ esponding to the eddy cunents in the coin 1714. This eliminates the need for additional circuitry to subtract out the common-mode voltage V com .
  • the common-mode voltage V com is effectively subtracted out because both the distal detection coil 1724 and the proximal detection coil 1722 receive the same level of induced voltage V com from the excitation coil 1712.
  • the voltages induced by the eddy cunent in the detector coils are not effectively the same. This is because the proximal detector coil 1722 is purposely positioned closer to the passing coin than the distal detector coil 1724. Thus, the voltage induced in the proximal detector coil 1722 is significantly stronger, i.e. has greater amplitude, than the voltage induced in the distal detector coil 1724. Although the present invention subtracts the eddy cunent- induced voltage on the distal coil 1724 from the eddy cunent-induced voltage on the proximal coil 1722, the voltage amplitude difference is sufficientiy great to permit detailed resolution of the eddy cunent response.
  • the excitation coil 1712 is radially sunounded by a magnetic shield 1734.
  • the magnet shield 1734 has a high level of magnetic permeability in order to help contain the magnetic field sunounding the excitation coil 1712.
  • the magnetic shield 1734 has the advantage of preventing stray magnetic field from interfering with other nearby eddy cunent sensors.
  • the magnetic shield is itself radially sunounded by a steel outer case 1736.
  • the excitation coil utilizes a cylindrical ceramic (e.g. , alumina) core 1738.
  • alumina has the advantages of being impervious to humidity and providing a good wear surface. It is desirable that the core 1748 be able to withstand wear because it may come into frictional contact with the coin 1714. Alumina withstands frictional contact well because of its high degree of hardness, i.e. , approximately 9 on mohs scale.
  • the detection coils 1722, 1724 are wound on a coil form (not shown).
  • a prefened form is a cylinder having a length of 0.5 inch, a maximum diameter of 0.2620 inch, a minimum diameter of 0J660 inch, and two grooves of 0.060 inch width spaced apart by 0.060 inch and spaced from one end of the form by 0.03 inch.
  • Both the proximal detection coil 1722 and the distal detector coil 1724 have 350 turns of #44 AWG enamel covered magnet wire layer wound to generally uniformly fill the available space in the grooves.
  • Each of the detector coils 1722, 1724 are wound in the same direction with the finish ends 1728, 1732 being connected together by the conductive wire 1734.
  • the start ends 1726, 1730 of the detector coils 1722, 1724 are connected to separately identified wires in a connecting cable.
  • the excitation coil 1712 is a generally uniformly layer wound on a cylindrical alumina ceramic coil form having a length of 0.5 inch, an outside diameter of 0.2750 inch, and a wall thickness of 0.03125 inch.
  • the excitation coil 1712 is wound with 135 turns of #42 AWG enamel covered magnet wire in the same direction as the detector coils 1722, 1724.
  • the excitation coil voltage V g ⁇ is applied across the start end 1716 and the finish end 1718. After the excitation coil 1712 and detector coils 1722, 1724 are wound, the excitation coil 1712 is slipped over the detector coils 1722, 1724 around a common center axis.
  • the sensor 1710 is connected to a test oscillator (not shown) which applies the excitation voltage V £ ⁇ to the excitation coil 1712.
  • the excitation coil's position is adjusted along the axis of the coil to give a null response from the detector coils 1722, 1724 on an a-c. voltmeter with no metal near the coil windings.
  • the magnetic shield 1644 is the slipped over the excitation coil 1712 and adjusted to again give a null response from the detector coils 1722, 1724.
  • the magnetic shield 1744 and coils 1712, 1722, 1724 within the magnetic shield 1744 are then placed in the steel outer case 1746 and encapsulated with a polymer resin (not shown) to "freeze" the position of the magnetic shield 1744 and coils 1712, 1722, 1724.
  • an end of the eddy cunent sensor 1710 nearest the proximal detector coil 1722 is sanded and lapped to produce a flat and smooth surface with the coils 1712, 1722 slightly recessed within the resin.
  • it is prefened to use a combination of phase and amplitude analysis of the detected voltage. This type of analysis minimizes the effects of variations in coin surface geometry and in the distance between the coin and the coils.
  • the voltage applied to the excitation coil 1712 causes cu ⁇ ent to flow in the coil 1712 which lags behind the voltage 1720.
  • the cunent may lag the voltage 1720 by 90 degrees in a superconductive coil.
  • the coin's 1714 eddy cunents impose a resistive loss on the cunent in the excitation coil 1712. Therefore, the initial phase difference between the voltage and cunent in the excitation coil 1712 is decreased by the presence of the coin 1714.
  • the detector coils 1724, 1726 have a voltage induced upon them, the phase difference between the voltage applied to the excitation coil 1712 and that of the detector coils is reduced due to the jm eddy cunent effect in the coin.
  • the amount of reduction in the phase difference is proportional to the electrical and magnetic characteristics of the coin and thus the composition of the coin.
  • FIGS. 67A and 67B illustrate a prefened phase-sensitive detector 1750 for sampling the differential output signal V ⁇ from the two detector coils 1722, 1724.
  • the differential output signal is passed through a buffer amplifier 252 to a switch 1754, where the buffered V ⁇ j ⁇ is sampled once per cycle by momentarily closing the switch 1754.
  • the switch 1754 is controlled by a series of reference pulses produced from the V £ ⁇ signal, one pulse per cycle.
  • the reference pulses 1758 are synchronized with excitation voltage V e ⁇ .
  • the amplitude of the differential output signal during the sampling interval is a function not only of the amplitude of the detector coil voltages 1736, 1738, but also of the phase difference between the signals in excitation coil 1712 and the detection coils 1736, 1738.
  • the pulses derived from V £ ⁇ are delayed by an "offset angX' which can be adjusted to minimize the sensitivity of V ⁇ to variations in the gap between the proximal face of the sensor 1710 and the surface of the coin 1714 being sensed.
  • the value of the offset angle for any given coin can be determined empirically by moving a standard metal disc, made of the same material as the coin 1714, from a position where it contacts the sensor face, to a position where it is spaced about 0.001 to 0.020 inch from the sensor face.
  • the signal sample from the detector 1750 is measured at both positions, and the difference between the two measurements is noted. This process is repeated at several different offset angles to determine the offset angle which produces the minimum difference between the two measurements.
  • Each time buffered is sampled the resulting sample is passed through a second buffer amplifier 1756 to an analog-to-digital converter (not shown).
  • the resulting digital value is supplied to a microprocessor (not shown) which compares that value with several different ranges of values stored in a lookup table (not shown).
  • Each stored range of values conesponds to a particular coin material, and thus the coin material represented by any given sample value is determined by the particular stored range into which the sample value falls.
  • the stored ranges of values can be determined empirically by simply measuring a batch of coins of each denomination and storing the resulting range of values measured for each denomination.
  • the coin sorting and counting module 8 may be replaced with a coin discriminating module which does not sort the coins.
  • a coin discriminating module which does not sort the coins.
  • Such a module would align the coins of all denominations in a single file and guide them past a single coin discrimination sensor to dete ⁇ nine whether the coins are genuine. The coins of all denominations would then be discharged into a single storage receptacle and sorted at a later time. Coins that are detected to be non-genuine would be diverted and retumed to the customer at the coin retum station 4.
  • the curved exit chute 1800 includes two slots 1802, 1804 separated by an intemal partition 1806.
  • the intemal partition 1806 is pivotally mounted to a stationary base 1808 so that the intemal partition 1806 may be moved, perpendicular to the plane of the coins, by an actuator 1810 between an up position (FIG. 70) and a down position (FIG. 69).
  • the exit chute 1800 is positioned adjacent an exit channel of the coin sorter such that coins exiting the coin sorter are guided into the slot 1802 when the intemal partition 1806 is in the down position (FIG. 69).
  • the actuator 1810 moves the intemal partition 1806 to the up position (FIG. 66) so that the invalid coin now enters the slot 1804 of the exit chute 1800.
  • Coins entering the slot 1804 are discharged into the tube 9 that conveys those coins to the coin-return slot 4 at the front of the ATM . While FIGS. 67-70 illustrate only a single exit chute, it will be apparent that a similar exit chute is provided at each of the six coin exit locations around the circumference of the sorting disc.
  • the actuator 1810 moves the intemal partition 1806 between the up and down positions in response to detection of invalid and valid coins.
  • the intemal partition 1806 is in the down position and an invalid coin is detected, the partition 1806 is moved to the up position so that the invalid coin will be diverted into the slot 1804.
  • an invalid coin may be separated from the valid coins by use of inboard actuators in the sorting head, activated by signals derived from one or more sensors mounted in the sorting head upstream of the actuators. Such an anangement is described in U.S. Patent No. 5,299,977, which is incorporated herein by reference.

Abstract

An apparatus for currency discrimination comprises first and second stationary scanheads (18a, 18b), disposed on opposite sides of a bill transport path, for scanning respective first and second opposing surfaces of a bill (17) traveling along the bill transport path and for producing respective output signals. The bill (17) travels along the transport path in the direction of a predetermined dimension of the bill (17). A memory (34) stores master characteristic patterns corresponding to associated predetermined surfaces of a plurality of denominations of genuine bills. Sampling circuitry (30, 32) samples the output signals associated with the respective first and second opposing surfaces of the scanned bill (17). A signal processor (30) is programmed to determine which one of the first and second opposing surfaces corresponds to the associated predetermined surfaces of the plurality of denominations of genuine bills. The processor (30) then correlates the output signal associated with the one of the first and second opposing surfaces corresponding to the associated predetermined surfaces with the master characteristic patterns to identify the denomination of the scanned bill (17).

Description

AUTOMATIC CURRENCY PROCESSING SYSTEM
Cr oss-Reference to Related Applications
This application is a continuation-in-part of the following pending United States patent applications:
Serial No. 08/399,854 filed March 7, 1995 for a "Method and Apparatus For Discriminating and Counting Documents";
Serial No. 08/394,752 filed Febniary 27, 1995 for a "Method of Generating Modified Patterns and Method and Apparatus for Using the Same in a Currency Identification System";
Serial No. 08/362,848 filed December 22, 1994, for a "Method And Apparatus For Discriminating and Counting Documents" ;
Serial No. 08/340,031 filed November 14, 1994, for a "Method And Apparatus For Discriminating and Counting Documents";
Serial No. 08/317,349 filed October 4, 1994, for a "Method And Apparatus For Authenticating Documents Including Currency"; Serial No. 08/287,882 filed August 9, 1994 for a "Method and Apparatus for
Document Identification";
Serial No. 08/243,807 filed May 16, 1994, for "Method And Apparatus For Currency Discrimination";
Serial No. 08/226,660 filed April 12, 1994, for "Method And Apparatus For Currency Discrimination";
Serial No. 08/219,093 filed on March 29, 1994, for a "Currency Discriminator and Authenticator"; and
Serial No. 08/207,592 filed March 8, 1994 for "Method and Apparatus for Currency Discrimination". United States patent application Serial No. 08/399,854 filed March 7, 1995 for a "Method and Apparatus For Discriminating and Counting Documents" is a continuation-in-part of pending United States patent applications Serial No. 08/394,752 filed February 27, 1995 for a "Method of Generating Modified Patterns and Method and Apparatus for Using the Same in a Currency Identification System," Serial No. 08/340,031, and Serial No. 08/287,882. United States patent application Serial No. 08/394,752 filed February 27, 1995 for a "Method of Generating Modified Patterns and Method and Apparatus for Using the Same in a Currency Identification System" is a continuation-in-part of pending United States patent applications Serial No. 08/340,031 and Serial No. 08/127,334 filed September 27, 1993, for a "Method and Apparatus for Currency Discrimination and Counting."
United States patent application Serial No. 08/362,848 filed December 22, 1994, for a "Method And Apparatus For Discriminating and Counting Documents" is a continuation-in-part of pending United States patent application Serial No. 08/340,031.
United States patent application Serial No. 08/340,031 filed November 14, 1994, for a "Method And Apparatus For Discriminating and Counting Documents" is a continuation-in-part of pending United States patent applications Serial No. 08/243,807 and Serial No. 08/207,592. United States patent application Serial No. 08/287,882 filed August 9, 1994 for a "Method and Apparatus for Document Identification" is a continuation-in-part of pending United States patent applications Serial No. 08/207,592, Serial No. 08/219,093, and Serial No. 08/127,334 filed September 27, 1993, for a "Method and Apparatus for Currency Discrimination and Counting." United States patent application Serial No. 08/243,807 filed May 16, 1994, for "Method And Apparatus For Currency Discrimination" is a continuation-in-part of pending United States patent applications Serial No. 08/219,093 and Serial No. 08/127,334.
United States patent application Serial No. 08/226,660 filed April 12, 1994, for "Method And Apparatus For Currency Discrimination" is a continuation-in-part of pending United States patent application Serial No. 08/127,334.
United States patent application Serial No. 08/219,093 filed on March 29, 1994, for a "Currency Discrirninator and Authenticator" is a continuation-in-part of pending United States patent application Serial No. 08/127,334. United States patent application Serial No. 08/207,592 filed March 8, 1994 for "Method and Apparatus for Currency Discrimination" is a continuation-in-part of pending United States patent application Serial No. 08/127,334. United States patent application Serial No. 08/127,334 filed September 27, 1993, for a "Method and Apparatus for Currency Discrimination and Counting," is a continuation of United States patent application Serial No. 07/885,648, filed on May 19, 1992, and issued as United States Patent No. 5,295, 196, for a "Method and Apparatus for Currency Discrimination and Counting," which is a continuation-in- part of abandoned United States patent application Serial No. 07/475,111, filed February 5, 1990, for a "Method and Apparatus for Currency Discrimination and Counting. "
United States patent application Serial No. 08/201,350 filed February 24, 1994, for "Coin Handling System With Shunting Mechanism," is a continuation-in- part of copending U.S. patent application Serial No. 08/149,660, filed November 9, 1993, and entitled "Coin Handling System With Coin Sensor Discriminator", which is in turn a continuation-in-part of copending U.S. patent application Serial No. 08/115,319, filed September 1, 1993, and entitled "Coin Handling System With Controlled Coin Discharge" , which is in turn a continuation-in-part of copending
U.S. patent application Serial No. 07/951,731, filed September 25, 1992 (allowed on October 1, 1993), and entitled "Coin Handling System," which is in turn a continuation-in-part of copending U.S. patent application Serial No. 07/904,161 filed August 21 , 1992 (now issued as U.S. patent number 5,277,651), and entitled "Coin Sorter with Automatic Bag-Switching or Stopping," which in turn is a continuation of U.S. patent application Serial No. 07/524,134 filed May 14, 1990 (now issued as U.S. patent number 5,141,443), and entitled "Coin Sorter With Automatic Bag- Switching Or Stopping."
United States patent application Serial No. 08/399,771 filed March 7, 1995, for "Improved Coin Discrimination Sensor And Coin Handling System."
Field of the Invention
The present invention relates to currency processing systems such as automatic teller machines and currency redemption machines. Summary Of The Invention
The primary object of the present invention is to provide an improved automatic teller machine ("ATM") or currency redemption machine that is capable of processing cash deposits as well as withdrawals. Another object of this invention is to provide such machines that are capable of accepting and dispensing coins as well as bills.
A further object of this invention is to provide such machines that automatically evaluate the authenticity, as well as the denomination, of the cash that is deposited, whether in the form of bills or coins. Still another object of the invention is to provide such machines that are coupled to the cash accounting system of a bank or other financial institution so that the customer's account can be immediately credited with verified cash deposit amounts.
In accordance with the present invention, the foregoing objectives are realized by providing a currency processing machine for receiving and dispensing cash and substantially immediately furnishing an associated cash accounting system with data, including the value of the currency processed, for each transaction. The machine includes a bill dispenser having a bill storage device and controllable transport means for dispensing selected numbers of bills from the storage device, a bill receptacle for receiving stacks of bills to be deposited, and a bill counter and scanner for rapidly removing the bills one at a time from the receptacle and counting the bills while determining the denomination of each bill. The counter and scanner also generates data representing the denomination of each bill, and the number of bills of each denomination, passed through the counter and scanner. A memory receives and stores data representing the number of bills of each denomination passed through the counter and scanner in each transaction, and data representing the total value of the bills passed through the counter and scanner in each transaction. A control system transfers data from the memory to an associated cash accounting system so that the deposits and withdrawals executed at the currency processing machine are entered in the accounting system substantially immediately after the execution of those transactions. The preferred control system checks the genuineness of each bill and coin that is counted, and produces a control signal in response to the detection of a non-genuine bill or coin. The processing of the bill or coin detected to be non- genuine is altered in response to such control signals.
Brief Description Of The Drawings FIG. la is a perspective view of an automatic teller machine embodying the present invention;
FIG. lb is a diagrammatic side elevation of the machine of FIG. la; FIG. lc is a more detailed diagrammatic side elevation of the machine of FIG. la; FIG. ld is a flow chart illustrating the sequential procedure involved in the execution of a transaction in the machine of FIG. la;
FIG. le is a flow chart illustrating the sequential procedure involved in the execution of a deposit of bills in the machine of FIG. la;
FIG. lf is a flow chart illustrating an altemative sequential procedure involved in the execution of a deposit of bills in the machine of FIG. la;
FIG. 2a is a functional block diagram of the currency scanning and counting subassembly in the machine of FIG. 1, including a scanhead arranged on each side of a transport path;
FIG. 2b is a functional block diagram of a currency scanning and counting device that includes a scanhead arranged on a single side of a transport path;
FIG. 2c is a functional block diagram of a currency scanning and counting machine similar to that of FIG. 2b, but adapted to feed and scan bills along their wide dimension;
FIG. 2d is a functional block diagram of a currency scanning and counting device similar to those of FIGs. 2a-2c but including a second type of scanhead for detecting a second characteristic of the currency;
FIG. 3 is a diagrammatic perspective illustration of the successive areas scanned during the traversing movement of a single bill across an optical sensor according to a preferred embodiment of the primary scanhead; FIGs. 4a and 4b are perspective views of a bill and a preferred area to be optically scanned on the bill; FIGs. 5a and 5b are diagrammatic side elevation views of the preferred areas to be optically scanned on a bill according to a preferred embodiment of the invention;
FIG. 6a is a perspective view of a bill showing the preferred area of a first surface to be scanned by one of the two scanheads employed in the preferred embodiment of the present invention;
FIG. 6b is another perspective view of the bill in FIG. 6a showing the preferred area of a second surface to be scanned by the other of the scanheads employed in the preferred embodiment of the present invention; FIG. 6c is a side elevation showing the first surface of a bill scanned by an upper scanhead and the second surface of the bill scanned by a lower scanhead;
FIG. 6d is a side elevation showing the first surface of a bill scanned by a lower scanhead and the second surface of the bill scanned by an upper scanhead; FIGs. 7a and 7b form a block diagram illustrating a preferred circuit arrangement for processing and correlating reflectance data according to the optical sensing and counting technique of this invention;
FIGs. 8a and 8b comprise a flowchart illustrating the sequence of operations involved in implementing a discrimination and authentication system according to a preferred embodiment of the present invention; FIG. 9 is a flow chart illustrating the sequential procedure involved in detecting the presence of a bill adjacent the lower scanhead and the borderline on the side of the bill adjacent to the lower scanhead;
FIG. 10 is a flow chart illustrating the sequential procedure involved in detecting the presence of a bill adjacent the upper scanhead and the borderline on the side of the bill adjacent to the upper scanhead;
FIG. Ila is a flow chart illustrating the sequential procedure involved in the analog-to-digital conversion routine associated with the lower scanhead;
FIG. lib is a flow chart illustrating the sequential procedure involved in the analog-to-digital conversion routine associated with the upper scanhead; FIG. 12 is a flow chart illustrating the sequential procedure involved in determining which scanhead is scanning the green side of a U.S. currency bill; FIG. 13 is a flow chart illustrating the sequence of operations involved in deteπnining the bill denomination from the correlation results;
FIG. 14 is a flow chart illustrating the sequential procedure involved in decelerating and stopping the bill transport system in the event of an error; FIG. 15a is a graphical illustration of representative characteristic pattems generated by narrow dimension optical scanning of a $1 currency bill in the forward direction;
FIG. 15b is a graphical illustration of representative characteristic pattems generated by narrow dimension optical scanning of a $2 currency bill in the reverse direction;
FIG. 15c is a graphical illustration of representative characteristic pattems generated by narrow dimension optical scanning of a $100 currency bill in the forward direction;
FIG. 15d is a graph illustrating component pattems generated by scanning old and new $20 bills according a second method according to a preferred embodiment of the present invention;
FIG. 15e is a graph illustrating an pattem for a $20 bill scanned in the forward direction derived by averaging the pattems of FIG. 15d according a second method according to a preferred embodiment of the present invention; FIGs. 16a-16e are graphical illustrations of the effect produced on correlation pattem by using the progressive shifting technique, according to an embodiment of this invention;
FIGs. 17a-17c are a flowchart illustrating a preferred embodiment of a modified pattem generation method according to the present invention; FIG. 18a is a flow chart illustrating the sequential procedure involved in the execution of multiple correlations of the scan data from a single bill;
FIG. 18b is a flow chart illustrating a modified sequential procedure of that of FIG. 18a;
FIG. 19a is a flow chart illustrating the sequence of operations involved in determining the bill denomination from the correlation results using data retrieved from the green side of U.S. bills according to one prefened embodiment of the present invention; FIGs. 19b and 19c are a flow chart illustrating the sequence of operations involved in determining the bill denomination from the conelation results using data retrieved from the black side of U.S. bills;
FIG. 20a is an enlarged vertical section taken approximately through the center of the machine, but showing the various transport rolls in side elevation;
FIG. 20b is a top plan view of the interior mechanism of the machine of FIG. 1 for transporting bills across the optical scanheads, and also showing the stacking wheels at the front of the machine;
FIG. 21a is an enlarged perspective view of the bill transport mechanism which receives bills from the stripping wheels in the machine of FIG. 1;
FIG. 21b is a cross-sectional view of the bill transport mechanism depicted in FIG. 21 along line 21b;
FIG. 22 is a side elevation of the machine of FIG. 1, with the side panel of the housing removed; FIG. 23 is an enlarged bottom plan view of the lower support member in the machine of FIG. 1 and the passive transport rolls mounted on that member;
FIG. 24 is a sectional view taken across the center of the bottom support member of FIG. 23 across the narrow dimension thereof;
FIG. 25 is an end elevation of the upper support member which includes the upper scanhead in the machine of FIG. 1, and the sectional view of the lower support member mounted beneath the upper support member;
FIG. 26 is a section taken through the centers of both the upper and lower support members, along the long dimension of the lower support member shown in FIG. 23; FIG. 27 is a top plan view of the upper support member which includes the upper scanhead;
FIG. 28 is a bottom plan view of the upper support member which includes the upper scanhead;
FIG. 29 is an illustration of the light distribution produced about one of the optical scanheads; FIGs. 30a and 30b are diagrammatic illustrations of the location of two auxiliary photo sensors relative to a bill passed thereover by the transport and scanning mechanism shown in FIGs. 20a-28;
FIG. 31 is a flow chart illustrating the sequential procedure involved in a ramp-up routine for increasing the transport speed of the bill transport mechanism from zero to top speed;
FIG. 32 is a flow chart illustrating the sequential procedure involved in a ramp-to-slow-speed routine for decreasing the transport speed of the bill transport mechanism from top speed to slow speed; FIG. 33 is a flow chart illustrating the sequential procedure involved in a ramp-to-zero-speed routine for decreasing the transport speed of the bill transport mechanism to zero;
FIG. 34 is a flow chart illustrating the sequential procedure involved in a pause-after-ramp routine for delaying the feedback loop while the bill transport mechanism changes speeds;
FIG. 35 is a flow chart illustrating the sequential procedure involved in a feedback loop routine for monitoring and stabilizing the transport speed of the bill transport mechanism;
FIG. 36 is a flow chart illustrating the sequential procedure involved in a doubles detection routine for detecting overlapped bills;
FIG. 37 is a flow chart illustrating the sequential procedure involved in a routine for detecting sample data representing dark blemishes on a bill;
FIG. 38 is a flow chart illustrating the sequential procedure involved in a routine for maintaining a desired readhead voltage level; FIG. 39 is a top view of a bill and size determining sensors according to a prefened embodiment of the present invention;
FIG. 40 is a top view of a bill illustrating multiple areas to be optically scanned on a bill according to a prefened embodiment of the present invention; FIG. 41a is a graph illustrating a scanned patte which is offset from a conesponding master pattem;
FIG. 41b is a graph illustrating the same pattems of FIG. 41a after the scanned patte is shifted relative to the master pattem; FIG. 42 is a side elevation of a multiple scanhead anangement according to a prefened embodiment of the present invention;
FIG. 43 is a side elevation of a multiple scanhead arrangement according to another prefened embodiment of the present invention; FIG. 44 is a side elevation of a multiple scanhead arrangement according to another prefened embodiment of the present invention;
FIG. 45 is a side elevation of a multiple scanhead arrangement according to another prefened embodiment of the present invention;
FIG. 46 is a top view of a staggered scanhead arrangement according to a prefened embodiment of the present invention;
FIG. 47a is a top view of a linear array scanhead according to a prefened embodiment of the present invention illustrating a bill being fed in a centered fashion;
FIG. 47b is a side view of a linear array scanhead according to a prefened embodiment of the present invention illustrating a bill being fed in a centered fashion; FIG. 48 is a top view of a linear array scanhead according to another prefened embodiment of the present invention illustrating a bill being fed in a non- centered fashion;
FIG. 49 is a top view of a linear array scanhead according to another prefened embodiment of the present invention illustrating a bill being fed in a skewed fashion;
FIGs. 50a and 50b are a flowchart of the operation of a currency discrimination system according to a preferred embodiment of the present invention;
FIG. 51 is a top view of a triple scanhead arrangement utilized in a discriminating device able to discriminate both Canadian and German bills according to a prefened embodiment of the present invention;
FIG. 52 is a top view of Canadian bill illustrating the areas scanned by the triple scanhead arrangement of FIG. 51 according to a prefened embodiment of the present invention;
FIG. 53 is a flowchart of the threshold tests utilized in calling the denomination of a Canadian bill according to a prefened embodiment of the present invention; FIG. 54a illustrates the general areas scanned in generating master 10 DM German pattems according to a prefened embodiment of the present invention;
FIG. 54b illustrates the general areas scanned in generating master 20 DM, 50 DM, and 100 DM German pattems according to a prefened embodiment of the present invention;
FIG. 55 is a flowchart of the threshold tests utilized in calling the denomination of a German bill;
FIG. 56 is a functional block diagram illustrating a first embodiment of a document authenticator and discriminator; FIG. 57 is a functional block diagram illustrating a second embodiment of a document authenticator and discriminator;
FIG. 58a is a side view of a document authenticating system utilizing ultraviolet light;
FIG. 58b is a top view of the system of FIG. 58a along the direction 58b; FIG. 58c is a top view of the system of FIG. 58a along the direction 58c; and
FIG. 59 is a functional block diagram of the optical and electronic components of the document authenticating system of FIGS. 58a-58c.
FIG. 60 is perspective view of a disc-type coin sorter embodying the present invention, with a top portion thereof broken away to show intemal structure; FIG. 61 is an enlarged horizontal section taken generally along line 61-61 in
FIG. 60;
FIG. 62 is an enlarged section taken generally along line 62-62 in FIG. 61, showing the coins in full elevation;
FIG. 63 is an enlarged section taken generally along line 63-63 in FIG. 61, showing in full elevation a nickel registered with an ejection recess;
FIG. 64 is a diagrammatic cross-section of a coin and an improved coin discrimination sensor embodying the invention;
FIG. 65 is a schematic circuit diagram of the coin discrimination sensor of FIG. 64; FIG. 66 is a diagrammatic perspective view of the coils in the coin discrimination sensor of FIG. 64; FIG. 67a is a circuit diagram of a detector circuit for use with the discrirnination sensor of this invention;
FIG. 67b is a waveform diagram of the input signals supplied to the circuit of FIG. 67a; FIG. 68 is a perspective view of an outboard shunting device embodying the present invention;
FIG. 69 is a section taken generally along line 69-69 in FIG. 68; FIG. 70 is a section taken generally along line 70-70 in FIG. 68, showing a movable partition in a nondiverting position; and FIG. 71 is the same section illustrated in FIG. 70, showing the movable portion in a diverting position.
Detailed Description Of The Preferred Embodiments
While the invention is susceptible to various modifications and altemative forms, specific embodiments thereof have been shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that it is not intended to limit the invention to the particular forms disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as defined by the appended claims. Tuming now to the drawings and referring first to FIGs. la, lb and lc, there is shown an automatic teller machine ("ATM") having a bill deposit receptacle 1 as well as a bill withdrawal or return slot 2. The ATM has the conventional slot 3 for receiving the customer's identification card so that the data on the card can be automatically read by a card reader. A video display 4 provides the customer with a menu of options, and also prompts the customer to carry out the various actions required to execute a transaction, including the use of a keypad 5.
The illustrative ATM also has a coin deposit receptacle 6 and a coin return pocket 7. The deposit receptacles 1 and 6 are normally retracted within the machine but are advanced to their open positions (shown in FIG. la) when a customer initiates a transaction and selects a "cash deposit" mode of operation. Bills and coins can then be deposited by the customer into the deposit receptacles 1 and 6, respectively. After the customer has placed a stack of bills into the receptacle 1, the customer is prompted to push that receptacle into the machine, to its retracted position. This inward movement of the receptacle 1 positions the stack of bills at the feed station of a bill scanning and counting module 8 which automatically feeds, counts, scans and authenticates the bills one at a time at a high speed (e.g., at least 350 bills per minute). The bills that are recognized by the scanning module 8 are delivered to a conventional currency canister 9 (FIG. lc) which is periodically removed from the machine and replaced with an empty canister. When a bill cannot be recognized by the scanning module, a diverter 10 is actuated to divert the unidentified bill to the retum slot 2 so that it can be removed from the machine by the customer. Altematively, unrecognizable bills can be diverted to a separate cunency canister rather than being returned to the customer. Bills that are detected to be counterfeit are treated in the same manner as unrecognizable bills.
Though not shown in FIGs. la-lc, the bill transport system may also include an escrow holding area where the bills being processed in a pending deposit transaction are held until the transaction is complete. Then if the declared balance entered by the customer does not agree with the amount verified by the machine, the entire stack of bills can be returned to the customer. If desired, this decision can be controlled by the customer via the keypad. When coins are deposited by the customer in the receptacle 6, the customer again is prompted to push that receptacle into the machine. This causes the coins to be fed by gravity into the receiving hopper of a coin-sorting and counting module 11 which physically separates the coins by size (denomination) while separately counting the number of coins of each denomination in each separate transaction. The module 11 also includes a coin discriminator which detects coins that are counterfeit or otherwise non-genuine. These unacceptable coins are discharged from the sorter at a common exit, and the coins from that exit are guided by a tube 12 to the coin return slot 7.
The ATM also preferably includes a conventional loose currency dispensing module 13 for dispensing loose bills, and/or a strapped cunency dispensing module 14 for dispensing strapped cunency, into a receptacle 15 at the front of the machine, in response to a withdrawal transaction. If desired, a loose coin dispensing module 16 and/or a rolled coin dispensing module 17, may also be included for dispensing coins via the coin return pocket 7. Additional modules that may be included in the ATM or a redemption machine using the same system are modules for verifying and accepting checks, food stamps, tokens and/or tickets containing bar codes. As will be described in more detail below, each of the modules 8 and 11 accumulates data representing both the number and the value of each separate cunency item processed by these modules in each separate transaction. At the end of each transaction, this data and the account number for the transaction are downloaded to an associated cash accounting system by a modem link, so that the customer's account can be immediately adjusted to reflect both the deposits and the withdrawals effected by the cunent transaction. Altematively, the data from the cuπency- processing modules and the card reader can be temporarily stored within a temporary memory within the ATM, so that the data can be downloaded at intervals controlled by the computing system on which the cash accounting system is n. FIG. ld is a flow chart of a subroutine for transferring data from the ATM to the cash accounting system. This subroutine is entered at step 10a each time a customer inserts an identification card into the ATM. The customer's account number is stored at step 10b, and step 10c then initiates a transaction by prompting the customer to select from a menu of available deposit or withdrawal transactions, and step lOd then monitors the ATM system to determine when the transaction is complete. When the answer is affirmative, the bill deposit amount B^, the bill withdrawal amount Bw, the coin deposit amount C<j, and the coin withdrawal amount Cw are stored at steps lOe, lOf, lOg and lOh, and then downloaded to the cash accounting system at step IOi. If desired, these amounts may be loaded into a buffer memory for later retrieval by the computer that controls the cash accounting system. The cash accounting system then enters these amounts in the customer's account, and immediately adjusts the balance in that account accordingly.
A subroutine for executing a cash deposit of bills is shown in FIG. le. When this type of transaction is selected by the customer, the video display prompts the customer to place the stack of bills being deposited into the receptacle 1 and to push that receptacle into the machine. The bill counting and scanning module then automatically withdraws one bill at a time from the bottom of the stack, and scans each bill for denomination and authentication.
Each successive bill that is withdrawn from the deposit stack is scanned at step 1 la to determine the denomination of the bill, and checked for authentication at step lib. The results of the authentication are checked at step l ie. If the bill cannot be authenticated, it is a counterfeit suspect and thus step 1 lc produces an affirmative answer. This advances the system to step lid, which determines whether the owner of the ATM or redemption machine has opted to retum counterfeit-suspect bills to the customer. If this option has been selected, the suspect bill is returned to the customer at step lie. If the return option has not been selected at step lid, the resulting negative response advances the system to step 1 lf which transports the bill to a suspect bill canister.
If the bill is not a counterfeit suspect, the resulting negative answer at step l ie advances the system to step 1 lg to check the results of the scanning step. This step determines whether the bill is a "no call," i.e. , whether it was impossible for the scanning operation to determine the denomination of the bill. If the bill is a "no call," step l lg produces an affirmative answer, and step l lh determines whether the option to retum "no calls" to the customer has been selected. If the answer is affirmative, the "no call" bill is returned to the customer at step lie. If the answer is negative, the "no call" bill is transported to a "no call" canister at step Hi. If the denomination of the bill has been determined by the scanner, the resulting negative response at step 1 lg causes the counter for that particular denomination to be incremented at step llj. The dollar value of that denomination is then added to the verified deposit amount at step I lk to maintain a cunent cumulative total of the cunency deposit that is being processed. The bill is then transported to an escrow holding area for the cunent deposit, at step 111.
To determine when the processing of a deposit has been completed, step 11m determines when the last bill in a deposited stack of bills has been counted. When this step produces an affirmative answer, step 11m then determines whether the final verified deposit amount agrees with the declared balance that was entered by the customer through the key pad. If the answer is affirmative, the deposited bills are transported from the escrow holding area to a verified deposit canister at step l lo. A negative answer at step lln advances the system to step l ip where again the system determines whether a "retum" option has been selected. This option may be preselected by the owner of the ATM or redemption machine, or it may be an option that is available to the customer. In any event, if the option has been selected, the bills are returned to the customer at step llq to enable the customer to determine why the verified deposit amount does not agree with the customer's declared balance. At this time, the verified deposit amount is displayed to the customer along with an appropriate message. A negative response at step lip causes the bills to be transported from the escrow holding area to a disputed balance canister at step llr. FIG. lf illustrates a modification of the routine of FIG. le which permits the use of a single storage canister for all of the bills, regardless of whether they are verified bills, no calls, or counterfeit suspects. In this system the various bills are identified within the single canister by placing different colored markers on top of different bills. These markers are inserted into the bill transport path so that they follow the respective bills to be marked into the canister. Specifically, a first marker, e.g., a marker of a first color, is inserted at step 11s following an affirmative response at step l ie and a negative response at step lid to indicate that the bill is a counterfeit suspect that is not to be returned to the customer. A second type of marker, e.g., a marker of a second color, is inserted at step lit in response to an affirmative response at step 1 lg and a negative answer at step 1 lh, to indicate that the marked bill is a counterfeit suspect. A third type of marker, e.g., of a third color, is inserted at step llu in response to negative answers at steps lln and lip, to indicate that the marked batch of bills represents a deposit whose verified amount did not agree with the customer's declared balance. Because this third type of marker identifies a batch of bills instead of a single bill, it is necessary to insert a marker at both the beginning and end of the marked batch.
In the event that the customer wishes to deposit "no call" bills that are returned to the customer, the customer may key in the value and number of such bills and deposit them in an envelope for later verification by the bank. A message on the display screen may advise the customer of this option. For example, if four $10 bills are returned, and then re-deposited by the customer in an envelope, fhe customer may press a "$10" key four times. The customer then receives immediate credit for all the bills denominated and authenticated by the scanner. Credit for the re-deposited "no call" bills is given only after the bank picks up the deposit envelope and manually verifies the amount. Altematively, at least prefened customers can be given full credit immediately, subject to later verification, or immediate credit can be given up to a certain dollar limit. In the case of counterfeit bills that are not retumed to the customer, the customer can be notified of the detection of a counterfeit suspect at the ATM or later by a written notice or personal call, depending upon the preferences of the financial institution.
The ATM or redemption machine may also have a "verify mode" in which it simply denominates and totals all the cunency (bills and/or coins) deposited by the customer and retums it all to the customer. If the customer agrees with the amount and wishes to proceed with an actual deposit, the customer selects the "deposit mode" and re-deposits the same batch of cunency in the machine. Altematively, the "verify mode" may hold the initially deposited cunency in an escrow area until the customer decides whether to proceed with an actual deposit.
In the event that the machine jams or otherwise malfunctions while cunency is being processed, the message display screen advises the customer of the number and value of the cunency items processed prior to the jam. The customer is instructed to retrieve the cunency not yet processed and to manually deposit it in a sealed envelope which is then deposited into the machine for subsequent verification. The machine malfunction is automatically reported via modem to the home office. Referring now to FIG. 2a, there is shown a prefened embodiment of a cunency scanning and counting module 8. The module 8 includes a bill accepting station 12 for receiving stacks of cunency bills from the deposit receptacle 1. A feed mechanism functions to pick out or separate one bill at a time for transfer to a bill transport mechanism 16 (FIG. 2a) which transports each bill along a precisely predetermined transport path, between a pair of scanheads 18a, 18b where the denomination of the bill is identified. In the prefened embodiment, bills are scanned and identified at a rate in excess of 350 bills per minute. In the prefened embodiment depicted, each scanhead 18a, 18b is an optical scanhead that scans for characteristic information from a scanned bill 17 which is used to identify the denomination of the bill. The scanned bill 17 is then transported to a cassette or bill stacking station 20 where bills so processed are stacked for subsequent removal.
Each optical scanhead 18a, 18b preferably comprises a pair of light sources 22 directing light onto the bill transport path so as to illuminate a substantially rectangular light strip 24 upon a cunency bill 17 positioned on the transport path adjacent the scanhead 18. Light reflected off the illuminated strip 24 is sensed by a photodetector 26 positioned between the two light sources. The analog output of the photodetector 26 is converted into a digital signal by means of an analog-to-digital (ADC) convertor unit 28 whose output is fed as a digital input to a central processing unit (CPU) 30.
While the scanheads 18a, 18b of FIG. 2a are optical scanheads, it should be understood that the scanheads and the signal processing system may be designed to detect a variety of characteristic information from cunency bills. Additionally, the scanheads may employ a variety of detection means such as magnetic, optical, electrical conductivity, and capacitive sensors. Use of such sensors is discussed in more detail below (see, e.g., FIG. 2d).
Referring again to FIG. 2a, the bill transport path is defined in such a way that the transport mechanism 16 moves cunency bills with the naπow dimension of the bills being parallel to the transport path and the scan direction. Altematively, the system may be designed to scan bills along their long dimension or along a skewed dimension. As a bill 17 traverses the scanheads 18a, 18b, the coherent light strip 24 effectively scans the bill across the naπow dimension of the bill. In the prefeπed embodiment depicted, the transport path is so arranged that a cunency bill 17 is scanned across a central section of the bill along its naπow dimension, as shown in FIG. 2a. Each scanhead functions to detect light reflected from the bill as it moves across the illuminated light strip 24 and to provide an analog representation of the variation in reflected light, which, in turn, represents the variation in the dark and light content of the printed pattem or indicia on the surface of the bill. This variation in light reflected from the naπow-dimension scanning of the bills serves as a measure for distinguishing, with a high degree of confidence, among a plurality of cunency denominations which the system is programmed to handle. A series of such detected reflectance signals are obtained across the naπow dimension of the bill, or across a selected segment thereof, and the resulting analog signals are digitized under control of the CPU 30 to yield a fixed number of digital reflectance data samples. The data samples are then subjected to a normalizing routine for processing the sampled data for improved conelation and for smithing out variations due to "contrast" fluctuations in the printed pattem existing on the bill surface. The normalized reflectance data represents a characteristic pattem that is unique for a given bill denomination and provides sufficient distinguishing features among characteristic pattems for different cunency denominations. In order to ensure strict coπespondence between reflectance samples obtained by naπow dimension scanning of successive bills, the reflectance sampling process is preferably controlled through the CPU 30 by means of an optical encoder 32 which is linked to the bill transport mechanism 16 and precisely tracks the physical movement of the bill 17 between the scanheads 18a, 18b. More specifically, the optical encoder 32 is linked to the rotary motion of the drive motor which generates the movement imparted to the bill along the transport path. In addition, the mechanics of the feed mechanism ensure that positive contact is maintained between the bill and the transport path, particularly when the bill is being scanned by the scanheads. Under these conditions, the optical encoder 32 is capable of precisely tracking the movement of the bill 17 relative to the light strips 24 generated by the scanheads 18a, 18b by monitoring the rotary motion of the drive motor.
The outputs of the photodetectors 26 are monitored by the CPU 30 to initially detect the presence of the bill adjacent the scanheads and, subsequently, to detect the starting point of the printed patte on the bill, as represented by the thin borderline 17a which typically encloses the printed indicia on U.S. cunency bills. Once the borderline 17a has been detected, the optical encoder 32 is used to control the timing and number of reflectance samples that are obtained from the outputs of the photodetectors 26 as the bill 17 moves across the scanheads.
FIG. 2b illustrates a modified cunency scanning and counting device similar to that of FIG. 2a but having a scanhead on only a single side of the transport path. 20 FIG. 2c illustrates anodier modified cunency scanning and counting device similar to that of FIG. 2b but illustrating feeding and scanning of bills along their wide direction.
As illustrated in FIGs. 2b-2c, the transport mechanism 16 moves cunency bills with a preselected one of their two dimensions (naπow or wide) being parallel to the transport path and the scan direction. FIGs. 2b and 4a illustrate bills oriented with their naπow dimension "W" parallel to the direction of movement and scanning, while FIGs. 2c and 4b illustrate bills oriented with their wide dimension "L" parallel to the direction of movement and scanning. Referring now to FIG. 2d, there is shown a functional block diagram illustrating a prefeπed embodiment of a cunency discriminating and authenticating system. The operation of the system of FIG. 2d is the same as that of FIG. 2a except as modified below. The system includes a bill accepting station 12 where stacks of cunency bills that need to be identified, authenticated, and counted are positioned. Accepted bills are acted upon by a bill separating station 14 which functions to pick out or separate one bill at a time for transfer to a bill transport mechanism 16 which transports each bill along a precisely predetermined transport path, across two scanheads 18 and 39 where the cunency denomination of the bill is identified and the genuineness of the bill is authenticated. In the prefeπed embodiment depicted, scanhead 18 is an optical scanhead that scans for a first type of characteristic information from a scanned bill 17 which is used to identify the bill's denomination. A second scanhead 39 scans for a second type of characteristic information from the scanned bill 17. While the illustrated scanheads 18 and 39 are separate and distinct, they may be incorporated into a single scanhead. For example, where the first characteristic sensed is intensity of reflected light and the second characteristic sensed is color, a single optical scanhead having a plurality of detectors, one or more without filters and one or more with colored filters, may be employed (U.S. Pat. No. 4,992, 860 incorporated herein by reference). The scanned bill is then transported to a bill stacking station 20 where bills so processed are stacked for subsequent removal.
The optical scanhead 18 of the embodiment depicted in FIG. 2d comprises at least one light source 22 directing a beam of coherent light downwardly onto the bill transport path so as to illuminate a substantially rectangular light strip 24 upon a cunency bill 17 positioned on the transport path below the scanhead 18. Light reflected off die iUurninated strip 24 is sensed by a photodetector 26 positioned directly above the strip. The analog output of photodetector 26 is converted into a digital signal by means of an analog-to-digital (ADC) convertor unit 28 whose output is fed as a digital input to a central processing unit (CPU) 30.
The second scanhead 39 comprises at least one detector 41 for sensing a second type of characteristic infoπnation from a bill. The analog output of the detector 41 is converted into a digital signal by means of a second analog-to-digital converter 43 whose output is also fed as a digital input to the central processing unit (CPU) 30.
While the scanhead 18 in the embodiment of FIG. 2d is an optical scanhead, it should be understood that the first and second scanheads 18 and 39 may be designed to detect a variety of characteristic information from cunency bills. Additionally these scanheads may employ a variety of detection means such as magnetic or optical sensors. For example, a variety of cunency characteristics can be measured using magnetic sensing. These include detection of pattems of changes in magnetic flux (U.S. Pat. No. 3,280,974), pattems of vertical grid lines in the portrait area of bills (U.S. Pat. No. 3,870,629), the presence of a security thread (U.S. Pat. No. 5,151,607), total amount of magnetizable material of a bill (U.S. Pat. No.
4,617,458), pattems from sensing the strength of magnetic fields along a bill (U.S. Pat. No. 4,593,184), and other pattems and counts from scanning different portions of the bill such as the area in which the denomination is written out (U.S. Pat. No. 4,356,473). With regard to optical sensing, a variety of cunency characteristics can be measured such as density (U.S. Pat. No. 4,381,447), color (U.S. Pat. Nos. 4,490,846; 3,496,370; 3-,480,785), length and thickness (U.S. Pat. No. 4,255,651), the presence of a security thread (U.S. Pat. No. 5,151,607) and holes (U.S. Pat. No. 4,381,447), and other pattems of reflectance and transmission (U.S. Pat. No. 3,496,370; 3,679,314; 3,870,629; 4,179,685). Color detection techniques may employ color filters, colored lamps, and/or dichroic beamsplitters (U.S. Pat. Nos. 4,841,358; 4,658,289; 4,716,456; 4,825,246, 4,992,860 and EP 325,364). Prescribed hues or intensities of a given color may be detected. Reflection and/or fluorescence of ultraviolet light may also be used, as described in detail below. Absorption of infrared light may also be used as an authenticating technique.
In addition to magnetic and optical sensing, other techniques of detectmg characteristic information of cunency include electrical conductivity sensing, capacitive sensing (U.S. Pat. No. 5,122,754 [watermark, security thread]; 3,764,899 [thickness]; 3,815,021 [dielectric properties]; 5,151,607 [security thread]), and mechanical sensing (U.S. Pat. Nos. 4,381,447 [limpness]; 4,255,651 [thickness]), and hologram, kinegram and moviegram sensing. The detection of the borderline 17a realizes improved discrimination efficiency in systems designed to accommodate U.S. cunency since the borderline 17a serves as an absolute reference point for initiation of sampling. When the edge of a bill is used as a reference point, relative displacement of sampling points can occur because of the random manner in which the distance from the edge to the borderline 17a varies from bill to bill due to the relatively large range of tolerances permitted during printing and cutting of cunency bills. As a result, it becomes difficult to establish direct coπespondence between sample points in successive bill scans and the discrimination efficiency is adversely affected. Accordingly, the modified pattem generation method discussed below is useful in discrimination systems designed to accommodate bills other than U.S. cunency because many non-U.S. bills lack a borderline around the printed indicia on their bills. Likewise, the modified patte generation method may be important in discrimination systems designed to accommodate bills other than U.S. cunency because the printed indicia of many non- U.S. bills lack sharply defined edges which in turns inhibits using the edge of the printed indicia of a bill as a trigger for the initiation of the scanning process and instead promotes reliance on using the edge of the bill itself as the trigger for the initiation of the scanning- process.
The use of the optical encoder 32 for controlling the sampling process relative to the physical movement of a bill 17 across the scanheads 18a, 18b is also advantageous in that the encoder 32 can be used to provide a predetermined delay following detection of the borderline 17a prior to initiation of samples. The encoder delay can be adjusted in such a way that the bill 17 is scanned only across those segments which contain the most distinguishable printed indicia relative to the different cunency denominations.
In the case of U.S. cunency, for instance, it has been deteπnined that the central, approximately two-inch (approximately 5 cm) portion of cunency bills, as scanned across the central section of the naπow dimension of the bill, provides sufficient data for distinguishing among the various U.S. cunency denominations. Accordingly, the optical encoder can be used to control the scanning process so that reflectance samples are taken for a set period of time and only after a certain period of time has elapsed after the borderline 17a is detected, thereby restricting the scanning to the desired central portion of the naπow dimension of the bill.
FIGs. 3-5b illustrate the scanning process in more detail. Referring to FIG. 4a, as a bill 17 is advanced in a direction parallel to the naπow edges of the bill, scanning via a slit in the scanhead 18a or 18b is effected along a segment S of the central portion of the bill 17. This segment S begins a fixed distance D inboard of the borderline 17a. As the bill 17 traverses the scanhead, a strip s of the segment S is always illuminated, and the photodetector 26 produces a continuous output signal which is proportional to the intensity of the light reflected from the illuminated strip s at any given instant. This output is sampled at intervals controlled by the encoder, so that the sampling intervals are precisely synchronized with the movement of the bill across the scanhead. FIG. 4b is similar to FIG. 4a but illustrates scanning along the wide dimension of the bill 17.
As illustrated in FIGs. 3, 5a, and 5b, it is prefeπed that the sampling intervals be selected so that the strips s that are illuminated for successive samples overlap one another. The odd-numbered and even-numbered sample strips have been separated in FIGs. 3, 5a, and 5b to more clearly illustrate this overlap. For example, the first and second strips si and s2 overlap each other, the second and third strips s2 and s3 overlap each other, and so on. Each adjacent pair of strips overlap each other. In the illustrative example, this is accomplished by sampling strips that are 0.050 inch (0J27 cm) wide at 0.029 inch (0.074 cm) intervals, along a segment S that is 1.83 inch (4.65 cm) long (64 samples).
FIGs. 6a and 6b illustrate two opposing surfaces of U.S. bills. The printed pattems on the black and green surfaces of the bill are each enclosed by respective thin borderlines Bj and B9. As a bill is advanced in a direction parallel to the naπow edges of the bill, scanning via the wide slit of one of the scanheads is effected along a segment SA of the central portion of die black surface of the bill (FIG. 6a). As previously stated, the orientation of the bill along the transport path determines whether the upper or lower scanhead scans the black surface of the bill. This segment SA begins a fixed distance D^ inboard of die borderline B^, which is located a distance Wj from the edge of the bill. The scanning along segment SA is as described in connection with FIGs. 3, 4a, and 5a.
Similarly, the other of the two scanheads scans a segment Sg of die central portion of the green surface of the bill (FIG. 6b). The orientation of the bill along the transport path determines whether the upper or lower scanhead scans the green surface of the bill. This segment Sg begins a fixed distance D9 inboard of the border line B2, which is located a distance W2 from the edge of the bill. For U.S. cunency, the distance W2 on die green surface is greater than the distance Wj on the black surface. It is this feature of U.S. cunency which permits one to determine die orientation of the bill relative to the upper and lower scanheads 18, thereby permitting one to select only the data samples coπesponding to me green surface for conelation to the master characteristic pattems in the EPROM 34. The scanning along segment Sg is as described in connection with FIGs. 3, 4a, and 5a. FIGs. 6c and 6d are side elevations of FIG. 2a. FIG. 6c shows the first surface of a bill scanned by an upper scanhead and the second surface of the bill scanned by a lower scanhead, while FIG. 6d shows the first surface of a bill scanned by a lower scanhead and the second surface of the bill scanned by an upper scanhead. FIGs. 6c and 6d illustrate the pair of optical scanheads 18a, 18b disposed on opposite sides of the transport path to permit optical scanning of both surfaces of a bill. With respect to United States cunency, these opposing surfaces coπespond to the black and green surfaces of a bill. One of the optical scanheads 18 (the "upper" scanhead 18a in FIGs. 6c-6d) is positioned above die transport path and illuminates a light strip upon a first surface of the bill, while the other of the optical scanheads 18 (the "lower" scanhead 18b in FIGs. 6c-6d) is positioned below the transport path and illuminates a light strip upon the second surface of the bill. The surface of die bill scanned by each scanhead 18 is determined by the orientation of the bill relative to the scanheads 18. The upper scanhead 18a is located slightiy upstream relative to the lower scanhead 18b.
The photodetector of the upper scanhead 18a produces a first analog output conesponding to the first surface of the bill, while die photodetector of the lower scanhead 18b produces a second analog output conesponding to the second surface of the bill. The first and second analog outputs are converted into respective first and second digital outputs by means of respective analog-to-digital (ADC) convertor units 28 whose outputs are fed as digital inputs to a central processing unit (CPU) 30. As described in detail below, the CPU 30 uses the sequence of operations illustrated in FIG. 12 to determine which of die first and second digital outputs coπesponds to die green surface of the bill, and then selects the "green" digital output for subsequent conelation to a series of master characteristic pattems stored in EPROM 34. As explained below, the master characteristic pattems are preferably generated by performing scans on the green surfaces, not black surfaces, of bills of different denominations. According to a prefeπed embodiment, fhe analog output coπesponding to die black surface of the bill is not used for subsequent conelation. The optical sensing and conelation technique is based upon using the above process to generate a series of stored intensity signal pattems using genuine bills for each denomination of cunency mat is to be detected. According to a prefened embodiment, two or four sets of master intensity signal samples are generated and stored within the system memory, preferably in the form of an EPROM 34 (see FIG. 2a), for each detectable cunency denomination. According to one prefened embodiment these are sets of master green-surface intensity signal samples. In the case of U.S. cunency, the sets of master intensity signal samples for each bill are generated from optical scans, performed on the green surface of the bill and taken along both the "forward" and "reverse" directions relative to the pattem printed on the bill. Altematively, the optical scanning may be performed on the black side of U.S. cunency bills or on either surface of foreign bills. Additionally, me optical scanning may be performed on both sides of a bill. In adapting this technique to U.S. cunency, for example, sets of stored intensity signal samples are generated and stored for seven different denominations of U.S. cunency, i.e., $1, $2, $5, $10, $20, $50 and $100. For bills which produce significant pattem changes when shifted slightiy to the left or right, such as the $2, the $10 and/or the $100 bills in U.S. cunency, it is prefened to store two green-side pattems for each of the "forward" and "reverse" directions, each pair of pattems for the same direction represent two scan areas that are slightly displaced from each other along the long dimension of the bill. Accordingly, a set of 16 [or 18] different green-side master characteristic pattems are stored within the EPROM for subsequent conelation purposes (four master pattems for the $10 bill [or four master pattems for the $10 bill and the $2 bill and/or the $100 bill] and two master pattems for each of the other denominations). The generation of the master pattems is discussed in more detail below. Once fhe master pattems have been stored, the pattem generated by scanning a bill under test is compared by fhe CPU 30 with each of the 16 [or 18] master pattems of stored intensity signal samples to generate, for each comparison, a conelation number representing the extent of conelation, i.e., similarity between conesponding ones of the plurality of data samples, for the sets of data being compared.
According to a prefened embodiment, in addition to the above set of 18 original green-side master pattems, five more sets of green-side master pattems are stored in memory. These sets are explained more fully in conjunction with FIGs. 18a and 18b below. The CPU 30 is programmed to identify the denomination of the scanned bill as conesponding to e set of stored intensity signal samples for which me conelation number resulting from pattem comparison is found to be the highest. In order to preclude the possibility of mischaracterizing the denomination of a scanned bill, as well as to reduce the possibility of spurious notes being identified as belonging to a valid denomination, a bi-level threshold of conelation is used as me basis for making a "positive" call. If a "positive" call can not be made for a scanned bill, an eπor signal is generated.
According to a prefeπed embodiment, master pattems are also stored for selected denominations coπesponding to scans along the black side of U.S. bills. More particularly, according to a prefened embodiment, multiple black-side master pattems are stored for $20, $50 and $100 bills. For each of these denominations, three master pattems are stored for scans in the forward and reverse directions for a total of six pattems for each denomination. For a given scan direction, black-side master pattems are generated by scanning a conesponding denominated bill along a segment located about the center of the naπow dimension of the bill, a segment slightly displaced (0.2 inches) to the left of center, and a segment slightiy displaced (0.2 inches) to the right of center. When die scanned pattem generated from me green side of a test bill fails to sufficientiy conelate with one of the green-side master pattems, me scanned pattem generated from the black side of a test bill is then compared to black-side master pattems in some situations as described in more detail below in conjunction with FIGs. 19a-19c. Using the above sensing and conelation approach, the CPU 30 is programmed to count the number of bills belonging to a particular cunency denomination as part of a given set of bills that have been scanned for a given scan batch, and to determine d e aggregate total of the cunency amount represented by the bills scanned during a scan batch. The CPU 30 is also linked to an output unit 36 (FIGs. 2a and FIG. 2b) which is adapted to provide a display of the number of bills counted, die breakdown of die bills in terms of cunency denomination, and die aggregate total of the cunency value represented by counted bills. The output unit 36 can also be adapted to provide a print-out of the displayed information in a desired format.
Referring again to the prefened embodiment depicted in FIG. 2d, as a result of the first comparison described above based on the reflected light intensity information retrieved by scanhead 18, the CPU 30 will have either determined die denomination of fhe scanned bill 17 or determined that die first scanned signal samples fail to sufficiently conelate with any of the sets of stored intensity signal samples in which case an enor is generated. Provided diat an enor has not been generated as a result of diis first comparison based on reflected light intensity characteristics, a second comparison is performed. This second comparison is performed based on a second type of characteristic information, such as alternate reflected light properties, similar reflected light properties at alternate locations of a bill, light transmissivity properties, various magnetic properties of a bill, the presence of a security thread embedded within a bill, the color of a bill, the thickness or other dimension of a bill, etc. The second type of characteristic infoπnation is retrieved from a scanned bill by the second scanhead 39. The scanning and processing by scanhead 39 may be controlled in a manner similar to that described above with regard to scanhead 18.
In addition to the sets of stored first characteristic information, in this example stored intensity signal samples, the EPROM 34 stores sets of stored second characteristic information for genuine bills of me different denominations which the system 10 is capable of handling. Based on the denomination indicated by die first comparison, the CPU 30 retrieves the set or sets of stored second characteristic data for a genuine bill of the denomination so indicated and compares the retrieved information with die scanned second characteristic information. If sufficient conelation exists between the retrieved information and the scanned information, the CPU 30 verifies the genuineness of the scanned bill 17. Otherwise, the CPU generates an eπor. While the prefeπed embodiment illustrated in FIG. 2d depicts a single CPU 30 for malting comparisons of first and second characteristic information and a single EPROM 34 for storing first and second characteristic information, it is understood that two or more CPUs and/or EPROMs could be used, including one CPU for making first characteristic information comparisons and a second CPU for making second characteristic information comparisons. Using the above sensing and conelation approach, the CPU 30 is programmed to count die number of bills belonging to a particular cunency denomination whose genuineness has been verified as part of a given set of bills that have been scanned for a given scan batch, and to determine the aggregate total of the cunency amount represented by the bills scanned during a scan batch.
Referring now to FIGs. 7a and 7b, there is shown a representation, in block diagram form, of a prefened circuit arrangement for processing and coπelating reflectance data according to the system of this invention. The CPU 30 accepts and processes a variety of input signals including those from the optical encoder 32, the sensor 26 and the erasable programmable read only memory (EPROM) 60. The EPROM 60 has stored within it die conelation program on the basis of which pattems are generated and test pattems compared with stored master programs in order to identify the denomination of test cunency. A crystal 40 serves as the time base for the CPU 30, which is also provided with an extemal reference voltage "VREF 42 on the basis of which peak detection of sensed reflectance data is performed.
According to one embodiment, the CPU 30 also accepts a timer reset signal from a reset unit 44 which, as shown in FIG. 7b, accepts the output voltage from the photodetector 26 and compares it, by means of a threshold detector 44a, relative to a pre-set voltage threshold, typically 5.0 volts, to provide a reset signal which goes "high" when a reflectance value conesponding to the presence of paper is sensed. More specifically, reflectance sampling is based on the premise that no portion of the illuminated light strip (24 in FIG. 2a) is reflected to the photodetector in the absence of a bill positioned below the scanhead. Under these conditions, fhe output of the photodetector represents a "dark" or "zero" level reading. The photodetector output changes to a "white" reading, typically set to have a value of about 5.0 volts, when the edge of a bill first becomes positioned below die scanhead and falls under the light strip 24. When this occurs, the reset unit 44 provides a "high" signal to die CPU 30 and marks the initiation of the scanning procedure.
The machine-direction dimension, that is, the dimension parallel to fhe direction of bill movement, of the illuminated strip of light produced by the light sources within me scanhead is set to be relatively small for the initial stage of the scan when the thin borderline is being detected, according to a prefened embodiment. The use of the nanow slit increases the sensitivity with which the reflected light is detected and allows minute variations in the "gray" level reflected off the bill surface to be sensed. This ensures mat the thin borderline of fhe pattem, i.e., fhe starting point of fhe printed pattem on the bill, is accurately detected. Once the borderline has been detected, subsequent reflectance sampling is performed on die basis of a relatively wider light strip in order to completely scan across the nanow dimension of the bill and obtain the desired number of samples, at a rapid rate. The use of a wider slit for the actual sampling also smoothes out the output characteristics of the photodetector and realizes the relatively large magnitude of analog voltage which is desirable for accurate representation and processing of fhe detected reflectance values.
The CPU 30 processes die output of fhe sensor 26 through a peak detector 50 which essentially functions to sample the sensor output voltage and hold the highest, i.e., peak, voltage value encountered after the detector has been enabled. For U.S. cunency, the peak detector is also adapted to define a scaled voltage on the basis of which the printed borderline on the cunency bills is detected. The output of the peak detector 50 is fed to a voltage divider 54 which lowers the peak voltage down to a scaled voltage Vg representing a predefined percentage of this peak value. The voltage Vg is based upon die percentage drop in output voltage of the peak detector as it reflects the transition from the "high" reflectance value resulting from fhe scanning of the unprinted edge portions of a cunency bill to the relatively lower "gray" reflectance value resulting when the thin borderline is encountered. Preferably, the scaled voltage Vg is set to be about 70 - 80 percent of the peak voltage.
The scaled voltage Vg is supplied to a line detector 56 which is also provided with the incoming instantaneous output of die sensor 26. The line detector 56 compares the two voltages at its input side and generates a signal LD£T which normally stays "low" and goes "high" when the edge of the bill is scanned. The signal LDET goes "low" when the incoming sensor output reaches the pre-defined percentage of the peak output up to that point, as represented by the voltage Vg. Thus, when die signal
Figure imgf000032_0001
goes "low", it is an indication that the borderline of fhe bill pattem has been detected. At this point, the CPU 30 initiates me actual reflectance sampling under control of the encoder 32, and the desired fixed number of reflectance samples are obtained as fhe cunency bill moves across the iUuminated light strip and is scanned along die central section of its naπow dimension.
When master characteristic pattems are being generated, the reflectance samples resulting from fhe scanning of one or more genuine bills for each denomination are loaded into conesponding designated sections within a system memory 60, which is preferably an EPROM. During cunency discrimination, the reflectance values resulting from fhe scanning of a test bill are sequentially compared, under control of die conelation program stored within the EPROM 60, with die conesponding master characteristic pattems stored within the EPROM 60. A pattem averaging procedure for scanning bills and generating characteristic pattems is described below in connection with FIGs. 15a-15e. The intenelation between d e use of the first and second type of characteristic information can be seen by considering FIGs. 8a and 8b which comprise a flowchart illustrating the sequence of operations involved in implementing a discrimination and authentication system according to a prefened embodiment of the present invention. Upon the initiation of the sequence of operations (step 1748), reflected light intensity information is retrieved from a bill being scanned (step 1750). Similarly, second characteristic information is also retrieved from the bill being scanned (step 1752). Denomination eπor and second characteristic enor flags are cleared (steps 1753 and 1754). Next the scanned intensity information is compared to each set of stored intensity information coπesponding to genuine bills of all denominations e system is programmed to accommodate (step 1758). For each denomination, a conelation number is calculated. The system then, based on the conelation numbers calculated, determines either the denomination of the scanned bill or generates a denomination enor by setting me denomination enor flag steps 1760 and 1762). In die case where the denomination enor flag is set (step 1762), the process is ended (step 1772). Altematively, if based on this first comparison, die system is able to determine die denomination of the scanned bill, the system proceeds to compare the scanned second characteristic information with the stored second characteristic information conesponding to die denomination determined by the first comparison (step 1764). For example, if as a result of the first comparison the scanned bill is deteπnined to be a $20 bill, fhe scanned second characteristic information is compared to the stored second characteristic information conesponding to a genuine $20 bill. In this manner, the system need not make comparisons with stored second characteristic information for die other denominations the system is programmed to accommodate. If based on this second comparison (step 1764) it is determined that the scanned second characteristic information does not sufficientiy match diat of die stored second characteristic information (step 1766), then a second characteristic eπor is generated by setting the second characteristic eπor flag (step 1768) and the process is ended (step 1772). If the second comparison results in a sufficient match between the scanned and stored second characteristic information (step 1766), then the denomination of the scanned bill is indicated (step 1770) and the process is ended (step 1772).
An example of an inteπelationship between authentication based on first and second characteristics can be seen by considering Table 1. The denomination determined by optical scanning of a bill is preferably used to facilitate authentication of the bill by magnetic scanning, using the relationship set forth in Table 1.
Table 1
Figure imgf000034_0001
Table 1 depicts relative total magnetic content thresholds for various denominations of genuine bills. Columns 1-5 represent varying degrees of sensitivity. The values in Table 1 are set based on fhe scanning of genuine bills of varying denominations for total magnetic content and setting required thresholds based on die degree of sensitivity selected. The information in Table 1 is based on the total magnetic content of a genuine $1 being 1000. The following discussion is based on a sensitivity setting of 4. In this example it is assumed that magnetic content represents the second characteristic tested. If the comparison of first characteristic information, such as reflected light intensity, from a scanned billed and stored information coπesponding to genuine bills results in an indication that the scanned bill is a $10 denomination, then the total magnetic content of die scanned bill is compared to die total magnetic content threshold of a genuine $10 bill, i.e., 200. If the magnetic content of the scanned bill is less than 200, the bill is rejected. Otherwise it is accepted as a $10 bill.
Referring now to FIGs. 9-1 lb, there are shown flow charts illustrating the sequence of operations involved in implementing the above-described optical sensing and conelation technique. FIGs. 9 and 10, in particular, illustrate fhe sequences involved in detecting the presence of a bill adjacent die scanheads and die borderlines on each side of fhe bill. Turning to FIG. 9, at step 70, the lower scanhead fine line interrupt is initiated upon die detection of die fine line by the lower scanhead. An encoder counter is maintained that is incremented for each encoder pulse. The encoder counter scrolls from 0 - 65,535 and then starts at 0 again. At step 71 die value of the encoder counter is stored in memory upon die detection of the fine line by the lower scanhead. At step 72 the lower scanhead fine line interrupt is disabled so that it will not be triggered again during the interrupt period. At step 73, it is determined whether fhe magnetic sampling has been completed for die previous bill. If it has not, the magnetic total for die previous bill is stored in memory at step 74, and the magnetic sampling done flag is set at step 75 so that magnetic sampling of the present bill may thereafter be performed. Steps 74 and 75 are skipped if it is determined at step 73 that die magnetic sampling has been completed for the previous bill. At step 76, a lower scanhead bit in die trigger flag is set. This bit is used to indicate that die lower scanhead has detected the fine line. The magnetic sampler is initialized at step 77, and the magnetic sampling interrupt is enabled at step 78. A density sampler is initialized at step 79, and a density sampling interrupt is enabled at step 80. The lower read data sampler is initialized at step 81, and a lower scanhead data sampling interrupt is enabled at step 82. At step 83, the lower scanhead fine line interrupt flag is reset, and at step 84 the program returns from the interrupt. Turning to FIG. 10, at step 85, the upper scanhead fine line interrupt is initiated upon the detection of the fine line by the upper scanhead. At step 86 die value of the encoder counter is stored in memory upon die detection of the fine line by the upper scanhead. This information in connection with the encoder counter value associated with die detection of the fine line by the lower scanhead may then be used to determine the face orientation of a bill, mat is whether a bill is fed green side up or green side down in the case of U.S. bills, as is described in more detail below in connection with FIG. 12. At step 87 die upper scanhead fine line interrupt is disabled so diat it will not be triggered again during the interrupt period. At step 88, die upper scanhead bit in the trigger flag is set. This bit is used to indicate that the upper scanhead has detected the fine line. By checking the lower and upper scanhead bits in the trigger flag, it can be determined whether each side has detected a respective fine line. Next, the upper scanhead data sampler is initialized at step 89, and me upper scanhead data sampling interrupt is enabled at step 90. At step 91, the upper scanhead fine line interrupt flag is reset, and at step 92 the program retums from the interrupt. Referring now to FIGs. Ila and lib, there are shown, respectively, the digitizing routines associated with the lower and upper scanheads. FIG. Ila is a flow chart illustrating the sequential procedure involved in the analog-to-digital conversion routine associated with the lower scanhead. The routine is started at step 93a. Next, the sample pointer is decremented at step 94a so as to maintain an indication of the number of samples remaining to be obtained. The sample pointer provides an indication of the sample being obtained and digitized at a given time. At step 95a, the digital data conesponding to die output of the photodetector associated with the lower scanhead for the cunent sample is read. The data is converted to its final form at step 96a and stored widiin a pre-defined memory segment as Xflvf-L at steP 97a. Next, at step 98a, a check is made to see if the desired fixed number of samples "N" has been taken. If the answer is found to be negative, step 99a is accessed where the interrupt authorizing the digitization of die succeeding sample is enabled, and die program retums from interrupt at step 100a for completing fhe rest of the digitizing process. However, if the answer at step 98a is found to be positive, i.e., the desired number of samples have already been obtained, a flag, namely the lower scanhead done flag bit, indicating die same is set at step 101a, and the program returns from interrupt at step 102a.
FIG. lib is a flow chart illustrating the sequential procedure involved in the analog-to-digital conversion routine associated with the upper scanhead. The routine is started at step 93b. Next, the sample pointer is decremented at step 94b so as to maintain an indication of the number of samples remaining to be obtained. The sample pointer provides an indication of fhe sample being obtained and digitized at a given time. At step 95b, the digital data conesponding to die output of the photodetector associated with the upper scanhead for the cunent sample is read. The data is converted to its final form at step 96b and stored within a pre-defined memory segment as jjsf.Tj at step 97b. Next, at step 98b, a check is made to see if die desired fixed number of samples "N" has been taken. If the answer is found to be negative, step 99b is accessed where the interrupt authorizing the digitization of die succeeding sample is enabled and the program retums from interrupt at step 100b for completing the rest of the digitizing process. However, if the answer at step 98b is found to be positive, i.e., die desired number of samples have already been obtained, a flag, namely the upper scanhead done flag bit, indicating the same is set at step 101b, and the program retums from interrupt at step 102b.
The CPU 30 is programmed with fhe sequence of operations in FIG. 12 to conelate at least initially only the test pattem conesponding to the green surface of a scanned bill. As shown in FIGs. 6c-6d, the upper scanhead 18a is located slighdy upstream adjacent the bill transport path relative to the lower scanhead 18b. The distance between the scanheads 18a, 18b in a direction parallel to the transport path coπesponds to a predetermined number of encoder counts. It should be understood diat die encoder 32 produces a repetitive tracking signal synchronized with incremental movements of the bill transport mechanism, and diis repetitive tracking signal has a repetitive sequence of counts (e.g., 65,535 counts) associated therewith. As a bill is scanned by the upper and lower scanheads 18a, 18b, the CPU 30 monitors the output of the upper scanhead 18a to detect die borderline of a first bill surface facing the upper scanhead 18a. Once this borderline of the first surface is detected, the CPU 30 retrieves and stores a first encoder count in memory.
Similarly, the CPU 30 monitors the output of the lower scanhead 18b to detect the borderline of a second bill surface facing the lower scanhead 18b. Once die borderline of the second surface is detected, the CPU 30 retrieves and stores a second encoder count in memory. Referring to FIG. 12, the CPU 30 is programmed to calculate the difference between the first and second encoder counts (step 105a). If this difference is greater than the predetermined number of encoder counts coπesponding to the distance between the scanheads 18a, 18b plus some safety factor number "X", e.g., 20 (step 106), the bill is oriented with its black surface facing the upper scanhead 18a and its green surface facing the lower scanhead 18b. This can best be understood by reference to FIG. 6c which shows a bill with the foregoing orientation. In diis situation, once die borderline B^ of the black surface passes beneath die upper scanhead 18a and die first encoder count is stored, the borderline B2 still must travel for a distance greater than die distance between die upper and lower scanheads 18a, 18b in order to pass over die lower scanhead 18b. As a result, the difference between the second encoder count associated with the borderline B and the first encoder count associated with the borderline B^ will be greater dian the predetermined number of encoder counts coπesponding to the distance between die scanheads 18a, 18b. With die bill oriented with its green surface facing the lower scanhead, the CPU 30 sets a flag to indicate that the test pattem produced by fhe lower scanhead 18b should be conelated (step 107). Next, this test pattem is conelated with the green-side master characteristic pattems stored in memory (step 109).
If at step 106 the difference between the first and second encoder counts is less than the predetermined number of encoder counts conesponding to me distance between the scanheads 18a, 18b, die CPU 30 is programmed to determine whedier the difference between the first and second encoder counts is less than the predetermined number minus some safety number "X", e.g., 20 (step 108). If the answer is negative, the orientation of the bill relative to fhe scanheads 18a, 18b is uncertain, so the CPU 30 is programmed to conelate die test pattems produced by both die upper and lower scanheads 18a, 18b with die green-side master characteristic pattems stored in memory (steps 109, 110, and 111).
If fhe answer is affirmative, the bill is oriented with its green surface facing the upper scanhead 18a and its black surface facing the lower scanhead 18b. This can best be understood by reference to FIG. 6d, which shows a bill with the foregoing orientation. In this situation, once the borderline B2 of fhe green surface passes beneath die upper scanhead 18a and the first encoder count is stored, die borderline Bj must travel for a distance less than die distance between the upper and lower scanheads 18a, 18b in order to pass over the lower scanhead 18b. As a result, the difference between fhe second encoder count associated with die borderline B^ and the first encoder count associated with die borderline Bo should be less dian the predetermined number of encoder counts coπesponding to die distance between the scanheads 18a, 18b. To be on die safe side, it is required that fhe difference between first and second encoder counts be less d an the predetermined number minus the safety number "X". Therefore, the CPU 30 is programmed to conelate die test pattem produced by die upper scanhead 18a widi die green-side master characteristic pattems stored in memory (step 111).
After coπelating fhe test pattem associated widi eidier the upper scanhead 18a, the lower scanhead 18b, or both scanheads 18a, 18b, the CPU 30 is programmed to perform the bi-level threshold check (step 112).
A simple conelation procedure is utilized for processing digitized reflectance values into a form which is conveniently and accurately compared to conesponding values pre-stored in an identical format. More specifically, as a first step, the mean value X for the set of digitized reflectance samples (comparing "n" samples) obtained for a bill scan run is first obtained as below:
n X x ~ ∑ ^ (1)
1=0 "
Subsequently, a normalizing factor Sigma ("σ") is determined as being equivalent to die sum of the square of the difference between each sample and the mean, as normalized by die total number n of samples. More specifically, the normalizing factor is calculated as below:
σ = ∑ '*'--X|2 (2) i=0 n In the final step, each reflectance sample is normalized by obtaining the difference between the sample and the above-calculated mean value and dividing it by the square root of die normalizing factor σ as defined by die following equation:
Figure imgf000040_0001
The result of using the above conelation equations is that, subsequent to the normalizing process, a relationship of conelation exists between a test pattem and a master pattem such mat die aggregate sum of the products of coπesponding samples in a test pattem and any master pattem, when divided by the total number of samples, equals unity if the pattems are identical. Odierwise, a value less than unity is obtained. Accordingly, die conelation number or factor resulting from the comparison of normalized samples within a test pattem to mose of a stored master pattem provides a clear indication of fhe degree of similarity or conelation between the two pattems.
According to a prefened embodiment of diis invention, die fixed number of reflectance samples which are digitized and normalized for a bill scan is selected to be 64. It has experimentally been found diat the use of higher binary orders of samples (such as 128, 256, etc.) does not provide a coπespondingly increased discrimination efficiency relative to the increased processing time involved in implementing die above-described conelation procedure. It has also been found diat fhe use of a binary order of samples lower than 64, such as 32, produces a substantial drop in discrimination efficiency.
The conelation factor can be represented conveniently in binary terms for ease of conelation. In a prefened embodiment, for instance, me factor of unity which results when a hundred percent conelation exists is represented in terms of fhe binary number 2 , which is equal to a decimal value of 1024. Using die above procedure, die normalized samples within a test pattem are compared to tiie master characteristic pattems stored within the system memory in order to determine me particular stored pattem to which the test pattem conesponds most closely by identifying the comparison which yields a conelation number closest to 1024. A bi-level threshold of conelation is required to be satisfied before a particular call is made, for at least certain denominations of bills. More specifically, the conelation procedure is adapted to identify fhe two highest conelation numbers resulting from the comparison of the test pattem to one of the stored pattems. At that point, a minimum threshold of conelation is required to be satisfied by these two conelation numbers. It has experimentally been found diat a conelation number of about 850 serves as a good cut-off threshold above which positive calls may be made with a high degree of confidence and below which die designation of a test pattem as coπesponding to any of the stored pattems is uncertain. As a second threshold level, a minimum separation is prescribed between the two highest conelation numbers before making a call. This ensures that a positive call is made only when a test pattem does not conespond, within a given range of conelation, to more than one stored master pattem. Preferably, me minimum separation between conelation numbers is set to be 150 when the highest conelation number is between 800 and 850. When the highest conelation number is below 800, no call is made.
The procedure involved in comparing test pattems to master pattems is discussed below in connection with FIG. 18a.
Next a routine designated as "CORRES" is initiated. The procedure involved in executing the routine CORRES is illustrated at FIG. 13 which shows the routine as starting at step 114. Step 115 determines whether the bill has been identified as a $2 bill, and, if the answer is negative, step 116 determines whether the best conelation number ("call #1") is greater than 799. If the answer is negative, the conelation number is too low to identify fhe denomination of the bill with certainty, and thus step 117 generates a "no call" code. A "no call previous bill" flag is then set at step 118, and the routine retums to the main program at step 119.
An affirmative answer at step 116 advances the system to step 120, which determines whether the sample data passes an ink stain test (described below). If the answer is negative, a "no call" code is generated at step 117. If die answer is affirmative, the system advances to step 121 which determines whether the best conelation number is greater than 849. An affirmative answer at step 121 indicates diat die conelation number is sufficiently high that the denomination of the scanned bill can be identified with certainty without any further checking. Consequently, a "denomination" code identifying the denomination represented by the stored pattem resulting in the highest conelation number is generated at step 122, and me system retums to the main program at step 119.
A negative answer at step 121 indicates that the conelation number is between 800 and 850. It has been found diat conelation numbers within this range are sufficient to identify all bills except the $2 bill. Accordingly, a negative response at step 121 advances die system to step 123 which determines whether the difference between the two highest conelation numbers ("call #1" and "call #2") is greater dian
149. If the answer is affirmative, the denomination identified by die highest conelation number is acceptable, and tiius die "denomination" code is generated at step 122. If the difference between the two highest conelation numbers is less than
150, step 123 produces a negative response which advances the system to step 117 to generate a "no call" code.
Returning to step 115, an affirmative response at this step indicates diat the initial call is a $2 bill. This affirmative response initiates a series of steps 124-127 which are identical to steps 116, 120, 121 and 123 described above, except that die numbers 799 and 849 used in steps 116 and 121 are changed to 849 and 899, respectively, in steps 124 and 126. The result is either the generation of a "no call" code at step 117 or the generation of a $2 "denomination" code at step 122. One problem encountered in cunency recognition and counting systems is the difficulty involved in interrupting (for a variety of reasons) and resuming the scanning and counting procedure as a stack of bills is being scanned. If a particular cunency recognition unit (CRU) has to be halted in operation due to a "major" system enor, such as a bill being jammed along the transport path, there is generally no concern about the outstanding transitional status of the overall recognition and counting process. However, where the CRU has to be halted due to a "minor" eπor, such as the identification of a scanned bill as being a counterfeit (based on a variety of monitored parameters) or a "no call" (a bill which is not identifiable as belonging to a specific cunency denomination based on the plurality of stored master pattems and/or other criteria), it is desirable that the transitional status of the overall recognition and counting process be retained so that die CRU may be restarted without any effective disruptions of the recognition/counting process. More specifically, once a scanned bill has been identified as a "no call" bill Bj) based on some set of predefined criteria, it is desirable that this bill Bj be transported directly to a retum conveyor or to the system stacker, and die CRU brought to a halt, while at the same time ensuring that the following bills are maintained in positions along the bill transport padi whereby CRU operation can be conveniently resumed without any disruption of the recognition/counting process.
Since fhe bill processing speeds at which cunency recognition systems must operate are substantially high (speeds of the order of 350 to 1500 bills per minute), it is practically impossible to totally halt the system following a "no call" without the following bill B2 already overlapping the optical scanhead and being partially scanned. As a result, it is virtually impossible for the CRU system to retain the transitional status of the recognition/counting process (particularly with respect to bill B2) in order that the process may be resumed once the bad bill B^ has been dealt with, and the system restarted. The basic problem is that if the CRU is halted with bill B2 only partially scanned, it is difficult to reference the data reflectance samples extracted therefrom in such a way that the scanning may be later continued (when die CRU is restarted) from exactly fhe same point where the sample extraction process was interrupted when die CRU was stopped.
Even if an attempt were made at immediately halting the CRU system following a "no call," any subsequent scanning of bills would be totally unreliable because of mechamcal backlash effects and the resultant disruption of the optical encoder routine used for bill scanning. Consequently, when the CRU is restarted, the call for the following bill is also likely to be bad and the overall recognition/counting process is totally disrupted as a result of an endless loop of "no calls." The above problems are solved by the use of a cunency detecting and counting technique whereby a scanned bill identified as a "no call" is transported directly to the return conveyor which retums the bill to the customer, while the CRU is halted without adversely affecting the data collection and processing steps for a succeeding bill. Accordingly, when the CRU is restarted, the overall bill recognition and counting procedure can be resumed without any disruption as if the CRU had never been halted at all. According to a prefened technique, if the bill is identified as a "no call" based on any of a variety of conventionally defined bill criteria, the CRU is subjected to a controlled deceleration process whereby the speed at which bills are moved across the scanhead is reduced from the normal operating speed. During diis deceleration process die "no call" bill (Bj) is transported to the retum conveyor, at the same time, the following bill B2 is subjected to die standard scanning procedure in order to identify fhe denomination.
The rate of deceleration is such that optical scanning of bill B2 is completed by the time the CRU operating speed is reduced to a predefined operating speed. While the exact operating speed at die end of the scanning of bill B2 is not critical, the objective is to permit complete scanning of bill B2 without subjecting it to backlash effects that would result if the ramping were too fast, while at the same time ensuring that bill B^ has in fact been transported to the return conveyor.
It has been experimentally determined that at nominal operating speeds of the order of 1000 bills per minute, die deceleration is preferably such that die CRU operating speed is reduced to about one-fifth of its normal operating speed at the end of e deceleration phase, i.e., by the time optical scanning of bill B2 has been completed. It has been determined diat at these speed levels, positive calls can be made as to the denomination of bill B2 based on reflectance samples gathered during the deceleration phase widi a relatively high degree of certainty (i.e., with a conelation number exceeding about 850).
Once die optical scanning of bill B2 has been completed, the speed is reduced to an even slower speed until die bill B2 has passed bill-edge sensors SI and S2 described below, and the bill B2 is then brought to a complete stop. At the same time, the results of the processing of scanned data coπesponding to bill B2 are stored in system memory. The ultimate result of this stopping procedure is that the CRU is brought to a complete halt following the point where the scanning of bill B2 has been reliably completed, and the scan procedure is not subjected to die disruptive effects (backlash, etc.) which would result if a complete halt were attempted immediately after bill Bj is identified as a "no call."
The reduced operating speed of the machine at the end of fhe deceleration phase is such diat the CRU can be brought to a total halt before the next following bill B3 has been transported over die optical scanhead. Thus, when die CRU is in fact halted, bill B^ is in die retum conveyor, bill B2 is maintained in transit between d e optical scanhead and die stacking station after it has been subjected to scanning, and the following bill B3 is stopped short of die optical scanhead. When the CRU is restarted, the overall scanning operation can be resumed in an uninterrupted fashion by using the stored call results for bill B2 as the basis for updating die system count appropriately, moving bill B2 from its earlier transitional position along the transport path into the stacking station, and moving bill B3 along the transport path into the optical scanhead area where it can be subjected to normal scanning and processing. A routine for executing the deceleration/ stopping procedure described above is illustrated by the flow chart in FIG. 14. This routine is initiated at step 170 with die CRU in its normal operating mode. At step 171 , a test bill Bj is scanned and the data reflectance samples resulting therefrom are processed. Next, at step 172, a determination is made as to whether or not test bill B^ is a "no call" using predefined criteria in combination with the overall bill recognition procedure, such as the routine of FIG. 13. If the answer at step 172 is negative, i.e. , the test bill Bj can be identified, step 173 is accessed where normal bill processing is continued in accordance with the procedures described above. If, however, die test bill B s found to be a "no call" at step 172, step 174 is accessed where CRU deceleration is initiated, e.g., die transport drive motor speed is reduced to about one-fifth its normal speed.
Subsequently, die "no call" bill B s guided to die return conveyor while, at the same time, the following test bill B2 is brought under the optical scanhead and subjected to die scanning and processing steps. The call resulting from the scanning and processing of bill B2 is stored in system memory at this point. Step 175 determines whether the scanning of bill B2 is complete. When the answer is negative, step 176 determines whether a preselected "bill timeout" period has expired so that the system does not wait for the scanning of a bill that is not present. An affirmative answer at step 176 results in the transport drive motor being stopped at step 179 while a negative answer at step 176 causes steps 175 and 176 to be reiterated until one of them produces an affirmative response. After the scanning of bill B2 is complete and before stopping die transport drive motor, step 178 determines whedier eidier of die sensors SI or S2 (described below) is covered by a bill. A negative answer at step 178 indicates diat die bill has cleared botii sensors SI and S2, and tiius die transport drive motor is stopped at step 179. This signifies the end of die deceleration/ stopping process. At diis point in time, bill B2 remains in transit while the following bill B3 is stopped on die transport path just short of the optical scanhead.
Following step 179, conective action responsive to me identification of a "no call" bill is conveniently undertaken, and d e CRU is then in condition for resuming fhe scanning process. Accordingly, the CRU can be restarted and die stored results conesponding to bill B2, are used to appropriately update the system count. Next, die identified bill B2 is guided along the transport path to the stacking station, and die CRU continues with its normal processing routine. While the above deceleration process has been described in the context of a "no call" enor, other minor enors (e.g., suspect bills, stranger bills in stranger mode, etc.) are handled in die same manner.
In cunency discrimination systems in which discrimination is based on die comparison of a patte obtained from scanning a subject bill to stored master pattems conesponding to various denominations, die pattems which are designated as master pattems significantiy influence the performance characteristics of the discrimination system. According to a prefeπed technique, a master pattem for a given denomination is generated by averaging a plurality of component pattems. Each component patte is generated by scanning a genuine bill of fhe given denomination. According to a first method, master pattems are generated by scanning a standard bill a pluraUty of times, typically three (3) times, and obtaining the average of conesponding data samples before storing the average as representing a master pattem. In other words, a master pattem for a given denomination is generated by averaging a plurality of component pattems, wherein all of the component pattems are generated by scanning a single genuine bill of "standard" quality of fhe given denomination. The "standard" bill is a slightiy used bill, as opposed to a crisp new bill or one which has been subject to a high degree of usage. Rather, the standard bill is a bill of good to average quality. Component pattems generated according to this first methods are illustrated in FIGs. 15a-15c. More specifically, FIGs. 15a-15c show three test pattems generated, respectively, for the forward scanning of a $1 bill along its green side, die reverse scanning of a $2 bill on its green side, and die forward scanning of a $100 bill on its green side. It should be noted diat, for purposes of clarity fhe test pattems in FIGS. 15a-15c were generated by using 128 reflectance samples per bill scan, as opposed to the prefeπed use of only 64 samples. The marked difference existing among coπesponding samples for these three test pattems is indicative of fhe high degree of confidence widi which cunency denominations may be called using the foregoing optical sensing and conelation procedure.
According to a second method, a master pattem for a given denomination is generated by scanning two or more standard bills of standard quality and obtaining a plurality of component pattems. These component pattems are then averaged in deriving a master pattem. For example, it has been found tiiat some genuine $5 bills have dark stairs on the Lincoln Memorial while omer genuine $5 bills have light stairs. To compensate for this variation, standard bills for which component pattems are derived may be chosen with at least one standard bill scanned having dark stairs and with at least one standard bill having light stairs. It has been found that an alternate method can lead to improved performance in a discrimination systems, especially with regards to certain denominations. For example, it has been found diat the printed indicia on a $10 bill has changed slightly with 1990 series bills incorporating security threads. More specifically, 1990 series $10 bills have a borderline-to-borderline dimension which is slightly greater than previous series $10 bills. Likewise it has been found that the scanned pattem of an old, semi-shrunken $5 bill can differ significantly from the scanned pattem of a new $5 bill.
According to a diird method, a master pattem for a given denomination is generated by averaging a plurality of component pattems, wherein some of the component pattems are generated by scanning one or more new bills of me given denomination, and some of the component pattems are generated by scanning one or more old bills of the given denomination. New bills are bills of good quality which have been printed in recent years and have a security thread incorporated therein (for those denominations in which security threads are placed). New bills are preferably relatively crisp. A new $10 bill is preferably a 1990 series or later bill of very high quality, meaning that fhe bill is in near mint condition. Old bills are bills exhibiting some shrinkage and often some discoloration. Shrinkage may result from a bill having been subjected to a relatively high degree of use. A new bill utilized in diis third method is of higher quality than a standard bill of die previous methods, while an old bill in diis third mediod is of lower quality than a standard bill.
The third method can be understood by considering Table 2 which summarizes the manner in which component pattems are generated for a variety of denominations.
Table 2. Component Scans bv Denomination
Denomination Scan Direction CP1 CP2 CP3
$1 Forward -0.2 std 0.0 std +0.2 std
$1 Reverse -0.2 std 0.0 std +0.2 std
$2, left Forward -0.2 std -0J5 std -OJ std
$2, left Reverse -0.2 std -0J5 std -OJ std
$2, right Forward 0.0 std +0J std +0.2 std
$2, right Reverse 0.0 std +0.1 std +0.2 std
$5 Forward -0.2 old 0.0 new +0.2 old (It str) (dk str) (It str)
$5 Reverse -0.2 old 0.0 new +0.2 old (It str) (dk str) (It str)
$10, left Forward -0.2 old -OJ new 0.0 old
$10, left Reverse 0.0 old +0J new +0.2 old
$10, right Forward +0.1 old +0.2 new +0.3 old
$10, right Reverse -0.2 old -0J5 new -OJ old
$20 Forward -0.2 old 0.0 new +0.2 old
$20 Reverse -0.2 old 0.0 new +0.2 old
$50 Forward -0.2 std 0.0 std +0.2 std
$50 Reverse -0.2 std 0.0 std +0.2 std
$100 Forward -0.2 std 0.0 std +0.2 std
$100 Reverse -0.2 std 0.0 std +0.2 std
Table 2 summarizes the position of the scanhead relative to the center of the green surface of United States currency as well as the type of bill to be scanned for generating component patterns for various denominations. The three component patterns ("CP") for a given denomination and for a given scan direction are averaged to yield a coπesponding master pattern. The eighteen (18) rows coπespond to the preferred method of storing eighteen (18) master patterns. The scanhead position is indicated relative to the center of the borderlined area of die bdi. Thus a position of "0.0" indicates that die scanhead is centered over me center of die borderlined area of the bill. Displacements to the left of center are indicated by negative numbers, while displacements to tiie right are indicated by positive numbers. Thus a position of "-0.2" indicates a displacement of 2/lOth of an inch to the left of die center of a bill, while a position of "+0J" indicates a displacement of 1/lOtiιs of an inch to the right of the center of a bill.
Accordingly, Table 2 indicates that component pattems for a $20 bill scanned in the forward direction are obtained by scanning an old $20 bill 2/10ths of a inch to the right and to the left of the center of the bill and by scanning a new $20 bill directly down the center of the bill. FIG. 15d is a graph illustrating these three pattems. These three pattems are then averaged to obtain the master pattem for a $20 bill scanned in the forward direction. FIG. 15e is a graph illustrating a pattem for a $20 bill scanned in the forward direction derived by averaging the pattems of FIG. 15d. This pattem becomes the coπesponding $20 master pattem after undergoing normalization. In generating the master patte s, one may use a scanning device in which a bill to be scanned is held stationary and a scanhead is moved over the bill. Such a device permits the scanhead to be moved laterally, left and right, over a bill to be scanned and tiius permits the scanhead to be positioned over the area of the bill which one wishes to scan, for example, 2/lOths of inch to the left of the center of the borderlined area.
As discussed above, for $10 bills two pattems are obtained in each scan direction with one pattem being scanned slightiy to the left of the center and one pattem being scanned slightiy to the right of the center. For $5 bills, it has been found that some $5 bills are printed with darker stairs ("dk str") on the picture of fhe Lincoln Memorial while others are printed with lighter stairs ("It str"). The effect of this variance is averaged out by using an old bill having light stairs and a new bill having dark stairs.
As can be seen from Table 2, for some bills, the third method of using old and new bills is not used; rather, a standard ("std") bill is used for generating all three component pattems as with the first method. Thus, the master pattem for a $1 bill scanned in die forward direction is obtained by averaging three component pattems generated by scanning a standard bill three times, once 2/10ths of an inch to the left, once down die center, and once 2/10ths of an inch to the right.
As illustrated by Table 2, a discrimination system may employ a combmation of methods wherein, for example, some master patte s are generated according die first method and some master pattems are generated according to the diird memod. Likewise, a discrimination system may combine the scanning of new, standard, and old bills to generate component pattems to be averaged in obtaining a master pattem. Additionally, a discrimination system may generate master patte s by scanning bills of various qualities and/or having various characteristics and then averaging the resultant pattems. Altematively, a discrimination system may scan multiple bills of a given quality for a given denomination, e.g., three new $50 bills, while scanning one or more bills of a different quality for a different denomination, e.g. , tiiree old and worn $1 bills, to generate component pattems to be averaged in obtaining master pattems. In order to accommodate or nullify the effect of such bill shrinking, the above-described conelation technique can be modified by use of a progressive shifting approach whereby a test pattem which does not coπespond to any of the master pattems is partitioned into predefined sections, and samples in successive sections are progressively shifted and compared again to the stored pattems in order to identify the denomination. It has experimentally been determined that such progressive shifting effectively counteracts any sample displacement resulting from shrinkage of a bill along the preselected dimension.
The progressive shifting effect is best illustrated by the conelation pattems shown in FIGs. 16a-e. For purposes of clarity, the illustrated pattems were generated using 128 samples for each bill scan as compared to the prefened use of 64 samples. FIG. 16a shows the conelation between a test pattem (represented by a heavy line) and a conesponding master pattem (represented by a thin line). It is clear from FIG. 16a that the degree of conelation between the two pattems is relatively low and exhibits a conelation factor of 606. The manner in which the conelation between these pattems is increased by employing progressive shifting is best illustrated by considering the conelation at the reference points designated as A-E along the axis defining the number of samples. The effect on conelation produced by "single" progressive shifting is shown in FIG. 16b which shows "single" shifting of the test pattem of FIG. 16a. This is effected by dividing the test patte into two equal segments each comprising 64 samples. The first segment is retained without any shift, whereas the second segment is shifted by a factor of one data sample. Under these conditions, it is found tiiat the conelation factor at the reference points located in the shifted section, particularly at point E, is improved.
FIG. 16c shows die effect produced by "double" progressive shifting whereby sections of the test pattem are shifted in three stages. This is accomplished by dividing the overall pattem into three approximately equal sized sections. Section one is not shifted, section two is shifted by one data sample (as in FIG. 16b), and section three is shifted by a factor of two data samples. With "double" shifting, it can be seen that the conelation factor at point E is further increased.
On a similar basis, FIG. 16d shows the effect on conelation produced by "triple" progressive shifting where the overall pattem is first divided into four approximately equal sized sections. Subsequently, section one is retained without any shift, section two is shifted by one data sample, section three is shifted by two data samples, and section four is shifted by three data samples. Under these conditions, the conelation factor at point E is seen to have increased again. FIG. 16e shows the effect on conelation produced by "quadruple" shifting, where the pattem is first divided into five approximately equal sized sections. The first four sections are shifted in accordance with the "triple" shifting approach of FIG. 16d, whereas the fifth section is shifted by a factor of four data samples. From FIG. 16e it is clear mat the conelation at point E is increased almost to me point of superimposition of the compared data samples.
In an altemative progressive shifting approach, the degree of shrinkage of a scanned bill is determined by comparing the length of the scanned bill, as measured by the scanhead, with die length of an "unshrunk" bill. This "unshrunk" length is pre-stored in the system memory. The type of progressive shifting, e.g., "single", "double", "triple", etc., applied to the test pattem is then directly based upon the measured degree of shrinkage. The greater the degree of shrinkage, the greater the number of sections into which the test pattem is divided. An advantage of this approach is mat only one conelation factor is calculated, as opposed to potentially calculating several conelation factors for different types of progressive shifting.
In yet another progressive shifting approach, instead of applying progressive shifting to the test patte , progressive shifting is applied to each of the master pattems. The master pattems in the system memory are partitioned into predefined sections, and samples in successive sections are progressively shifted and compared again to the scanned test pattem in order to identify the denomination. To reduce the amount of processing time, the degree of progressive shifting which should be applied to die master pattems may be determined by first measuring the degree of shrinkage of the scanned bill. By first measuring the degree of shrinkage, only one type of progressive shifting is applied to the stored master patte s.
Instead of rearranging the scanned test pattem or the stored master pattems, the system memory may contain pre-stored pattems coπesponding to various types of progressive shifting. The scanned test pattem is then compared to all of these stored pattems in the system memory. However, to reduce the time required for processing the data, this approach may be modified to first measure the degree of shrinkage and to dien select only tiiose stored pattems from the system memory which conespond to the measured degree of shrinkage for comparison with the scanned test patte . The advantage of using the progressive shifting approach, as opposed to merely shifting by a set amount of data samples across the overall test pattem, is that the improvement in conelation achieved in the initial sections of the pattem as a result of shifting is not neutralized or offset by any subsequent shifts in the test pattem. It is apparent from the above figures that the degree of conelation for sample points falling within the progressively shifted sections increases coπespondingly.
More importantly, the progressive shifting realizes substantial increases in the overall conelation factor resulting from pattem comparison. For instance, the original conelation factor of 606 (FIG. 16a) is increased to 681 by the "single" shifting shown in FIG. 16b. The "double" shifting shown in FIG. 16c increases the conelation number to 793, the "triple" shifting of FIG. 16d increases the conelation number to 906, and, finally, the "quadruple" shifting shown in FIG. 16e increases the overall conelation number to 960. Using the above approach, it has been determined tiiat used cunency bills which exhibit a high degree of shrinkage and which cannot be accurately identified as belonging to the conect cunency denomination when the conelation is performed without any shifting, can be identified with a high degree of certainty by using a progressive shifting approach, preferably by adopting "triple" or "quadruple" shifting.
The degree of conelation between a scanned pattem and a master pattem may be negatively impacted if the two pattems are not properly aligned with each odier. Such misalignment between pattems may in turn negatively impact upon the performance of a cunency identification system. Misalignment between patte s may result from a number of factors. For example, if a system is designed so that the scanning process is initiated in response to the detection of the thin borderline sunounding U.S. cunency or the detection of some other printed indicia such as die edge of printed indicia on a bill, stray marks may cause initiation of the scanning process at an improper time. This is especially true for stray marks in the area between the edge of a bill and die edge of the printed indicia on die bill. Such stray marks may cause the scanning process to be initiated too soon, resulting in a scanned pattem which leads a conesponding master patte . Altematively, where the detection of fhe edge of a bill is used to trigger the scanning process, misalignment between pattems may result from variances between the location of printed indicia on a bill relative to the edges of a bill. Such variances may result from tolerances permitted during the printing and/or cutting processes in the manufacture of currency. For example, it has been found that location of the leading edge of printed indicia on Canadian cunency relative to the edge of Canadian cunency may vary up to approximately 0.2 inches (approximately 0.5 cm). The problems associated with misaligned pattems may be overcome by removing data samples from one end of a pattem to be modified and adding data values on the opposite end equal to die data values contained in the conesponding sequence positions of the pattem to which the modified pattem is to be compared. This process may be repeated, up to a predetermined number of times, until a sufficientiy high conelation is obtained between the two pattems so as to permit the identity of a bill under test to be called. A prefened embodiment of the technique can be further understood by considering Table 3. Table 3 contains data samples generated by scanning the nanow dimension of Canadian $2 bills along a segment positioned about the center of the bill on the side opposite the portrait side. More specifically, the second column of Table 3 represents a scanned pattem generated by scanning a test Canadian $2 bill. The scanned pattem comprises 64 data samples aπanged in a sequence. Each data sample has a sequence position, 1-64, associated therewith. The fifth column represents a master pattem associated with a Canadian $2 bill. The master pattem likewise comprises a sequence of 64 data samples. The third and fourth columns represent the scanned patte after it has been modified in the forward direction one and two times, respectively. In the embodiment depicted in Table 3, one data sample is removed from the beginning of the preceding patte during each modification.
Table 3
Sequence Scanned Scanned Pattern Scanned Pattern Master Position Pattern Modified Once Modified Twice Pattern
1 93 50 -21 161
2 50 -21 50 100
3 -21 50 93 171
4 50 93 65 191
5 93 65 22 252
6 65 22 79 403
7 22 79 136 312
8 79 136 193 434
9 136 193 278 90
10 193 278 164 0
11 278 164 136 20
12 164 136 278 444
. . . . .
. . . ,
. . . . . Sequence Scanned Scanned Pattern Scanned Pattern Master Position Pattern Modified Once Modified Twice Pattern
52 -490 -518 -447 -1090
53 -518 -447 -646 -767
54 -447 -646 -348 -575
55 -646 -348 -92 -514
56 -348 -92 -63 -545
57 -92 -63 -205 -40
58 -63 -205 605 1665
59 -205 605 1756 1705
60 605 1756 1401 1685
61 1756 1401 1671 2160
62 1401 1671 2154 2271
63 1671 2154 *2240 2240
64 2154 *2210 *2210 2210
The modified pattem represented in the third column is generated by adding an additional data value to the end of the original scanned pattern sequence which effectively removes the first data sample of the original pattern, e.g., 93, from the modified patte . The added data value in the last sequence position, 64, is set equal to the data value contained in the 64th sequence position of the master pattern, e.g., 2210. This copying of the 64th data sample is indicated by an asterisk in the third column. The second modified pattern represented in the fourth column is generated by adding two additional data values to the end of the original scanned pattem which effectively removes the first two data samples of the original scanned, e.g., 93 and 50, from the second modified pattem. The last two sequence positions, 63 and 64, are filled with the data values contained in the 63rd and 64th sequence positions of the master pattem, e.g., 2240 and 2210, respectively. The copying of the 63rd and 64th data samples is indicated by asterisks in the fourth column.
In the example of Table 3, the printed area of the bill under test from which the scanned patte was generated was farther away from the leading edge of the bill than was the printed area of die bill from which the master pattem was generated. As a result, fhe scanned pattem trailed the master pattem. The prefened embodiment of the pattem generation method described in conjunction with Table 3 compensates for the variance of the distance between the edge of the bill and e edge of me printed indicia by modifying the scanned pattem in the forward direction. As a result of the modification mediod employed, the conelation between die original and modified versions of the scanned pattem and die master pattem increased from 705 for the original, unmodified scanned pattem to 855 for the first modified patte and to 988 for the second modified pattem. Accordingly, the bill under test which would otherwise have been rejected may now be properly called as a genuine $2 Canadian bill tiirough the employment of the pattem generation method discussed above.
Another modified discrimination technique can be understood with reference to the flowchart of FIGs. 17a-17c. The process of FIGs. 17a-17c involves a method of identifying a bill under test by comparing a scanned pattem retrieved from a bill under test with one or more master pattems associated with one or more genuine bills. After the process begins at step 128a, the scanned pattem is compared wim one or more master pattems associated with genuine bills (step 128b). At step 129 it is determined whether the bill under test can be identified based on the comparison at step 128b. This may be accomplished by evaluating the conelation between the scanned pattem and each of the master pattems. If the bill can be identified, the process is ended at step 130. Otherwise, one or more of the master pattems are designated for further processing at step 131. For example, all of the master pattems may be designated for further processing. Altematively, less than all of the master pattems may be designated based on a preliminary assessment about the identity of the bill under test. For example, only the master pattems which had the four highest conelation values with respect to the scanned pattem at step 128b might be chosen for further processing. In any case, the number of master pattems designated for further processing is Ml. At step 132, either the scanned pattem is designated for modification or the
Ml master pattems designated at step 131 are designated for modification. In a prefened embodiment, the scanned pattem is designated for modification and the master pattems remain unmodified. At step 133, it is designated whether forward modification or reverse modification is to be performed. This determination may be made, for example, by analyzing the beginning or ending data samples of the scanned pattem to determine whether the scanned pattem trails or leads the master pattems. At step 134, the iteration counter, I, is set equal to one. The iteration counter is used to keep track of how many times the working pattems have been modified. Then at step 135, the number of incremental data samples, R, to be removed during each iteration is set. For example, only one additional data sample may be removed from each working patte during each iteration in which case R is set equal to one. At step 136, it is determined whether the scanned pattem has been designated for modification. If it has, then the scanned pattem is replicated Ml times and the Ml replicated pattems, one for each of the Ml master pattems, are designated as working pattems at step 137. If the scanned pattem has not been designated for modification, then the Ml master pattems have been so designated, and die Ml master pattems are replicated and designated as working pattems at step 138.
Regardless of which patte or pattems were designated for modification, at step 139, it is determined whether forward or reverse modification is to be performed on the working pattems.
If forward modification is to be performed, the first R x I data samples from each working pattem are removed at step 140. The first R x I data samples may either be explicitly removed from the working patte s or be removed as a result of adding additional data samples (step 141) to the end of the patte and designating the beginning of the modified pattem to be the (R x I) + 1 sequence position of die original pattem. As a result of the modification, the data sample which was in the 64th sequence position in the original working pattem will be in the 64 - (R x I) sequence position. The added data values in the last R x I sequence positions of a working pattem are copied from the data samples in tiie last R x I sequence positions of a coπesponding non-designated pattem at step 141. After the above described modification, the working pattems are compared with either respective ones of the non-designated pattems (scanned pattem modified/Ml master pattems not designated for modification) or the non-designated pattem (Ml master pattems designated for modification/scanned pattem not designated for modification) at step 142. Altematively, if reverse modification is to be performed, the last R x I data samples from each working pattem are removed at step 143. The last R x I data samples may either be explicitly removed from me working pattems or be removed as a result of adding additional data samples (step 144) to the beginning of the pattem and designating the beginning of the modified pattem to start with the added data samples. As a result of the modification, the data sample which was in the 1st sequence position in die original working patte will be in the (R x I) + 1 sequence position. The added data samples in die first R x I sequence positions of a working pattem are copied from the data samples in the first R x I sequence positions of a coπesponding non-designated pattem at step 144. After the above described modification, the working pattems are compared with either respective ones of the non-designated pattems (scanned pattem modified/Ml master pattems not designated for modification) or the non-designated pattem (Ml master pattems designated for modification/ scanned pattem not designated for modification) at step 142. For example, if the scanned pattem is designated for forward modification and four master pattems are designated for further processing, four working pattems are generated from the scanned pattem at step 137, one for each of the four master patte s. If R is set to two at step 135, during the first iteration the last two data samples from each of the Ml master pattems are copied and added to die end of fhe Ml working pattems so as to become the last two sequence positions of the Ml working patte s, one working pattem being associated with each of fhe Ml master pattems. As a result, after the first iteration, four different working pattems are generated witii each working patte coπesponding to a modified version of the scanned pattem but with each having data values in its last two sequence positions copied from the last two sequence positions of a respective one of the Ml master patte s. After a second iteration, the last four sequence positions of each of the Ml master pattems are copied and added to die end of the Ml working pattems so as to become the last four sequence positions of a respective one of the Ml working pattems. As another example, if four master pattems are designated for further processing and the four designated master pattems are designated for forward modification, four working pattems are generated at step 138, one from each of the four designated master pattems. If R is set to two at step 135, during the first iteration the last two data samples of the scanned pattem are copied and added to the end of the Ml working pattems so as to become the last two sequence positions of the Ml working pattems, one working pattem being associated with each of the Ml master pattems. As a result, after the first iteration, four different working patte s are generated with each working pattem coπesponding to a modified version of a conesponding master pattem but widi each having data values in its last two sequence position copied from the last two sequence positions of the scanned pattem. After a second iteration, the last four sequence positions of the scanned pattem are copied and added to the end of the Ml working pattems so as to become the last four sequence positions of the Ml working pattems.
After the comparison at step 142, it is determined whether die bill under test can be identified at step 145. If the bill can be identified the process is ended at step 146. Otherwise, fhe iteration counter, I, is incremented by one (step 147), and die incremented iteration counter is compared to a maximum iteration number, T (step 148). If the iteration counter, I, is greater than the maximum iteration number, T, then a no call is issued (step 149a), meaning that a match sufficient to identify the bill under test was not obtained, and the process is ended (step 149b). Otherwise, if the iteration is not greater than the maximum iteration number, the modification process is repeated beginning with step 136.
The flowchart of FIGs. 17a- 17c is intended to illustrate one prefened embodiment of the above technique. However, it is recognized diat there are numerous ways in which the steps of the flowchart of FIGs. 17a- 17c may be rearranged or altered and yet still result in the comparison of the same pattems as would be compared if the steps of FIGs. 17a-17c were followed exactly. For example, instead of generating multiple working pattems, a single working pattem may be generated and the leading or trailing sequence positions successively altered before comparisons to conesponding non-designated pattems. Likewise, instead of generating multiple modified pattems directly from unmodified pattems, multiple modified pattems may be generated from the preceding modified pattems. For example, instead of generating a twice forward modified scanned pattem by removing the first two data samples from the original scanned pattem and copying the last 2R sequence positions of a conesponding master pattem and adding these data values to the end of the original scanned pattem, die first data sample of die single forward modified scanned pattem may be removed and one data sample added to the end of the single modified scanned pattem, and then die data samples in the last two sequence positions may be set equal to the data samples in the last 2R sequence positions of a coπesponding master pattem.
In a modification of the above technique, instead of copying data values from a scanned pattem into coπesponding sequence positions of modified master pattems, leading or trailing sequence positions of modified master pattems are filled with zeros.
In an alternate embodiment, modified master patte s are stored, for example in EPROM 60 of FIG. 7a, before a bill under test is scanned. In such an embodiment, a scanned pattem retrieved from a bill under test is compared to die modified master pattems stored in memory. Modified master pattems are generated by modifying a coπesponding master pattem in either the forward or backward direction, or both, and filling in any trailing or leading sequence positions widi zeros. An advantage of such a prefeπed embodiment is that no modification needs to be performed during the normal operation of an identification device incorporating such an embodiment. An example of a procedure involved in comparing test pattems to master pattems is illustrated at FIG. 18a which shows the routine as starting at step 150a. At step 151a, the best and second best conelation results (refened to in FIG. 18a as the "#1 and #2 answers") are initialized to zero and, at step 152a, the test pattem is compared with each of the sixteen or eighteen original master pattems stored in the memory. At step 153a, the calls conesponding to the two highest conelation numbers obtained up to that point are determined and saved. At step 154a, a post¬ processing flag is set. At step 155a the test pattem is compared with each of a second set of 16 or 18 master pattems stored in the memory. This second set of master pattems is the same as the 16 or 18 original master pattems except mat the last sample is dropped and a zero is inserted in front of the first sample. If any of the resulting conelation numbers is higher than the two highest numbers previously saved, the #1 and #2 answers are updated at step 156. Steps 155a and 156a are repeated at steps 157a and 158a, using a tiiird set of master pattems formed by dropping the last two samples from each of the 16 original master pattems and inserting two zeros in front of me first sample. At steps 159a and 160a the same steps are repeated again, but using only $50 and $100 master pattems formed by dropping die last tiiree samples from the original master pattems and adding three zeros in front of the first sample. Steps 161a and 162a repeat the procedure once again, using only $1, $5, $10 and $20 master patte s formed by dropping the 33rd sample, whereby original samples 34-64 become samples 33-63, and inserting a 0 as the new last sample. Finally, steps 163a and 164a repeat the same procedure, using master pattems for $10 and $50 bills printed in 1950, which differ significantly from bills of the same denominations printed in later years. This routine then retums to die main program at step 165a. The above multiple sets of master pattems may be pre-stored in EPROM 60.
A modified procedure involved in comparing test pattems to green-side master pattems is illustrated at FIG. 18b which shows the routine as starting at step 150b. At step 151b, the best and second best conelation results (refeπed to in FIG. 18b as the "#1 and #2 answers") are initialized to zero and, at step 152b, die test pattem is compared with each of the eighteen original green-side master pattems stored in the memory. At step 153b, the calls coπesponding to the two highest conelation numbers obtained up to that point are determined and saved. At step 154b, a post¬ processing flag is set. At step 155b the test pattem is compared with each of a second set of 18 green-side master pattems stored in the memory. This second set of master pattems is the same as the 18 original green-side master pattems except that the last sample is dropped and a zero is inserted in front of the first sample. If any of the resulting conelation numbers is higher than the two highest numbers previously saved, the #1 and #2 answers are updated at step 156b.
Steps 155b and 156b are repeated at steps 157b and 158b, using a third set of green-side master pattems formed by dropping the last two samples from each of the 18 original master pattems and inserting two zeros in front of die first sample. At steps 159b and 160b the same steps are repeated again, but using only $50 and $100 master pattems (two pattems for the $50 and four pattems for the $100) formed by dropping the last three samples from the original master pattems and adding tiiree zeros in front of the first sample. Steps 161b and 162b repeat the procedure once again, using only $1, $5, $10, $20 and $50 master pattems (four pattems for the $10 and two pattems for the otiier denominations) formed by dropping the 33rd sample whereby original samples 34-64 become samples 33-63, and inserting a 0 as the new last sample. Finally, steps 163b and 164b repeat die same procedure, using master pattems for $10 and $50 bills printed in 1950 (two pattems scanned along a center segment for each denomination), which differ significantly from bills of the same denominations printed in later years. This routine then retums to the main program at step 165b. The above multiple sets of master pattems may be pre-stored in EPROM 60.
In another modified embodiment where conditional black-side conelation is to be performed, a modified version of the routine designated as "CORRES" is initiated. The procedure involved in executing the modified version of CORRES is illustrated at FIG. 19a, which shows the routine as starting at step 180. Step 181 determines whether the bill has been identified as a $2 bill, and, if the answer is negative, step 182 determines whether die best conelation number ("call #1") is greater than 799. If the answer is negative, the conelation number is too low to identify the denomination of die bill with certainty, and at step 183b a black side conelation routine is called (described in more detail below in conjunction with FIGs. 19b-19c). An affirmative answer at step 182 advances the system to step 186, which determines whether the sample data passes an ink stain test (described below). If the answer is negative, a "no call" bit is set in a conelation result flag at step 183a. A "no call previous bill" flag is then set at step 184, and the routine retums to the main program at step 185. If die answer at step 186 is affirmative, the system advances to step 187 which determines whether the best conelation number is greater than 849. An affirmative answer at step 187 indicates tiiat the conelation number is sufficientiy high that the denomination of the scanned bill can be identified widi certainty without any further checking. Consequently, a "good call" bit is set in the conelation result flag at step 188. A separate register associated with the best conelation number (#1) may men be used to identify the denomination represented by the stored pattem resulting in the highest conelation number. The system retums to the main program at step 185. A negative answer at step 187 indicates that the conelation number is between 800 and 850. It has been found that conelation numbers within this range are sufficient to identify all bills except the $2 bill. Accordingly, a negative response at step 187 advances fhe system to step 189 which determines whether the difference between the two highest conelation numbers ("call #1" and "call #2") is greater man 149. If die answer is affirmative, the denomination identified by the highest conelation number is acceptable, and tiius the "good call" bit is set in die conelation result flag at step 188. If the difference between the two highest conelation numbers is less than 150, step 189 produces a negative response which advances the system to step 183b where the black side conelation routine is called.
Returning to step 181, an affirmative response at this step indicates diat die initial call is a $2 bill. This affirmative response initiates a series of steps 190-193 which are similar to steps 182, 186, 187 and 189 described above, except that die numbers 799 and 849 used in steps 182 and 187 are changed to 849 and 899, respectively, in steps 190 and 192. The result is either the setting of a "no call" bit in a conelation result flag at step 183a, the setting of the "good call" bit in the conelation result flag at step 188, or the calling of the black side conelation routine at step 183b.
Turning now to FIGs. 19b and 19c there is shown a flowchart illustrating the steps of the black side conelation routine called at step 183b of FIG. 19a. After the black side conelation routine is initiated at step 600, it is determined at step 602 whether the lower read head was the read head that scanned the black side of the test bill. If it was, the lower read head data is normalized at step 604. Otherwise, it is determined at step 606 whedier the upper read head was the read head that scanned the black side of the test bill. If it was, the upper read head data is normalized at step 608. If it cannot be determined which read head scanned the black side of the bill, then the pattems generated from both sides of the test bill are conelated against the green-side master pattems (see, e.g., step 110 of FIG. 12). Under such a circumstance, the "no call" bit in the conelation result flag is set at step 610, the "no call previous bill" flag is set at step 611, and the program retums to the calling point at step 612. After die lower read head data is normalized at step 604, or the upper read head data is normalized at step 608, it is determined whether the best green-side conelation number is greater than 700 at step 614. A negative response at step 614 results in the "no call" bit in the conelation result flag being set at step 610, the "no call previous bill" flag being set at step 611, and the program returning to the calling point at step 612. An affirmative response at step 614 results in a determination being made as to whether the best call from the green side conelation conesponds to a $20, $50, or $100 bill at step 616. A negative response at step 616 results in the "no call" bit in the conelation result flag being set at step 610, the "no call previous bill" flag being set at step 611, and the program retuming to the calling point at step 612.
If it is determined at step 616 diat die best call from the green side conelation conesponds to a $20, $50, or $100 bill, the scanned pattem from the black side is conelated against the black-side master pattems associated with the specific denomination and scan direction associated with the best call from tiie green side. According to a prefened embodiment, multiple black-side master pattems are stored for $20, $50 and $100 bills. For each of these denominations, three master pattems are stored for scans in the forward direction, and tiiree master pattems are stored for scans in die reverse direction, for a total of six pattems for each denomination. For a given scan direction, black-side master patte s are generated by scanning a coπesponding denominated bill along a segment located about the center of the naπow dimension of the bill, a segment slightiy displaced (0.2 inches) to the left of center, and a segment slightly displaced (0.2 inches) to die right of center.
For example, at step 618, it is determined whether the best call from the green side is associated with a forward scan of a $20 bill and, if it is, die normalized data from the black side of the test bill is coπelated against die black-side master pattems associated with a forward scan of a $20 bill at step 620. Next it is determined whether the black-side conelation number is greater than 900 at step 622. If it is, the "good call" bit in the conelation result flag is set at step 648, and the program returns to the calling point at step 646. If the black-side conelation number is not greater than 900, then the "no call bit" in the conelation result flag is set at step 642, the "no call previous bill" flag is set at step 644, and the program returns to the calling point at step 646. If it is determined that the best call from the green side is not associated with a forward scan of $20 bill at step 618, the program branches accordingly at steps 624 - 640 so diat the normalized data from the black side of the test bill is conelated against die appropriate black-side master pattems. The mechanical portions of the cunency scanning and counting module are shown in FIGS. 20a - 22. From the input receptacle, the bills are moved in seriatim from the bottom of the stack along a curved guideway 211 which receives bills moving downwardly and rearwardly and changes die direction of travel to a forward direction. The curvature of the guideway 211 conesponds substantially to the curved periphery of the drive roll 223 so as to form a nanow passageway for the bills along the rear side of the drive roll. The exit end of the guideway 211 directs die bills onto a linear path where the bills are scanned. The bills are transported with the nanow dimension of the bills maintained parallel to the transport path and the direction of movement at all times. Bills that are stacked on the bottom wall 205 of the input receptacle are stripped, one at a time, from the bottom of the stack. The bills are stripped by a pair of stripping wheels 220 mounted on a drive shaft 221 which, in turn, is supported across side plates 201, 202. The stripping wheels 220 project through a pair of slots formed in a cover 207. Part of the periphery of each wheel 220 is provided with a raised high-friction, serrated surface 222 which engages the bottom bill of die input stack as the wheels 220 rotate, to initiate feeding movement of the bottom bill from the stack. The serrated surfaces 222 project radially beyond the rest of the wheel peripheries so that the wheels "jog" the bill stack during each revolution so as to agitate and loosen die bottom cunency bill within the stack, diereby facilitating the stripping of the bottom bill from the stack.
The stripping wheels 220 feed each stripped bill B (FIG. 21a) onto a drive roll 223 mounted on a driven shaft 224 supported across the side plates 201 and 202. As can be seen most clearly in FIGs. 21a and 21b, the drive roll 223 includes a central smooth friction surface 225 formed of a material such as mbber or hard plastic. This smooth friction surface 225 is sandwiched between a pair of grooved surfaces 226 and 227 having serrated portions 228 and 229 formed from a high-friction material. The serrated surfaces 228, 229 engage each bill after it is fed onto the drive roll 223 by the stripping wheels 220, to frictionally advance the bill into the naπow arcuate passageway formed by die curved guideway 211 adjacent the rear side of the drive roll 223. The rotational movement of the drive roll 223 and the stripping wheels 220 is synchronized so that die serrated surfaces on the drive roll and the stripping wheels maintain a constant relationship to each other. Moreover, the drive roll 223 is dimensioned so diat the circumference of the outermost portions of the grooved surfaces is greater than the width W of a bill, so that the bills advanced by the drive roll 223 are spaced apart from each other. That is, each bill fed to the drive roll 223 is advanced by hat roll only when the serrated surfaces 228, 229 come into engagement with the bill, so that the circumference of the drive roll 223 determines die spacing between the leading edges of successive bills.
To avoid the simultaneous removal of multiple bills from the stack in the input receptacle, particularly when small stacks of bills are loaded into the machine, the stripping wheels 220 are always stopped with the raised, serrated portions 222 positioned below the bottom wall 205 of the input receptacle. This is accomplished by continuously monitoring the angular position of the serrated portions of the stripping wheels 220 via the encoder 32, and then controlling the stopping time of the drive motor so that the motor always stops the stripping wheels in a position where the serrated portions 222 are located beneath the bottom wall 205 of the input receptacle. Thus, each time a new stack of bills is loaded into the machine, those bills will rest on the smootii portions of the stripping wheels. This has been found to significantly reduce die simultaneous feeding of double or triple bills, particularly when small stacks of bills are involved. In order to ensure firm engagement between the drive roll 223 and the cunency bill being fed, an idler roll 230 urges each incoming bill against the smooth central surface 225 of tiie drive roll 223. The idler roll 230 is journalled on a pair of arms 231 which are pivotally mounted on a support shaft 232. Also mounted on the shaft 232, on opposite sides of the idler roll 230, are a pair of grooved guide wheels 233 and 234. The grooves in these two wheels 233, 234 are registered with the central ribs in the two grooved surfaces 226, 227 of the drive roll 223. The wheels 233, 234 are locked to the shaft 232, which in turn is locked against movement in the direction of the bill movement (clockwise as viewed in FIG. 20a) by a one-way spring clutch 235. Each time a bill is fed into die nip between the guide wheels 233, 234 and the drive roll 223, the clutch 235 is energized to turn the shaft 232 just a few degrees in a direction opposite the direction of bill movement. These repeated incremental movements distribute the wear uniformly around the circumferences of the guide wheels 233, 234. Although the idler roll 230 and the guide wheels 233, 234 are mounted behind die guideway 211, die guideway is apertured to allow the roll 230 and the wheels 233, 234 to engage the bills on the front side of die guideway. Beneadi the idler roll 230, a spring-loaded pressure roll 236 (FIGs. 20a and
21b) presses the bills into firm engagement with the smooth friction surface 225 of the drive roll as the bills curve downwardly along the guideway 211. This pressure roll 236 is journalled on a pair of arms 237 pivoted on a stationary shaft 238. A spring 239 attached to die lower ends of die arms 237 urges the roll 236 against die drive roll 233, through an aperture in the curved guideway 211.
At die lower end of fhe curved guideway 211, the bill being transported by the drive roll 223 engages a flat guide plate 240 which carries a lower scan head 18. Cunency bills are positively driven along the flat plate 240 by means of a transport roll arrangement which includes the drive roll 223 at one end of die plate and a smaller driven roll 241 at the other end of the plate. Both the driver roll 223 and the smaller roll 241 include pairs of smooth raised cylindrical surfaces 242 and 243 which hold the bill flat against the plate 240. A pair of O rings 244 and 245 fit into grooves formed in both the roll 241 and the roll 223 to engage the bill continuously between the two rolls 223 and 241 to transport the bill while helping to hold die bill flat against the guide plate 240.
The flat guide plate 240 is provided with openings through which the raised surfaces 242 and 243 of both the drive roll 223 and the smaller driven roll 241 are subjected to counter-rotating contact with conesponding pairs of passive transport rolls 250 and 251 having high-friction mbber surfaces. The passive rolls 250, 251 are mounted on the underside of the flat plate 240 in such a manner as to be freewheeling about their axes 254 and 255 and biased into counter-rotating contact with die conesponding upper rolls 223 and 241. The passive rolls 250 and 251 are biased into contact with the driven rolls 223 and 241 by means of a pair of H-shaped leaf springs 252 and 253 (see FIGs. 23 and 24). Each of the four roUs 250, 251 is cradled between a pair of parallel arms of one of the H-shaped leaf springs 252 and 253. The central portion of each leaf spring is fastened to the plate 240, which is fastened rigidly to the machine frame, so that the relatively stiff arms of the H- shaped springs exert a constant biasing pressure against the rolls and push diem against the upper rolls 223 and 241.
The points of contact between the driven and passive transport rolls are preferably coplanar with the flat upper surface of the plate 240 so that cunency bills can be positively driven along die top surface of the plate in a flat manner. The distance between the axes of the two driven transport rolls, and the conesponding counter- rotating passive rolls, is selected to be just short of the length of the most naπow dimension of the cunency bills. Accordingly, the bills are firmly gripped under uniform pressure between the upper and lower transport rolls within the scanhead area, thereby minimizing the possibility of bill skew and enhancing the reliability of the overall scanning and recognition process.
The positive guiding anangement described above is advantageous in that uniform guiding pressure is maintained on the bills as they are transported tiirough the optical scanhead area, and twisting or skewing of the bills is substantially reduced. This positive action is supplemented by die use of die H-springs 252, 253 for uniformly biasing the passive rollers into contact with the active rollers so that bill twisting or skew resulting from differential pressure applied to the bills along the transport path is avoided. The O-rings 244, 245 function as simple, yet extremely effective means for ensuring that the central portions of the bills are held flat. The location of a magnetic head 256 and a magnetic head adjustment screw
257 are illustrated in FIG. 23. The adjustment screw 257 adjusts the proximity of the magnetic head 256 relative to a passing bill and thereby adjusts the strength of the magnetic field in the vicinity of the bill.
FIG. 22 shows die mechanical aπangement for driving the various means for transporting cunency bills through the machine. A motor 260 drives a shaft 261 carrying a pair of pulleys 262 and 263. The pulley 262 drives the roll 241 through a belt 264 and pulley 265, and the pulley 263 drives the roll 223 through a belt 266 and pulley 267. Both pulleys 265 and 267 are larger than pulleys 262 and 263 in order to achieve the desired speed reduction from the typically high speed at which the motor 260 operates.
The shaft 221 of the stripping wheels 220 is driven by means of a pulley 268 provided thereon and linked to a coπesponding pulley 269 on the shaft 224 through a belt 270. The pulleys 268 and 269 are of the same diameter so that the shafts 221 and 224 rotate in unison.
As shown in FIG. 20b, the optical encoder 32 is mounted on the shaft of the roller 241 for precisely tracking the position of each bill as it is transported through the machine, as discussed in detail above in connection with the optical sensing and conelation technique.
The upper and lower scanhead assemblies are shown most clearly in FIGs. 25- 28. It can be seen that the housing for each scanhead is formed as an integral part of a unitary molded plastic support member 280 or 281 that also forms the housings for the light sources and photodetectors of the photosensors PSI and PS2. The lower member 281 also forms the flat guide plate 240 that receives the bills from the drive roll 223 and supports the bills as they are driven past the scanheads 18a and 18b.
The two support members 280 and 281 are mounted facing each other so that the lenses 282 and 283 of the two scanheads 18a, 18b define a nanow gap through which each bill is transported. Similar, but slightiy larger, gaps are formed by die opposed lenses of the light sources and photodetectors of the photosensors PSI and PS2. The upper support member 280 includes a tapered entry guide 280a which guides an incoming bill into the gaps between the various pairs of opposed lenses.
The lower support member 281 is attached rigidly to the machine frame. The upper support member 280, however, is mounted for limited vertical movement when it is lifted manually by a handle 284, to facilitate the clearing of any paper jams that occur beneath the member 280. To allow for such vertical movement, the member 280 is slidably mounted on a pair of posts 285 and 286 on the machine frame, with a pair of springs 287 and 288 biasing the member 280 to its lowermost position. Each of the two optical scanheads 18a and 18b housed in the support members
280, 281 includes a pair of light sources acting in combination to uniformly illuminate light strips of the desired dimension on opposite sides of a bill as it is transported across die plate 240. Thus, the upper scanhead 18a includes a pair of LEDs 22a, directing light downwardly through an optical mask on top of the lens 282 onto a bill traversing the flat guide plate 240 beneath the scanhead. The LEDs 22a are angularly disposed relative to the vertical axis of the scanhead so that their respective light beams combine to illuminate the desired light strip defined by an aperture in the mask. The scanhead 18a also includes a photodetector 26a mounted directly over the center of the illuminated strip for sensing the light reflected off die strip. The photodetector 26a is linked to the CPU 30 through the ADC 28 for processing the sensed data as described above. When die photodetector 26a is positioned on an axis passing through the center of the illuminated strip, the illumination by the LED's as a function of the distance from the central point "0" along the X axis, should optimally approximate a step function as illustrated by the curve A in FIG. 29. With die use of a single light source angularly displaced relative to a vertical axis through the center of the iUuminated strip, the variation in illumination by an LED typically approximates a Gaussian function, as illustrated by die curve B in FIG. 29.
The two LEDs 22a are angularly disposed relative to the vertical axis by angles α and β, respectively. The angles a and β are selected to be such that the resultant strip illumination by the LED's is as close as possible to the optimum distribution curve A in FIG. 29. The LED illumination distribution realized by this arrangement is illustrated by the curve designated as "C" in FIG. 29 which effectively merges the individual Gaussian distributions of each light source to yield a composite distribution which sufficiently approximates the optimum curve A.
In the particular embodiment of the scanheads 18a and 18b illustrated in die drawings, each scanhead includes two pairs of LEDs and two photodetectors for illuminating, and detecting light reflected from, strips of two different sizes. Thus, each mask also includes :two slits which are formed to allow light from the LEDs to pass through and illuminate light strips of the desired dimensions. More specifically, one slit illuminates a relatively wide strip used for obtaining the reflectance samples which coπespond to the characteristic pattem for a test bill. In a prefeπed embodiment, the wide slit has a length of about 0.500" and a width of about 0.050". The second slit forms a relatively naπow illuminated strip used for detecting the thin borderline sunounding the printed indicia on cunency bills, as described above in detail. In a prefened embodiment, the nanow slit 283 has a length of about 0.300" and a width of about 0.010".
In order to prevent dust from fouling the operation of the scanheads, each scanhead includes tiiree resilient seals or gaskets 290, 291, and 292. The two side seals 290 and 291 seal the outer ends of die LEDs 22, while the center seal 292 seals the outer end of the photodetector 26. Thus, dust cannot collect on either the light sources or the photodetectors, and cannot accumulate and block the slits dirough which light is transmitted from the sources to the bill, and from the bill to the photodetectors.
Doubling or overlapping of bills in the illustrative transport system is detected by two photosensors PSI and PS2 which are located on a common transverse axis that is perpendicular to the direction of bill flow (see e.g. , FIGs. 30a and 30b). The photosensors PSI and PS2 include photodetectors 293 and 294 mounted within the lower support member 281 in immediate opposition to coπesponding light sources 295 and 296 mounted in the upper support member 280. The photodetectors 293, 294 detect beams of light directed downwardly onto the bill transport path from the light sources 295, 296 and generate analog outputs which coπespond to the sensed light passing tiirough the bill. Each such output is converted into a digital signal by a conventional ADC convertor unit (not shown) whose output is fed as a digital input to and processed by the system CPU.
The presence of a bill adjacent the photosensors PSI and PS2 causes a change in the intensity of the detected light, and the conesponding changes in the analog outputs of the photodetectors 293 and 294 serve as a convenient means for density- based measurements for detecting the presence of "doubles" (two or more overlaid or overlapped bills) during die cunency scanning process. For instance, the photosensors may be used to collect a predefined number of density measurements on a test bill, and die average density value for a bill may be compared to predetermined density thresholds (based, for instance, on standardized density readings for master bills) to determine the presence of overlaid bills or doubles.
In order to prevent the accumulation of dirt on the light sources 295 and 296 and/or the photodetectors 293, 294 of the photosensors PS I and PS2, both the light sources and the photodetectors are enclosed by lenses mounted so close to the bill path that they are continually wiped by the bills. This provides a self-cleaning action which reduces maintenance problems and improves the reliability of the outputs from the photosensors over long periods of operation. The CPU 30, under control of software stored in d e EPROM 34, monitors and controls the speed at which the bill transport mechanism 16 transports bills from the bill separating station 14 to the bill stacking unit. Flowcharts of the speed control routines stored in the EPROM 34 are depicted in FIGs. 31-35. To execute more than the first step in any given routine, the cunency discriminating system 10 must be operating in a mode requiring the execution of the routine.
Referring first to FIG. 31, when a user places a stack of bills in the bill accepting station 12 for counting, the transport speed of the bill transport mechamsm 16 must accelerate or "ramp up" from zero to top speed. Therefore, in response to receiving the stack of bills in the bill accepting station 12, the CPU 30 sets a ramp-up bit in a motor flag stored in die memory unit 38. Setting the ramp-up bit causes the CPU 30 to proceed beyond step 300b of the ramp-up routine. If die ramp-up bit is set, the CPU 30 utilizes a ramp-up counter and a fixed parameter "ramp-up step" to incrementally increase the transport speed of the bill transport mechanism 16 until the bill transport mechanism 16 reaches its top speed. The "ramp-up step" is equal to the incremental increase in the transport speed of the bill transport mechamsm 16, and fhe ramp-up counter determines the amount of time between incremental increases in the bill transport speed. The greater the value of the "ramp-up step", the greater the increase in the transport speed of the bill transport mechanism 16 at each increment. The greater the maximum value of the ramp-up counter, the greater the amount of time between increments. Thus, the greater the value of the "ramp-up step" and the lesser the maximum value of the ramp-up counter, the lesser the time it takes the bill transport mechanism 16 to reach its top speed.
The ramp-up routine in FIG. 31 employs a variable parameter "new speed", a fixed parameter "full speed", and the variable parameter "transport speed". The "full speed" represents the top speed of die bill transport mechanism 16, while the "new speed" and "transport speed" represent the desired cuπent speed of the bill transport mechanism 16. To account for operating offsets of the bill transport mechanism 16, the "transport speed" of the bill transport mechanism 16 actually differs from the "new speed" by a "speed offset value". Outputting the "transport speed" to the bill transport mechanism 16 causes the bill transport mechanism 16 to operate at the transport speed. To incrementally increase me speed of the bill transport mechanism 16, the
CPU 30 first decrements the ramp-up counter from its maximum value (step 301). If the maximum value of the ramp-up counter is greater than one at step 302, the CPU 30 exits the speed control software in FIGs. 31-35 and repeats steps 300b, 301, and 302 during subsequent iterations of the ramp-up routine until the ramp-up counter is equal to zero. When the ramp-up counter is equal to zero, the CPU 30 resets the ramp-up counter to its maximum value (step 303). Next, the CPU 30 increases the "new speed" by the "ramp-up step" (step 304). If the "new speed" is not yet equal to the "full speed" at step 305, die "transport speed" is set equal to the "new speed" plus the "speed offset value" (step 306). The "transport speed" is output to the bill transport mechanism 16 at step 307 of the routine in FIG. 31 to change the speed of the bill transport mechanism 16 to the "transport speed". During subsequent iterations of the ramp-up routine, the CPU 30 repeats steps 300b-306 until the "new speed" is greater than or equal to the "full speed".
Once the "new speed" is greater than or equal to the "full speed" at step 305, the ramp-up bit in the motor flag is cleared (step 308), a pause-after-ramp bit in the motor flag is set (step 309), a pause-after-ramp counter is set to its maximum value (step 310), and the parameter "new speed" is set equal to the "full speed" (step 311). Finally, die "transport speed" is set equal to the "new speed" plus the "speed offset value" (step 306). Since fhe "new speed" is equal to the "full speed", outputting the "transport speed" to die bill transport mechanism 16 causes the bill transport mechanism 16 to operate at its top speed. The ramp-up routine in FIG. 31 smoothly increases the speed of fhe bill transport mechanism without causing jerking or motor spikes. Motor spikes could cause false triggering of the optical scanhead 18 such that the scanhead 18 scans non-existent bills. During normal counting, the bill transport mechanism 16 transports bills from the bill separating station 14 to the bill stacking unit at its top speed. In response to the optical scanhead 18 detecting a stranger, suspect or no call bill, however, the CPU 30 sets a ramp-to-slow-speed bit in the motor flag. Setting the ramp-to-slow- speεd bit causes die CPU 30 to proceed beyond step 312 of the ramp-to-slow-speed routine in FIG. 32 on the next iteration of the software in FIGs. 31-35. Using the ramp- o-slow-speed routine in FIG. 32, the CPU 30 causes the bill transport mechanism 16 to controllably decelerate or "ramp down" from its top speed to a slow speed. As the ramp-to-slow speed routine in FIG. 32 is similar to the ramp-up routine in FIG. 31, it is not described in detail herein.
It suffices to state that if the ramp-to-slow-speed bit is set in the motor flag, the CPU 30 decrements a ramp-down counter (step 313) and determines whether or not the ramp-down counter is equal to zero (step 314). If the ramp-down counter is not equal to zero, the CPU 30 exits the speed control software in FIGs. 31-35 and repeats steps 312, 313, and 314 of the ramp-to-slow-speed routine in FIG. 32 during subsequent iterations of the speed control software until the ramp-down counter is equal to zero. Once the ramp-down counter is equal to zero, the CPU 30 resets the ramp-down counter to its maximum value (step 315) and subtracts a "ramp-down step" from the variable parameter "new speed" (step 316). The "new speed" is equal to the fixed parameter "full speed" prior to initiating the ramp-to-slow-speed routine in FIG. 32.
After subtracting the "ramp-down step" from the "new speed", die "new speed" is compared to a fixed parameter "slow speed" (step 317). If die "new speed" is greater than the "slow speed", the "transport speed" is set equal to the "new speed" plus the "speed offset value" (step 318) and this "transport speed" is output to the bill transport mechanism 16 (step 307 of FIG. 31). During subsequent iterations of the ramp-to-slow-speed routine, the CPU 30 continues to decrement the "new speed" by die "ramp-down step" until the "new speed" is less than or equal to the
"slow speed". Once the "new speed" is less than or equal to the "slow speed" at step 317, the CPU 30 clears the ramp-to-slow-speed bit in the motor flag (step 319), sets die pause-after-ramp bit in the motor flag (step 320), sets the pause-after-ramp counter (step 321), and sets the "new speed" equal to the "slow speed" (step 322). Finally, die "transport speed" is set equal to d e "new speed" plus the "speed offset value" (step 318). Since the "new speed" is equal to the "slow speed", outputting the "transport speed" to the bill transport mechanism 16 causes the bill transport mechanism 16 to operate at its slow speed. The ramp-to-slow-speed routine in FIG. 32 smoothly decreases me speed of the bill transport mechanism 16 without causing jerking or motor spikes.
FIG. 33 depicts a ramp-to-zero-speed routine in which the CPU 30 ramps down the transport speed of the bill transport mechanism 16 to zero either from its top speed or its slow speed. In response to completion of counting of a stack of bills, the CPU 30 enters this routine to ramp down the transport speed of the bill transport mechanism 16 from its top speed to zero. Similarly, in response to the optical scanhead 18 detecting a stranger, suspect, or no call bill and the ramp-to- slow-speed routine in FIG. 32 causing the transport speed to be equal to a slow speed, die CPU 30 enters the ramp-to-zero-speεd routine to ramp down die transport speed from the slow speed to zero.
With the ramp-to-zero- speed bit set at step 323, the CPU 30 determines whether or not an initial-braking bit is set in the motor flag (step 324). Prior to ramping down die transport speed of die bill transport mechanism 16, the initial- braking bit is clear. Therefore, flow proceeds to fhe left branch of the ramp-to-zero- speed routine in FIG. 33. In this left branch, the CPU 30 sets the initial-braking bit in the motor flag (step 325), resets the ramp-down counter to its maximum value (step 326), and subtracts an "initial-braking step" from the variable parameter "new speed" (step 327). Next, the CPU 30 determines whether or not the "new speed" is greater than zero (step 328). If the "new speed" is greater than zero at step 328, fhe variable parameter "transport speed" is set equal to die "new speed" plus the "speed offset value" (step 329) and this "transport speed" is output to the bill transport mechanism 16 at step 307 in FIG. 31. During the next iteration of the ramp-to-zero-speed routine in FIG. 33, the
CPU 30 enters the right branch of the routine at step 324 because the initial-braking bit was set during die previous iteration of the ramp-to-zero-speed routine. With the initial-braking bit set, the CPU 30 decrements the ramp-down counter from its maximum value (step 330) and determines whether or not the ramp-down counter is equal to zero (step 331). If the ramp-down counter is not equal to zero, the CPU 30 immediately exits tiie speed control software in FIGs. 31-35 and repeats steps 323, 324, 330, and 331 of the ramp-to-slow-speed routine during subsequent iterations of the speed control software until the ramp-down counter is equal to zero. Once the ramp-down counter is equal to zero, the CPU 30 resets the ramp-down counter to its maximum value (step 332) and subtracts a "ramp-down step" from the variable parameter "new speed" (step 333). This "ramp-down step" is smaller than the "initial-braking step" so that the "initial-braking step" causes a larger decremental change in the transport speed of the bill transport mechanism 16 than that caused by the "ramp-down step".
Next, the CPU 30 determines whether or not the "new speed" is greater than zero (step 328). If the "new speed" is greater than zero, the "transport speed" is set equal to the "new speed" plus the "speed offset value" (step 329) and this "transport speed" is outputted to the bill transport mechanism 16 (step 307 in FIG. 31). During subsequent iterations of the speed control software, the CPU 30 continues to decrement the "new speed" by the "ramp-down step" at step 333 until the "new speed" is less than or equal to zero at step 328. Once the "new speed" is less dian or equal to the zero at step 328, the CPU 30 clears the ramp- to-zero-speed bit and the initial-braking bit in the motor flag (step 334), sets a motor-at-rest bit in the motor flag (step 335), and sets the "new speed" equal to zero (step 336). Finally, the "transport speed" is set equal to the "new speed" plus the "speed offset value" (step 329). Since the "new speed" is equal to zero, outputting the "transport speed" to the bill transport mechanism 16 at step 307 in FIG. 31 halts the bill transport mechanism 16.
Using the feedback loop routine in FIG. 35, the CPU 30 monitors and stabilizes the transport speed of the bill transport mechanism 16 when the bill transport mechanism 16 is operating at its top speed or at slow speed. To measure the transport speed of the bill transport mechanism 16, the CPU 30 monitors the optical encoder 32. While monitoring the optical encoder 32, it is important to synchronize the feedback loop routine with any transport speed changes of die bill transport mechanism 16. To account for the time lag between execution of the ramp- up or ramp-to-slow-speed routines in FIGs. 31-32 and the actual change in the transport speed of the bill transport mechanism 16, the CPU 30 enters a pause-after- ramp routine in FIG. 34 prior to entering the feedback loop routine in FIG. 35 if the bill transport mechanism 16 completed ramping up to its top speed or ramping down to slow speed during the previous iteration of the speed control software in FIGs. 31- 35.
The pause-after-ramp routine in FIG. 34 allows the bill transport mechanism 16 to "catch up" to the CPU 30 so that the CPU 30 does not enter the feedback loop routine in FIG. 35 prior to the bill transport mechanism 16 changing speeds. As stated previously, the CPU 30 sets a pause-after-ramp bit during step 309 of the ramp-up routine in FIG. 31 or step 320 of the ramp-to-slow-speed routine in FIG. 32. With the pause-after-ramp bit set, flow proceeds from step 337 of the pause- after-ramp routine to step 338, where the CPU 30 decrements a pause-after-ramp counter from its maximum value. If the pause-after-ramp counter is not equal to zero at step 339, the CPU 30 exits the pause-after-ramp routine in FIG. 34 and repeats steps 337, 338, and 339 of the pause-after-ramp routine during subsequent iterations of the speed control software until die pause-after-ramp counter is equal to zero. Once the pause-after-ramp counter decrements to zero, the CPU 30 clears the pause- after-ramp bit in the motor flag (step 340) and sets the feedback loop counter to its maximum value (step 341). The maximum value of the pause-after-ramp counter is selected to delay the CPU 30 by an amount of time sufficient to permit the bill transport mechanism 16 to adjust to a new transport speed prior to the CPU 30 monitoring the new transport speed with the feedback loop routine in FIG. 35. Referring now to the feedback loop routine in FIG. 35, if the motor-at-rest bit in the motor flag is not set at step 342, the CPU 30 decrements a feedback loop counter from its maximum value (step 343). If the feedback loop counter is not equal to zero at step 344, the CPU 30 immediately exits the feedback loop routine in FIG. 35 and repeats steps 342, 343, and 344 of the feedback loop routine during subsequent iterations of the speed control software in FIGs. 31-36 until the feedback loop counter is equal to zero. Once the feedback loop counter is decremented to zero, the CPU 30 resets :the feedback loop counter to its maximum value (step 345), stores the present count of the optical encoder 32 (step 346), and calculates a variable parameter "actual difference" between the present count and a previous count of the optical encoder 32 (step 347). The "actual difference" between the present and previous encoder counts represents the transport speed of the bill transport mechanism 16. The larger the "actual difference" between the present and previous encoder counts, die greater the transport speed of the bill transport mechanism. The CPU 30 subtracts the "actual difference" from a fixed parameter "requested difference" to obtain a variable parameter "speed difference" (step 348).
If die "speed difference" is greater than zero at step 349, the bill transport speed of the bill transport mechanism 16 is too slow. To counteract slower than ideal bill transport speeds, die CPU 30 multiplies the "speed difference" by a "gain constant" (step 354) and sets the variable parameter "transport speed" equal to the multiplied difference from step 354 plus the "speed offset value" plus a fixed parameter "target speed" (step 355). The "target speed" is a value tiiat, when added to the "speed offset value", produces the ideal transport speed. The calculated "transport speed" is greater than this ideal transport speed by the amount of the multiplied difference. If the calculated "transport speed" is nonetheless less than or equal to a fixed parameter "maximum allowable speed" at step 356, the calculated "transport speed" is output to the bill transport mechanism 16 at step 307 so that the bill transport mechanism 16 operates at the calculated "transport speed". If, however, die calculated "transport speed" is greater than the "maximum allowable speed" at step 356, the parameter "transport speed" is set equal to the "maximum allowable speed" (step 357) and is output to the bill transport mechanism 16 (step 307). If the "speed difference" is less than or equal to zero at step 349, the bill transport speed of the bill transport mechanism 16 is too fast or is ideal. To counteract faster than ideal bill transport speeds, the CPU 30 multiplies the "speed difference" by a "gain constant" (step 350) and sets the variable parameter "transport speed" equal to the multiplied difference from step 350 plus the "speed offset value" plus a fixed parameter "target speed" (step 351). The calculated "transport speed" is less than this ideal transport speed by the amount of the multiplied difference. If the calculated "transport speed" is nonetheless greater than or equal to a fixed parameter "minimum allowable speed" at step 352, the calculated "transport speed" is output to the bill transport mechanism 16 at step 307 so that the bill transport mechanism 16 operates at the calculated "transport speed". If, however, the calculated "transport speed" is less than die "minimum allowable speed" at step 352, the parameter "transport speed" is set equal to the "minimum allowable speed" (step 353) and is output to the bill transport mechanism 16 (step 307).
It should be apparent that the smaller the value of the "gain constant", the smaller the variations of the bill transport speed between successive iterations of the feedback control routine in FIG. 35 and, accordingly, the less quickly the bill transport speed is adjusted toward the ideal transport speed. Despite these slower adjustments in the bill transport speed, it is generally prefeπed to use a relatively small "gain constant" to prevent abrupt fluctuations in the bill transport speed and to prevent overshooting tiie ideal bill transport speed. A routine for using the outputs of the two photosensors PSI and PS2 to detect any doubling or overlapping of bills is illustrated in FIG. 36 by sensing the optical density of each bill as it is scanned. This routine starts at step 401 and retrieves the denomination determined for the previously scanned bill at step 402. This previously determined denomination is used for detecting doubles in the event that the newly scanned bill is a "no call", as described below. Step 403 determines whether the cunent bill is a "no call," and if the answer is negative, the denomination determined for the new bill is retrieved at step 404.
If the answer at step 403 is affirmative, the system jumps to step 405, so that the previous denomination retrieved at step 402 is used in subsequent steps. To permit variations in the sensitivity of the density measurement, a "density setting" is retrieved from memory at step 405. If the "density setting" has been turned off, this condition is sensed at step 406, and the system retums to the main program at step 413. If the "density setting" is not turned off, a denominational density comparison value is retrieved from memory at step 407. The memory preferably contains five different density values (for five different density settings, i.e., degrees of sensitivity) for each denomination. Thus, for a cunency set containing seven different denominations, the memory contains 35 different values. The denomination retrieved at step 404 (or step 402 in the event of a "no call"), and the density setting retrieved st step 405, determine which of the 35 stored values is retrieved at step 407 for use in the comparison steps described below. At step 408, the density comparison value retrieved at step 407 is compared to d e average density represented by the output of the photosensor PSI. The result of this comparison is evaluated at step 409 to determine whether the output of sensor SI identifies a doubling of bills for the particular denomination of bill determined at step 402 or 404. If the answer is negative, the system retums to the main program at step 413. If the answer is affirmative, step 410 then compares the retrieved density comparison value to die average density represented by the output of the second sensor PS2. The result of this comparison is evaluated at step 411 to determine whether the output of the photosensor PS2 identifies a doubling of bills. Affirmative answers at both step 409 and step 411 result in the setting of a "doubles eπor" flag at step 412, and die system then retums to the main program at step 413. The "doubles eπor" flag can, of course, be used to stop die bill transport motor.
FIG. 37 illustrates a routine that enables the system to detect bills which have been badly defaced by dark marks such as ink blotches, felt-tip pen marks and die like. Such severe defacing of a bill can result in such distorted scan data that the data can be interpreted to indicate the wrong denomination for the bill. Consequently, it is desirable to detect such severely defaced bills and dien stop die bill transport mechanism so that the bill in question can be examined by die operator. The routine of FIG. 37 retrieves each successive data sample at step 450b and then advances to step 451 to determine whether that sample is too dark. As described above, the output voltage from the photodetector 26 decreases as the darkness of the scanned area increases. Thus, the lower the output voltage from the photodetector, the darker the scanned area. For die evaluation carried out at step 451, a preselected threshold level for the photodetector output voltage, such as a threshold level of about 1 volt, is used to designate a sample that is "too dark."
An affirmative answer at step 451 advances the system to step 452 where a "bad sample" count is incremented by one. A single sample that is too dark is not enough to designate the bill as seriously defaced. Thus, the "bad sample" count is used to determine when:a preselected number of consecutive samples, e.g., ten consecutive samples, are determined to be too dark. From step 452, the system advances to step 453 to determine whether ten consecutive bad samples have been received. If fhe answer is affirmative, the system advances to step 454 where an eπor flag is set. This represents a "no call" condition, which causes die bill transport system to be stopped in the same manner discussed above. When a negative response is obtained at step 451, the system advances to step 455 where the "bad sample" count is reset to zero, so that diis count always represents the number of consecutive bad samples received. From step 455 the system advances to step 456 which determines when all the samples for a given bill have been checked. As long as step 456 yields a negative answer, the system continues to retrieve successive samples at step 450b. When an affirmative answer is produced at step 456, the system retums to the main program at step 457.
A routine for automatically monitoring and making any necessary coπections in various line voltages is illustrated in FIG. 38. This routine is useful in automatically compensating for voltage drifts due to temperature changes, aging of components and the like. The routine starts at step 550 and reads the output of a line sensor which is monitoring a selected voltage at step 550b. Step 551 determines whether the reading is below 0.60, and if the answer is affirmative, step 552 determines whether the reading is above 0.40. If step 552 also produces an affirmative response, the voltage is witiiin the required range and us the system returns to the main program step 553. If step 551 produces a negative response, an incremental coπection is made at step 554 to reduce the voltage in an attempt to retum it to the desired range. Similarly, if a negative response is obtained at step 552, an incremental coπection is made at step 555 to increase the voltage toward the desired range.
Because cuπencies come in a variety of sizes, sensors may be added to deterrnine fhe size of a bill to be scanned. These sensors are placed upstream of the scanheads. A prefened embodiment of size determining sensors is illustrated in FIG. 39. Two leading/trailing edge sensors 1062 detect the leading and trailing edges of a bill 1064 as it passes along the transport path. These sensors in conjunction with the encoder 32 (FIG. 2a-2b) may be used to determine the dimension of die bill along a direction parallel to the scan direction which in FIG. 39 is the nanow dimension (or width) of the bill 1064. Additionally, two side edge sensors 1066 are used to detect the dimension of a bill 1064 transverse to the scan direction which in FIG. 39 is the wide dimension (or lengdi) of the bill 1064. While the sensors 1062 and 1066 of FIG. 39 are optical sensors, other means of determining the size of a bill may be employed. Once the size of a bill is determined, the potential identity of the bill is limited to those bills having the same size. Accordingly, die area to be scanned can be tailored to the area or areas best suited for identifying the denomination and country of origin of a bill having the measured dimensions. While die printed indicia on U.S. cunency is enclosed within a thin borderline, the sensing of which may serve as a trigger to begin scanning using a wider slit, most cunencies of other cunency systems such as those from other countries do not have such a borderline. Thus the system described above may be modified to begin scanning relative to the edge of a bill for cunencies lacking such a borderline. Referring to FIG. 40, two leading edge detectors 1068 are shown. The detection of the leading edge 1069 of a bill 1070 by leading edge sensors 1068 triggers scanning in an area a given distance away from the leading edge of die bill 1070, e.g. , D^ or D2, which may vary depending upon the preliminary indication of the identity of a bill based on the dimensions of a bill. Altematively, the leading edge 1069 of a bill may be detected by one or more of the scanheads (to be described below) in a similar manner as that described with respect to FIGs. 7a and 7b. Altematively, the beginning of scanning may be triggered by positional information provided by the encoder 32 of FIG. 2a-2b, for example, in conjunction with the signals provided by sensors 1062 of FIG. 39, thus eliminating the need for leading edge sensors 1068.
However, when the initiation of scanning is triggered by die detection of the leading edge of a bill, the chance that a scanned pattem will be offset relative to a conesponding master pattem increases. Offsets can result from the existence of manufacturing tolerances which permit the location of printed indicia of a document to vary relative to the edges of the document. For example, the printed indicia on U.S. bills may vary relative to the leading edge of a bill by as much as 50 mils which is 0.05 inches (1.27 mm). Thus when scanning is triggered relative to the edge of a bill (rather than the detection of a certain part of the printed indicia itself, such as the printed borderline of U.S. bills), a scanned pattem can be offset from a coπesponding master pattem by one or more samples. Such offsets can lead to eπoneous rejections of genuine bills due to poor conelation between scanned and master pattems. To compensate, overall scanned patte s and master pattems can be shifted relative to each other as illustrated in FIGs. 41a and 41b. More particularly, FIG. 41a illustrates a scanned pattem which is offset from a coπesponding master pattem. FIG. 41b illustrates the same pattems after the scanned pattem is shifted relative to the master pattem, thereby increasing the conelation between the two pattems. Altematively, instead of shifting either scanned pattems or master pattems, master pattems may be stored in memory conesponding to different offset amounts. Thirdly, while it has been determined that die scanning of the central area on the green side of a U.S. bill (see segment S of FIG. 4) provides sufficientiy distinct pattems to enable discrimination among the plurality of U.S. denominations, the central area may not be suitable for bills originating in other countries. For example, for bills originating from Country 1, it may be determined that segment Sj (FIG. 40) provides a more preferable area to be scanned, while segment S2 (FIG. 40) is more preferable for bills originating from Country 2. Altematively, in order to sufficiently discriminate among a given set of bills, it may be necessary to scan bills which are potentially from such set along more than one segment, e.g. , scanning a single bill along both Sj_ and S2. To accommodate scanning in areas other than the central portion of a bill, multiple scanheads may be positioned next to each otiier. A prefeπed embodiment of such a multiple scanhead system is depicted in FIG. 42. Multiple scanheads 1072a-c and 1072d-f are positioned next to each other along a direction lateral to die direction of bill movement. Such a system permits a bill 1074 to be scanned along different segments. Multiple scanheads 1072a-f are arranged on each side of the transport path, thus permitting both sides of a bill 1074 to be scanned.
Two-sided scanning may be used to permit bills to be fed into a cunency discrimination system according to die present invention with either side face up. An example of a two-sided scanhead arrangement is described above in connection with FIGs. 2a, 6c, and 6d. Master pattems generated by scanning genuine bills may be stored for segments on one or both sides. In the case where master pattems are stored from the scanning of only one side of a genuine bill, the pattems retrieved by scanning both sides of a bill under test may be compared to a master set of single- sided master pattems. In such a case, a pattem retrieved from one side of a bill under test should match one of the stored master pattems, while a pattem retrieved from the other side of the bill under test should not match one of the master pattems. Altematively, master pattems may be stored for both sides of genuine bills. In such a two-sided system, a pattem retrieved by scanning one side of a bill under test should match with one of the master pattems of one side (Match 1) and a pattem retrieved from scanning the opposite side of a bill under test should match the master pattem associated widi tiie opposite side of a genuine bill identified by Match 1.
Altematively, in situations where the face orientation of a bill (i.e., whether a bill is "face up" or "face down") may be determined prior to or during characteristic pattem scanning, the number of comparisons may be reduced by limiting comparisons to pattems conesponding to the same side of a bill. That is, for example, when it is known that a bill is "face up", scanned patte s associated with scanheads above the transport path need only be compared to master pattems generated by scanning the "face" of genuine bills. By "face" of a bill it is meant a side which is designated as d e front surface of the bill. For example, the front or "face" of a U.S. bill may be designated as die "black" surface while the back of a U.S. bill may be designated as the "green" surface. The face orientation may be determinable in some situations by sensing the color of the surfaces of a bill. An altemative method of determining the face orientation of U.S. bills by detecting the borderline on each side of a bill is described above in connection with FIGs. 6c, 6d, and 12. The implementation of color sensing is discussed in more detailed below.
According to the embodiment of FIG. 42, the bill transport mechanism operates in such a fashion that the central area C of a bill 1074 is transported between central scanheads 1072b and 1072e. Scanheads 1072a and 1072c and likewise scanheads 1072d and 1072f are displaced the same distance from central scanheads 1072b and 1072e, respectively. By symmetrically arranging the scanheads about die central region of a bill, a bill may be scanned in either direction, e.g., top edge first (forward direction) or bottom edge first (reverse direction). As described above with respect to FIGs. l-7b, master pattems are stored from the scanning of genuine bills in both the forward and reverse directions. While a symmetrical aπangement is prefeπed, it is not essential provided appropriate master pattems are stored for a non-symmetrical system. While FIG. 42 illustrates a system having three scanheads per side, any number of scanheads per side may be utilized. Likewise, it is not necessary that there be a scanhead positioned over die central region of a bill. For example, FIG. 43 illustrates another prefeπed embodiment of the present invention capable of scanning the segments S1 and S2 of FIG. 40. Scanheads 1076a, 1076d, 1076e, and 1076h scan a bill 1078 along segment Sl while scanheads 1076b, 1076c, 1076f, and 1076g scan segment S2.
FIG. 44 depicts another prefeπed embodiment of a scanning system according to the present invention having laterally moveable scanheads 1080a-b. Similar scanheads may be positioned on the opposite side of the transport path. Moveable scanheads 1080a-b may provide more flexibility that may be desirable in certain scanning situations. Upon the determination of the dimensions of a bill as described in connection with FIG. 39, a preliminary determination of the identity of a bill may be made. Based on this preliminary determination, the moveable scanheads 1080a-b may be positioned over die area of the bill which is most appropriate for retrieving discrimination information. For example, if based on the size of a scanned bill, it is prek'minarily determined that the bill is a Japanese 5000 Yen bill-type, and if it has been determined that a suitable characteristic patte for a 5000 Yen bill-type is obtained by scanning a segment 2.0 cm to the left of center of the bill fed in the forward direction, scanheads 1080a and 1080b may be appropriately positioned for scanning such a segment, e.g., scanhead 1080a positioned 2.0 cm left of center and scanhead 1080b positioned 2.0 cm right of center. Such positioning permits proper discrimination regardless of the whether the scanned bill is being fed in the forward or reverse direction. Likewise scanheads on die opposite side of die transport path (not shown) could be appropriately positioned. Altematively, a single moveable scanhead may be used on one or both sides of the transport path. In such a system, size and color information (to be described in more detail below) may be used to properly position a single laterally moveable scanhead, especially where fhe orientation of a bill may be determined before scanning. FIG. 44 depicts a system in which the transport mechanism is designed to deliver a bill 1082 to be scanned centered witiiin the area in which scanheads 1080a-b are located. Accordingly, scanheads 1080a-b are designed to move relative to the center of the transport path with scanhead 1080a being moveable within the range Rj and scanhead 1080b being moveable witiiin range R2.
FIG. 45 depicts anodier prefeπed embodiment of a scanning system according to the present invention wherein bills to be scanned are transported in a left justified manner along the transport path, that is wherein the left edge L of a bill 1084 is positioned in the same lateral location relative to the transport path. Based on the dimensions of the bill, the position of the center of the bill may be determined and the scanheads 1086a-b may in rum be positioned accordingly. As depicted in FIG. 45, scanhead 1086a has a range of motion R3 and scanhead 1086b has a range of motion R4. The ranges of motion of scanheads 1086a-b may be influenced by the range of dimensions of bills which the discrimination system is designed to accommodate. Similar scanheads may be positioned on the opposite side of the transport path.
Altematively, the transport mechanism may be designed such that scanned bills are not necessarily centered or justified along die lateral dimension of the transport path. Rather the design of die transport mechanism may permit the position of bills to vary left and right within the lateral dimension of the transport path. In such a case, the edge sensors 1066 of FIG. 39 may be used to locate die edges and center of a bill, and tiius provide positional information in a moveable scanhead system and selection criteria in a stationary scanhead system.
In addition to the stationary scanhead and moveable scanhead systems described above, a hybrid system having both stationary and moveable scanheads may be used. Likewise, it should be noted diat the laterally displaced scanheads described above need not lie along die same lateral axis. That is, the scanheads may be, for example, staggered upstream and downstream from each other. FIG. 46 is a top view of a staggered scanhead arrangement according to a prefened embodiment of the present invention. As illustrated in FIG. 46, a bill 1130 is transported in a centered manner along the transport path 1132 so that the center 1134 of the bill 1130 is aligned widi die center 1136 of me transport path 1132. Scanheads 1140a-h are arranged in a staggered manner so as to permit scanning of the entire widdi of the transport path 1132. The areas illuminated by each scanhead are illustrated by strips 1142a, 1142b, 1142e, and 1142f for scanheads 1140a, 1140b, 1140e, and 1140f, respectively. Based on size determination sensors, scanheads 1140a and 1140h may either not be activated or their ouφut ignored.
In general, if prior to scanning a document, preliminary infonnation about a document can be obtained, such as its size or color, appropriately positioned stationary scanheads may be activated or laterally moveable scanheads may be appropriately positioned provided die preliminary information provides some indication as to die potential identity of the document. Alternatively, especially in systems having scanheads positioned over a significant portion of the transport path, many or all of die scanheads of a system may be activated to scan a document. Then subsequently, after some preliminary determination as to a document's identity has been made, only the output or derivations thereof of appropriately located scanheads may be used to generate scanned pattems. Derivations of output signals include, for example, data samples stored in memory generated by sampling output signals. Under such an altemative embodiment, information enabling a pretiminary determination as to a document's identity may be obtained by analyzing information either from sensors separate from the scanheads or from one or more of the scanheads themselves. An advantage of such preliminary determinations is that the number of scanned pattems which have to be generated or compared to a set of master pattems is reduced. Likewise the number of master pattems to which scanned pattems must be compared may also be reduced.
While the scanheads 1140a-h of FIG. 46 are arranged in a non-overlapping manner, they may altematively be arranged in an overlapping manner. By providing additional lateral positions, an overlapping scanhead arrangement may provide greater selectivity in the segments to be scanned. This increase in scanable segments may be beneficial in compensating for cunency manufacturing tolerances which result in positional variances of fhe printed indicia on bills relative to their edges. Additionally, in a prefened embodiment, scanheads positioned above fhe transport path are positioned upstream relative to their conesponding scanheads positioned below die transport path. FIGs. 47a and 47b illustrate another embodiment wherein a plurality of analog sensors 1150 such as photodetectors are laterally displaced from each other and are aπanged in a linear array within a single scanhead 1152. FIG. 47a is a top view while FIG. 47b is a side elevation view of such a linear array embodiment. The output of individual sensors 1150 are connected to photodetectors (not shown) through die use of graded index fibers, such as a "lens array" manufactured by MSG America, Inc., part number SLA20A1675702A3, and subsequently to analog-to- digital converters and a CPU (not shown) in a manner similar to that depicted in FIGs. 1 and 6a. As depicted in FIGs. 47a and 47b, a bill 1154 is transported past the linear array scanhead 1152 in a centered fashion. A prefened length for the linear array scanhead is about 6-7 inches (15 cm - 17 cm).
In a manner similar to that described above, based on the determination of the size of a bill, appropriate sensors may be activated and their output used to generate scanned pattems. Altematively many or all of the sensors may be activated with only the output or derivations thereof of appropriately located sensors being used to generate scanned pattems. Derivations of output signals include, for example, data samples stored in memory generated by sampling output signals. As a result, a discriminating system incorporating a linear array scanhead according the present invention would be capable of accommodating a wide variety of bill- types. Additionally, a linear array scanhead provides a great deal of flexibility in how information may be read and processed with respect to various bills. In addition to the ability to generate scanned patte s along segments in a direction parallel to the direction of bill movement, by appropriately processing scanned samples, scanned pattems may be "generated" or approximated in a direction perpendicular to the direction of bill movement. For example, if the linear array scanhead 1152 comprises one hundred and sixty (160) sensors 1150 over a length of 7 inches (17.78 cm) instead of taking samples for 64 encoder pulses from say 30 sensors, samples may be taken for 5 encoder pulses from all 160 cells (or all those positioned over the bill 1154). Altematively, 160 scanned pattems (or selected ones thereof) of 5 data samples each may be used for pattem comparisons. Accordingly, it can be seen that the data acquisition time is significantly reduced from 64 encoder pulses to only 5 encoder pulses. The time saved in acquiring data can be used to permit more time to be spent processing data and/or to reduce the total scanning time per bill thus enabling increased diroughput of the identification system. Additionally, the linear array scanhead permits a great deal of flexibility in tailoring the areas to be scanned. For example, it has been found diat the leading edges of Canadian bills contain valuable graphic information. Accordingly, when it is determined that a test bill may be a Canadian bill (or when the identification system is set to a Canadian cunency setting), the scanning area can be limited to the leading edge area of bills, for example, by activating many laterally displaced sensors for a relatively small number of encoder pulses.
FIG. 48 is a top view of anodier prefened embodiment of a linear array scanhead 1170 having a plurality of analog sensors 1172 such as photodetectors wherein a bill 1174 is transported past the scanhead 1170 in a non-centered manner. As discussed above, positional information from size-determining sensors may be used to select appropriate sensors. Altematively, the linear array scanhead itself may be employed to determine the size of a bill, thus eliminating the need for separate si-ze-determining sensors. For example, all sensors may be activated, data samples derived from sensors located on the ends of the linear array scanhead may be pretiminarily processed to determine the lateral position and die lengtii of a bill. The widdi of a bill may be determined either by employing separate leading/trailing edge sensors or pre-processing data samples derived from initial and ending cycle encoder pulses. Once size information is obtained about a bill under test, only the data samples retrieved from appropriate areas of a bill need be further processed. FIG. 49 is a top view of another embodiment of a linear scanhead 1180 having the ability to compensate for skewing of bills. Scanhead 1180 has a plurality of analog sensors 1182 and a bill 1184 is transported past scanhead 1180 in a skewed manner. Once the skew of a bill has been determined, for example through the use of leading edge sensors, readings from sensors 1182 along the scanhead 1180 may be appropriately delayed. For example, suppose it is determined that a bill is being fed past scanhead 1180 so that the left front comer of the bill reaches the scanhead five encoder pulses before the right front comer of the bill. In such a case, sensor readings along the right edge of the bill can be delayed for 5 encoder pulses to compensate for the skew. Where scanned pattems are to be generated over only a few encoder pulses, the bill may be treated as being fed in a non-skewed manner since the amount of lateral deviation between a scan along a skewed angle and a scan along a non-skewed angle is minimal for a scan of only a few encoder pulses. However, where it is desired to obtain a scan over a large number of encoder pulses, a single scanned pattem may be generated from me outputs of more than one sensor. For example, a scanned pattem may be generated by taking data samples from sensor 1186a for a given number of encoder pulses, then taking data samples from sensor 1186b for a next given number of encoder pulses, and then taking data samples from sensor 1186c for a next given number of encoder pulses. The number of given encoder pulses for which data samples may be taken from the same sensor is influenced by die degree of skew: die greater the degree of skew of the bill, the fewer the number of data samples which may be obtained before switching to the next sensor. Altematively, master pattems may be generated and stored for various degrees of skew, thus permitting a single sensor to generate a scanned patte from a bill under test.
With regard to FIGs. 47-49, while only a single linear array scanhead is shown, another linear array scanhead may be positioned on the opposite side of the transport path to permit scanning of either or both sides of a bill. Likewise, the benefits of using a linear array scanhead may also be obtainable using a multiple scanhead arrangement which is configured appropriately, such as depicted in FIG. 46 or a linear anangement of multiple scanheads.
In addition to size and scanned characteristic pattems, color may also be used to discriminate bills. For example, while all U.S. bills are printed in the same colors, e.g., a green side and a black side, bills from other countries often vary in color with the denomination of die bill. For example, a German 50 deutsche mark bill is brown in color while a German 100 deutsche mark bill is blue in color. Altematively, color detection may be used to determine die face orientation of a bill, such as where the color of each side of a bill varies. For example, color detection may be used to determine the face orientation of U.S. bills by detecting whether or not the "green" side of a U.S. bill is facing upwards. Separate color sensors may be added upstream of the scanheads described above. According to such an embodiment, color information may be used in addition to size information to preliminarily identify a bill. Likewise, color information may be used to deteπnine the face orientation of a bill, which determination may be used to select upper or lower scanheads for scanning a bill, or to compare scanned pattems retrieved from upper scanheads with a set of master pattems generated by scanning a coπesponding face while the scanned pattems retrieved from fhe lower scanheads are compared with a set of master pattems generated by scanning an opposing face. Altematively, color sensing may be incorporated into the scanheads described above. Such color sensing may be achieved by, for example, incorporating color filters, colored light sources, and/or dichroic beamsplitters into the cunency discrimination system of the present invention. Color information acquisition is described in more detail in co¬ pending U.S. application Serial No. 08/219,093 filed March 29, 1994, for a "Cunency Discriminator and Authenticator", incorporated herein by reference. Various color information acquisition techniques are described in U.S. Patent Nos. 4,841,358; 4,658,289; 4,716,456; 4,825,246; and 4,992,860.
The operation of a cunency discriminator according to one prefened embodiment may be further understood by referring to the flowchart of FIGs. 50a and 50b. In die process beginning at step 1100, a bill is fed along a transport path (step 1102) past sensors which measure die length and width of the bill (step 1104). These size determining sensors may be, for example, those illustrated in FIG. 39. Next at step 1106, it is deterrnined whether die measured dimensions of the bill match me dimensions of at least one bill stored in memory, such as EPROM 60 of FIG. 7a. If no match is found, an appropriate eπor is generated at step 1108. If a match is found, the color of the bill is scanned at step 1110. At step 1112, it is deterrnined whether the color of the bill matches a color associated with a genuine bill having the dimensions measured at step 1104. An enor is generated at step 1114 if no such match is found. However, if a match is found, a pretiminary set of potentially matching bills is generated at step 1116. Often, only one possible identity will exist for a bill having a given color and dimensions. However, the preUr nary set of step 1116 is not limited to tiie identification of a single bill-type, that is, a specific denomination of a specific cunency system; but rather, the preUrninary set may comprise a number of potential bill-types. For example, all U.S. bills have the same size and color. Therefore, the preliminary set generated by scanning a U.S. $5 bill would include U.S. bills of all denominations.
Based on die preliminary set (step 1116), selected scanheads in a stationary scanhead system may be activated (step 1118). For example, if the preliminary identification indicates that a bill being scanned has the color and dimensions of a German 100 deutsche mark bill, the scanheads over regions associated widi die scanning of an appropriate segment for a German 100 deutsche mark bill may be activated. Then upon detection of the leading edge of die bill by sensors 1068 of FIG. 40, the appropriate segment may be scanned. Altematively, all scanheads may be active widi only the scanning information from selected scanheads being processed. Altematively, based on the preliminary identification of a bill (step 1116), moveable scanheads may be appropriately positioned (step 1118).
Subsequently, the bill is scanned for a characteristic pattem (step 1120). At step 1122, the scanned pattems produced by the scanheads are compared with the stored master pattems associated with genuine bills as dictated by the prehminary set. By only malting comparisons with master pattems of bills within the preUminary set, processing time may be reduced. Thus for example, if the preliminary set indicated that the scanned bill could only possibly be a German 100 deutsche mark bill, then only the master patte or pattems associated with a German 100 deutsche mark bill need be compared to the scanned pattems. If no match is found, an appropriate enor is generated (step 1124). If a scanned pattem does match an appropriate master pattem, the identity of the bill is accordingly indicated (step 1126) and the process is ended (step 1128). While some of the embodiments discussed above entail a system capable of identifying a plurality of bill- types, the system may be adapted to identify a bill under test as either belonging to a specific bill-type or not. For example, the system may be adapted to store master information associated with only a single bill-type such as a United Kingdom 5 pound bill. Such a system would identify bills under test which were United Kingdom 5 pound bills and would reject all other bill- types.
The scanheads described above may be incorporated into a cunency identification system capable of identifying a variety of cunencies. For example, the system may be designed to accommodate a number of cunencies from different countries. Such a system may be designed to permit operation in a number of modes. For example, the system may be designed to permit an operator to select one or more of a plurality of bill-types which the system is designed to accommodate. Such a selection may be used to limit the number of master pattems with which scanned pattems are to be compared. Likewise, the operator may be permitted to select the manner in which bills will be fed, such as all bills face up, all bills top edge first, random face orientation, and/or random top edge orientation. Additionally, the system may be designed to permit output information to be displayed in a variety of formats to a variety of output devices, such as a monitor, LCD display, or printer. For example, the system may be designed to count the number of each specific bill-type identified and to tabulate the total amount of cunency counted for each of a plurality of cunency systems. For example, a stack of bills could be placed in the bill accepting station 12 of FIG. 2a-2b, and the output unit 36 of FIG. 2a-2b may indicate that a total of 370 British pounds and 650
German marks were counted. Altematively, the output from scanning the same batch of bills may provide more detailed information about the specific denominations counted, for example, one 100 pound bill, five 50 pound bills, and one 20 pound bill and thirteen 50 deutsche mark bills. In a cunency identification system capable of identifying a variety of bills from a number of countries, a manual selection device, such as a switch or a scrolling selection display, may be provided so that the customer may designate what type of cunency is to be discriminated. For example, in a system designed to accommodate botii Canadian and German cunency, the customer could turn a dial to the Canadian bill setting or scroll through a displayed menu and designate Canadian bills. By pre-declaring what type of cunency is to be discriminated, scanned pattems need only be compared to master pattems coπesponding to the indicated type of cunency, e.g., Canadian bills. By reducing the number of master pattems which have to be compared to scanned pattems, die processing time can be reduced. Altematively, a system may be designed to compare scanned pattems to all stored master pattems. In such a system, the customer need not pre-declare what type of cunency is to be scanned. This reduces die demands on the customer. Furthermore, such a system would permit the inputting of a mixture of bills from a number of countries. The system would scan each bill and automatically determine the issuing country and the denomination.
In addition to the manual and automatic bill-type discriminating systems, an alternate system employs a semi-automatic bill-type discriminating method. Such a system operates in a manner similar to the stranger mode described above. In such a system, a stack of bills is placed in the input hopper. The first bill is scanned and the generated scanned patte is compared with the master pattems associated with bills from a number of different countries. The discriminator identifies the country- type and the denomination of the bill. Then the discriminator compares all subsequent bills in the stack to the master pattems associated with bills only from the same country as the first bill. For example, if a stack of U.S. bills were placed in the input hopper and die first bill was a $5 bill, the first bill would be scanned. The scanned pattem would be compared to master pattems associated with bills from a number of countries, e.g. , U.S. , Canadian, and German bills. Upon determining that the first bill is a U.S. $5 bill, scanned patte s from the remaining bills in the stack are compared only to master patte s associated with U.S. bills, e.g., $1, $2, $5, $10, $20, $50, and $100 bills. When a bill fails to sufficiently match one of the compared patte s, the bill may be flagged as described above such as by stopping the transport mechanism while the flagged bill is retumed to the customer.
A cunency discriminating device designed to accommodate both Canadian and German cunency bills will now be described. According to this embodiment, a cunency discriminating device similar to that described above in connection with scanning U.S. cunency (see, e.g. , FIGs. 1-38 and accompanying description) is modified so as to be able to accept both Canadian and German cunency bills.
According to a prefeπed embodiment when Canadian bills are being discriminated, no mag:.≤tic sampling or authentication is performed.
Canadian bills have one side with a portrait (the portrait side) and a reverse side with a picture (the picture side). Likewise, German bills also have one side with a portrait (the portrait side) and a reverse side with a picture (the picture side). In a prefeπed embodiment, the discriminator is designed to accept either stacks of Canadian bills or stacks of German bills, the bills in the stacks being faced so that the picture side of all fhe bills will be scanned by a triple scanhead arrangement to be described in connection with FIG. 51. In a prefened embodiment, this triple scanhead replaces the single scanhead anangement housed in the unitary molded plastic support member 280 (see, e.g. , FIGs. 25 and 26). FIG. 51 is a top view of a triple scanhead arrangement 1200. The triple scanhead arrangement 1200 comprises a center scanhead 1202, a left scanhead 1204, and a right scanhead 1206 housed in a unitary molded plastic support member 1208. A bill 1210 passes under the anangement 1200 in the direction shown. O-rings are positioned near each scanhead, preferably two O-rings per scanhead, one on each side of a respective scanhead, to engage the bill continuously while transporting the bill between rolls 223 and 241 (FIG. 20a) and to help hold the bill flat against the guide plate 240 (FIG. 20a). The left 1204 and right 1206 scanhead are placed slightly upstream of the center scanhead 1202 by a distance D . In a prefened embodiment, D is 0.083 inches (0.21 cm). The center scanhead 1202 is centered over the center C of the transport path 1216. The center LQ of the left scanhead 1204 and the center RQ of the right scanhead 1206 are displaced laterally from center C of the transport path in a symmetrical fashion by a distance D4. In a prefened embodiment, D4 is 1.625 inches (4.128 cm). The scanheads 1202, 1204, and 1206 are each similar to the scanheads described above connection with FIGs. 1-38, except only a wide slit having a length of about 0.500 inch and a width of about 0.050 inch is utilized. The wide slit of each scanhead is used both to detect the leading edge of a bill and to scan a bill after the leading edge has been detected. Two photosensors 1212 and 1214 are located along the lateral axis of the left and right scanheads 1204 and 1206, one on either side of the center scanhead 1202. Photosensors 1212 and 1214 are same as the photosensors PSI and PS2 described above (see, e.g. , FIGs. 26 and 30). Photosensors 1212 and 1214 are used to detect doubles and also to measure the dimensions of bills in the direction of bill movement which in the prefened embodiment depicted in FIG. 51 is the nanow dimension of bills. Photosensors 1212 and 1214 are used to measure the nanow dimension of a bill by indicating when the leading and trailing edges of a bill passes by the photosensors 1212 and 1214. This information in combination with the encoder information permits the nanow dimension of a bill to be measured. All Canadian bills are 6 inches (15.24 cm) in their long dimension and 2.75 inches (6.985 cm) in their nanow dimension. German bills vary in size according to denomination. In a prefeπed embodiment of the cunency discriminating system, the discriminating device is designed to accept and discriminate $2, $5, $10, $20, $50. and $100 Canadian bills and 10 DM, 20 DM, 50 DM, and 100 DM German bills. These German bills vary in size from 13.0 cm (5J2 inches) in the long dimension by 6.0 cm (2.36 inches) in the nanow dimension for 10 DM bills to 16.0 cm (6.30 inches) in the long dimension by 8.0 cm (3.15 inches) in the nanow dimension for 100 DM bills. The input hopper of the discriminating device is made sufficientiy wide to accommodate all the above listed Canadian and German bills, e.g., 6.3 inches (16.0 cm) wide.
FIG. 52 is a top view of a Canadian bill illustrating the areas scanned by the triple scanhead anangement of FIG. 51. In generating scanned patte s from a
Canadian bill 1300 traveling along a transport path 1301 , segments SLj, SC^ , and SR^ are scanned by the left 1204, center 1202, and right 1206 scanheads, respectively, on the picture side of the bill 1300. These segments are similar to segment S in FIG. 4. Each segment begins a predetermined distance D^ inboard of the leading edge of the bill. In a prefened embodiment D^ is 0.5" (1.27 cm).
Segments SLj, SCj , and S j each comprise 64 samples as shown in FIGs. 3 and 5. In a prefened embodiment Canadian bills are scanned at a rate of 1000 bills per minute. The lateral location of segments SLj, SCj , and SR]_ is fixed relative to the transport path 1301 but may vary left to right relative to bill 1300 since the lateral position of bill 1300 may vary left to right within the transport path 1301.
A set of eighteen master Canadian pattems are stored for each type of Canadian bill that the system is designed to discriminate, three for each scanhead in both the forward and reverse directions. For example, three pattems are generated by scanning a given genuine Canadian bill in the forward direction with the center scanhead. One pattem is generated by scanning down the center of the bill along segment SC^, a second is generated by scanning along a segment SC2 initiated 1.5 samples before the beginning of SC]_, and a third is generated by scanning along a segment SC3 initiated 1.5 samples after the beginning of SCj. The second and third pattems are generated to compensate for the problems associated with triggering off the edge of a bill as discussed above.
To compensate for possible lateral displacement of bills to be scanned along a direction transverse to the direction of bill movement, the exact lateral location along which each of the above master pattems is generated is chosen after considering the conelation results achieved when a bill is displaced slightly to the left or to the right of the center of each scanhead, i.e., lines LQ, SQ, and RQ. For example, in generating a master pattem associated with segment SCi , a scan of a genuine bill may be taken down the center of a bill, a second scan may be taken along a segment 0J5 inch to the right of center (+0J5 inch), and a third scan may be taken along a segment 0J5 inch to the left of center (-0J5 inch). Based on the conelation result achieved, the actual scan location may be adjusted slightly to the right or left so the effect of the lateral displacement of a bill on the conelation results is minimized. Thus, for example, the master pattem associated with a forward scan of a Canadian $2 bill using the center scanhead 1202 may be taken along a line 0.05 inch to the right of the center of the bill.
Furthermore, the above stored master pattems are generated either by scanning both a relatively new crisp genuine bill and an older yellowed genuine bill and averaging the patte s generated from each or, altematively, by scanning an average looking bill.
Master pattems are stored for nine types of Canadian bills, namely, the newer series $2, $5, $10, $20, $50, and $100 bills and the older series $20, $50, and $100 bills. Accordingly, a total of 162 Canadian master pattems are stored (9 types x 18 per type).
FIG. 53 is a flowchart of the threshold test utilized in calling the denomination of a Canadian bill. When Canadian bills are being discriminated the flowchart of FIG. 53 replaces the flowchart of FIG. 13. The conelation results associated with conelating a scanned pattem to a master pattem of a given type of Canadian bill in a given scan direction and a given offset in the direction of bill movement from each of fhe three scanheads are summed. The highest of the resulting 54 summations is designated the #1 conelation and the second highest is preliminarily designated the #2 conelation. The #1 and #2 coπelations each have a given bill type associated with them. If the bill type associated with the #2 conelation is merely a different series from, but the same denomination as, the bill type associated with the #1 denomination, the preliminarily designated #2 conelation is substituted with the next highest conelation where the bill denomination is different from the denomination of the bill type associated with the #1 conelation.
The threshold test of FIG. 53 begins at step 1302. Step 1304 checks the denomination associated with the #1 conelation. If the denomination associated with the #1 conelation is not a $50 or $100, the #1 conelation is compared to a threshold of 1900 at step 1306. If the #1 conelation is less than or equal to 1900, the conelation number is too low to identify the denomination of the bill with certainty. Therefore, step 1308 sets a "no call" bit in a conelation result flag and the system retums to the main program at step 1310. If, however, the #1 conelation is greater than 1900 at step 1306, the system advances to step 1312 which determines whether the #1 conelation is greater than 2000. If the #1 conelation is greater than 2000, the conelation number is sufficiently high that the denomination of the scanned bill can be identified with certainty without any further checking. Consequently, a "good call" bit is set in the conelation result flag at step 1314 and the system retums to the main program at step 1310.
If the #1 conelation is not greater than 2000 at step 1312, step 1316 checks the denomination associated with the #2 conelation. If the denomination associated with the #2 conelation is not a $50 or $100, the #2 conelation is compared to a threshold of 1900 at step 1318. If the #2 conelation is less than or equal to 1900, the denomination identified by the #1 conelation is acceptable, and thus the "good call" bit is set in the conelation result flag at step 1314 and the system retums to the main program at step 1310. If, however, the #2 conelation is greater than 1900 at step 1318, the denomination of the scanned bill cannot be identified with certainty because the #1 and #2 conelations are both above 1900 and, yet, are associated with different denominations. Accordingly, the "no call" bit is set in the conelation result flag at step 1308.
If the denomination associated with the #2 conelation is a $50 or $100 at step 1316, the #2 conelation is compared to a threshold of 1500 at step 1320. If the #2 conelation is less than or equal to 1500, the denomination identified by the #1 conelation is acceptable, and thus the "good call" bit is set in the conelation result flag at step 1314 and the system retums to the main program at step 1310. If, however, the #2 conelation is greater than 1500 at step 1320, the denomination of the scanned bill cannot be identified with certainty. As a result, the "no call" bit is set in the conelation result flag at step 1308.
If the denomination associated with the #1 conelation is a $50 or $100 at step 1304, the #1 conelation is compared to a threshold of 1500 at step 1322. If the #1 conelation is less than or equal to 1500, the denomination of the scanned bill cannot be identified with certainty and, therefore, the "no call" bit is set in the conelation result flag at step 1308. If, however, the #1 conelation at step 1322 is greater than 1500, the system advances to step 1312 which determines whether the #1 conelation is greater than 2000. If the #1 conelation is greater than 2000, the conelation number is sufficiently high that the denomination of the scanned bill can be identified with certainty without any further checking. Consequently, a "good call" bit is set in the conelation result flag at step 1314 and the system retums to the main program at step 1310.
If the #1 conelation is not greater than 2000 at step 1312, step 1316 checks the denomination associated with the #2 conelation. If the denomination associated with the #2 conelation is not a $50 or $100, the #2 conelation is compared to a threshold of 1900 at step 1318. If the #2 conelation is less than or equal to 1900, the denomination identified by the #1 conelation is acceptable, and thus the "good call" bit is set in the conelation result flag at step 1314 and the system retums to the main program at step 1310. If, however, the #2 conelation is greater than 1900 at step 1318, the denomination of the scanned bill cannot be identified with certainty. Accordingly, the "no call" bit is set in the conelation result flag at step 1308.
If the denomination associated with the #2 conelation is a $50 or $100 at step 1316, the #2 conelation is compared to a threshold of 1500 at step 1320. If the #2 conelation is less than or equal to 1500, the denomination identified by the #1 conelation is acceptable, and thus the "good call" bit is set in the conelation result flag at step 1314 and the system retums to the main program at step 1310. If, however, the #2 conelation is greater than 1500 at step 1320, the denomination of the scanned bill cannot be identified with certainty. As a result, the "no call" bit is set in the conelation result flag at step 1308 and the system returns to the main program at step 1310. Now the use of the triple scanhead anangement 1200 in scanning and discriminating German cunency will be described. When scanning German bills, only the output of the center scanhead 1202 is utilized to generate scanned pattems. A segment similar to segment S of FIG. 4 is scanned over the center of the transport path at a predetermined distance Dg inboard after the leading edge of a bill is detected. In a prefened embodiment Dg is 0.25" (0.635 cm). The scanned segment comprises 64 samples as shown in FIGs. 3 and 5. In a prefened embodiment German bills are scanned at a rate of 1000 bills per minute. The lateral location of the scanned segment is fixed relative to the transport path 1216 but may vary left to right relative to bill 1210 since the lateral position of bill 1210 may vary left to right within the transport path 1216.
FIG. 54a illustrates the general areas scanned in generating master 10 DM German pattems. Due to the short length of 10 DM bills in their long dimension relative to the width of the transport path, thirty 10 DM master pattems are stored. A first set of five pattems are generated by scanning a genuine 10 DM bill 1400 in the forward direction along laterally displaced segments all beginning a predetermined distance Dg inboard of the leading edge of the bill 1400. Each of these five laterally displaced segments is centered about a respective one of lines Li -Lc. One such segment S10j centered about line L^ is illustrated in FIG. 54a. Line L^ is disposed down the center C of fhe bill 1400. In a prefeπed embodiment lines I-^-Lg are disposed in a symmetrical fashion about the center C of the bill 1400. In a prefeπed embodiment lines L2 and L3 are laterally displaced from Lj by a distance Dy where D7 is 0.24" (0.61 cm) and lines L and L5 are laterally displaced from L^ by a distance Dg where Dg is 0.48" (1.22 cm). A second set of five pattems are generated by scanning a genuine 10 DM bill
1400 in the forward direction along laterally displaced segments along lines L^-L^ all beginning at a second predetermined distance inboard of the leading edge of the bill 1400, the second predetermined distance being less than the predetermined distance Dg. One such segment SIO2 centered about line Lj is illustrated in FIG. 54a. In a prefened embodiment the second predetermined distance is such that scanning begins one sample earlier than Dg, that is about 30 mils before the initiation of the pattems in the first set of five pattems. A third set of five pattems are generated by scanning a genuine 10 DM bill 1400 in the forward direction along laterally displaced segments along lines j-L^ all beginning at a third predetermined distance inboard of the leading edge of the bill 1400, the third predetermined distance being greater than the predetermined distance Dg. One such segment SIO3 centered about line Lj is illustrated in FIG. 54a. In a prefened embodiment the third predetermined distance is such that scanning begins one sample later than Dg, that is about 30 mils after the initiation of the pattems in the first set of five pattems.
The above three sets of five pattems yield fifteen pattems in the forward direction. Fifteen additional 10 DM master pattems taken in the manner described above but in the reverse direction are also stored.
FIG. 54b illustrates the general areas scanned in generating master 20 DM, 50 DM, and 100 DM German pattems. Due to the lengths of 20 DM, 50 DM, and 100 DM bills in their long dimension being shorter than the width of the transport path, eighteen 20 DM master pattems, eighteen 50 DM master pattems, and eighteen 100 DM master pattems are stored. The 50 DM master patte s and the 100 DM master pattems are taken in the same manner as the 20 DM master pattems except mat the 50 DM master pattems and 100 DM master pattems are generated from respective genuine 50 DM bills and 100 DM bills while the 20 DM master pattems are generated from genuine 20 DM bills. Therefore, only the generation of the 20 DM master pattems will be described in detail.
A first set of three pattems are generated by scanning a genuine 20 DM bill 1402 in the forward direction along laterally displaced segments all beginning a predetermined distance Dg inboard of the leading edge of the bill 1402. Each of these three laterally displaced segments is centered about a respective one of lines Lg- Lg. One such segment S20j centered about line Lg is illustrated in FIG. 54b. Line Lg is disposed down the center C of the bill 1402. In a prefened embodiment lines Lγ-Lg are disposed in a symmetrical fashion about the center C of the bill 1402. In a prefened embodiment lines Lη and Lg are laterally displaced from Lg by a distance D9 where D9 is 0.30" (0.76 cm) for the 20 DM bill. The value of DQ is 0.20" (0.51 cm) for the 50 DM bill and 0.10" (0.25 cm) for the 100 DM bill. A second set of three pattems are generated by scanning a genuine 20 DM bill 1402 in the forward direction along laterally displaced segments along lines Lg-Lg all beginning at a second predetermined distance inboard of the leading edge of the bill 1402, the second predetermined distance being less than the predetermined distance Dg. One such segment S20-? centered about line Lg is illustrated in FIG. 54b. In a prefened embodiment the second predetermined distance is such that scanning begins one sample earlier than Dg, that is about 30 mils before the initiation of the pattems in the first set of three pattems.
A third set of three pattems are generated by scanning a genuine 20 DM bill 1402 in the forward direction along laterally displaced segments along lines Lg-Lg all beginning at a third predetermined distance inboard of the leading edge of the bill 1402, the third predetermined distance being greater than the predetermined distance Dg. One such segment S2O centered about line Lg is illustrated in FIG. 54b. In a prefened embodiment the third predetermined distance is such that scanning begins one sample later than Dg, that is about 30 mils after the initiation of the pattems in the first set of three patte s.
The above three sets of three pattems yield nine pattems in the forward direction. Nine additional 20 DM master patte s taken in the manner described above but in the reverse direction are also stored. Furthermore, the above stored master pattems are generated either by scanning both a relatively new crisp genuine bill and an older yellowed genuine bill and averaging the pattems generated from each or, altematively, by scanning an average looking bill.
This yields a total of 84 German master pattems (30 for 10 DM bills, 18 for 20 DM bills, 18 for 50 DM bills, and 18 for 100 DM bills). To reduce the number of master pattems that must compared to a given scanned pattem, the nanow dimension of a scanned bill is measured using photosensors 1212 and 1214. After a given bill has been scanned by the center scanhead 1202, the generated scanned patte is conelated only against certain ones of above described 84 master pattems based on the size of the naπow dimension of the bill as determined by the photosensors 1212 and 1214. The naπow dimension of each bill is measured independently by photosensors 1212 and 1214 and then averaged to indicate the length of the naπow dimension of a bill. In particular, a first number of encoder pulses occur between the detection of the leading and trailing edges of a bill by the photosensor 1212. Likewise, a second number of encoder pulses occur between the detection of the leading and trailing edges of the bill by the photosensor 1214. These first and second numbers of encoder pulses are averaged to indicate the length of the nanow dimension of the bill in terms of encoder pulses.
The photosensors 1212 and 1214 can also determine the degree of skew of a bill as it passes by the triple scanhead anangement 1200. By counting the number of encoder pulses between the time when photosensors 1212 and 1214 detect the leading edge of a bill, the degree of skew can be determined in terms of encoder pulses. If no or little skew is measured, a generated scanned pattern is only compared to master patterns associated with genuine bills having the same narrow dimension length. If a relatively large degree of skew is detected, a scanned pattern will be compared with master patte s associated with genuine bills having the next smaller denominational amount than would be indicated by the measured nanow dimension length.
Table 4 indicates which denominational set of master patterns are chosen for comparison to the scanned pattern based on the measured narrow dimension length in terms of encoder pulses and the measured degree of skew in terms of encoder pulses:
TABLE 4
Narrow Dimension Degree of Skew in Selected Set of Master Length in Encoder Pulses Encoder Pulses Patterns
< 1515 Not applicable 10 DM
> 1515 and < 1550 > 175 10 DM
> 1515 and < 1550 < 175 20 DM
> 1550 and < 1585 > 300 10 DM
> 1550 and < 1585 < 300 20 DM 1585 and < 1620 > 200 20 DM
> 1585 and < 1620 < 200 50 DM
> 1620 and < 1655 > 300 20 DM
> 1620 and < 1655 < 300 50 DM
> 1655 and < 1690 > 150 50 DM
> 1655 and < 1690 < 150 100 DM > 1690 and < 1725 > 300 50 DM
> 1690 and < 1725 < 300 100 DM
> 1725 Not applicable 100 DM
FIG. 55 is a flowchart of the threshold test utilized in calling the denomination of a German bill. It should be understood that this threshold test compares the scanned bill patte only to the set of master patterns selected in accordance with Table 4. Therefore, the selection made in accordance with Table 4 provides a preliminary indication as to the denomination of the scanned bill. The threshold test in FIG. 55, in effect, serves to confirm or overturn the preliminary indication given by Table 4.
The threshold test of FIG. 55 begins at step 1324. Step 1326 checks the nanow dimension length of the scanned bill in terms of encoder pulses. If the nanow dimension length is less than 1515 at step 1326, the preliminary indication is that the denomination of the scanned bill is a 10 DM bill. In order to confirm this preliminary indication, the #1 correlation is compared to 550 at step 1328. If the #1 conelation is greater than 550, the correlation number is sufficiently high to identify the denomination of the bill as a 10 DM bill. Accordingly, a "good call" bit is set in a correlation result flag at step 1330, and the system returns to the main program at step 1332. If, however, the #1 conelation is less than or equal to 550 at step 1328, the preliminary indication that the scanned bill is a 10 DM bill is effectively overturned. The system advances to step 1334 which sets a "no call" bit in the correlation result flag. If step 1326 determines that the nanow dimension length is greater than or equal to 1515, a conelation threshold of 800 is required to confirm the preliminary denominational indication provided by Table 4. Therefore, if the #1 conelation is greater than 800 at step 1336, the preliminary indication provided by Table 4 is confirmed. To confirm the preliminary indication, the "good call" bit is set in the conelation result flag. If, however, the #1 correlation is less than or equal to 800 at step 1336, the preliminary indication is rejected and the "no call" bit in the conelation result flag is set at step 1334. The system then retums to the main program at step 1332.
FIG. 56 is a functional block diagram illustrating another embodiment of a cunency discriminator system 1662. The discriminator system 1662 comprises an input receptacle 1664 for receiving a stack of cunency bills. A transport mechanism (as represented by aπows A and B) transports the bills in the input receptacle past an authenticating and discriminating unit 1666 to a canister 1668 where the bills are re- stacked. In addition to determining the denomination of each scanned bill, the authenticating and discriminating unit 1666 may additionally include various authenticating tests such as the ultraviolet authentication test described below.
Signals from the authenticating and discriminating unit 1666 are sent to a signal processor such as a central processor unit ("CPU") 1670. The CPU 1670 records the results of the authenticating and discriminating tests in a memory 1672. When the authenticating and discriminating unit 1666 is able to confirm the genuineness and denomination of a bill, the value of the bill is added to a total value counter in memory 1672 that keeps track of the total value of the stack of bills that was inserted in the input receptacle 1664 and scanned by the authenticating and discriminating unit 1666. Additionally, depending on the mode of operation of the discriminator system 1662, counters associated with one or more denominations are maintained in the memory 1672. For example, a $1 counter may be maintained to record how many $1 bills were scanned by the authenticating and discriminating unit 1666. Likewise, a $5 counter may be maintained to record how many $5 bills were scanned, and so on. In an operating mode where individual denomination counters are maintained, the total value of the scanned bills may be determined without maintaining a separate total value counter. The total value of the scanned bills and/or the number of each individual denomination may be displayed on a display 1674 such as a monitor or LCD display.
As discussed above, a discriminating unit such as the authenticating and discriminating unit 1666 may not be able to identify the denomination of one or more bills in the stack of bills loaded into the input receptacle 1664. For example, if a bill is excessively worn or soiled or if the bill is torn, a discriminating unit may not be able to identify the bill. Furthermore, some known discrimination methods do not have a high discrimination efficiency and thus are unable to identify bills which vary even somewhat from an "ideal" bill condition or which are even somewhat displaced by the transport mechanism relative to the scanning mechanism used to discriminate bills. Accordingly, such poorer performing discriminating units may yield a relatively large number of bills which are not identified.
The discriminator system 1662 may be designed so that when the authenticating and discriminating unit is unable to identify a bill, the transport mechanism is altered to divert the unidentified bill to a separate storage canister. Such bills may be "flagged" or "marked" to indicate that the bill is a no call or suspect bill. Altematively, the unidentified bill may be retumed to the customer. The discriminator system 1662 may be designed to continue operation automatically when a bill is diverted from the normal transport path because the bill is a "no call" or a counterfeit suspect, or the system may be designed to require a selection element to be depressed. For example, upon examination of a returned bill the customer may conclude that the retumed bill is genuine even though it was not identified by the discriminating unit. However, because the bill was not identified, the total value and/or denomination counters in the memory 1672 will not reflect its value. Nevertheless, the customer may wish to deposit the bill for subsequent verification by the bank. Tuming now to FIG. 57, there is shown a functional block diagram illustrating another embodiment of a document authenticator and discriminator according to the present invention. The discriminator system 1680 comprises an input receptacle 1682 for receiving a stack of cunency bills. A transport mechanism (as represented by aπow C) transports the bills from the input receptacle, one at a time, past an authenticating and discriminating unit 1684. Based on the results of the authenticating and discriminating unit 1684, a bill is either transported to a verified- deposit canister 1686 (aπow D), to an escrow canister 1688 (aπow E), or to a retum station 1690 (aπow F). When is bill is determined to be genuine and its denomination has been identified, the bill is transported to the verified-deposit canister 1686. Altematively, where the authenticating and discriminating unit determines that a bill is a fake, the bill is immediately routed (aπow E) to the escrow canister 1688. Finally, if a bill is not determined to be fake but for some reason the authenticating and discriminating unit 1684 is not able to identify the denomination of the bill, the flagged bill is retumed (aπow F) to the customer at station 1690. If the customer concludes that the bill is genuine, the customer may deposit the retumed bill or bills in an envelope for later verification by the bank and crediting to the customer's account. The discriminator system 1680 then resumes operation, and the suspect bills in the deposit envelope are held for manual pick-up without incrementing the counters associated with the various denomination and/or the total value counters.
Referring now to FIGs. 58-58, there is shown a document authenticating system using ultraviolet ("UV") light. A UV light source 2102 illuminates a document 2104. Depending upon the characteristics of the document, ultraviolet light mav be reflected off the document and/or fluorescent light mav be emitted from the document. A detection system 2106 is positioned so as to receive any light reflected or emitted toward it but not to receive any UV light directly from the light source 2102. The detection system 2106 comprises a UV sensor 2108, a fluorescence sensor 2110, filters, and a plastic housing. The light source 2102 and the detection system 2106 are both mounted to a printed circuit board 2112. The document 2104 is transported in the direction indicated by aπow A by a transport system (not shown). The document is transported over a transport plate 2114 which has a rectangular opening 2116 in it to permit passage of light to and from the document. In a prefeπed embodiment, the rectangular opening 2116 is 1.375 inches (3.493 cm) by 0.375 inches (0.953 cm). To minimize dust accumulation onto the light source 2102 and the detection system 2106 and to prevent document jams, the opening 2116 is covered with a transparent UV-transmitting acrylic window 2118. To further reduce dust accumulation, the UV light source 2102 and the detection system 2106 are completely enclosed within a housing (not shown) comprising the transport plate 2114.
Referring now to FIG. 59, there is shown a functional block diagram illustrating a prefeπed embodiment of a UV authenticating system. FIG. 59 shows a UV sensor 2202, a fluorescence sensor 2204, and filters 2206, 2208 of a detection system such as the detection system 2106 of FIG. 59. Light from the document passes through the filters 2206, 2208 before striking the sensors 2202, 2204, respectively. An ultraviolet filter 2206 filters out visible light and permits UV light to be transmitted and hence to strike the UV sensor 2202. Similarlv. a visible light filter 2208 filters out UV light and permits visible light to be transmitted and hence to strike fluorescence sensor 2204. Accordingly, UV light, which has a wavelength below 400 nm, is prevented from striking the fluorescence sensor 2204, and visible light, which has a wavelength greater than 400 nm, is prevented from striking the UV sensor 2202. In a prefeπed embodiment the UV filter 2206 transmits light having a wavelength between about 260 nm and about 380 nm and has a peak transmittance at 360 nm. In a prefened embodiment, the visible light filter 2208 is a blue filter and preferably transmits light having a wavelength between about 415 nm and about 620 nm and has a peak transmittance at 450 nm. The prefeπed blue filter comprises a combination of a blue component filter and a yellow component filter. The blue component filter transmits light having a wavelength between about 320 nm and about 620 nm and has a peak transmittance at 450 nm. The yellow component filter transmits light having a wavelength between about 415 nm and about 2800 nm. Examples of suitable filters are UG1 (UV filter), BG23 (blue bandpass filter), and GG420 (yellow longpass filter), all manufactured by Schott.
The UV sensor 2202 outputs an analog signal proportional to the amount of light incident thereon, and this signal is amplified by amplifier 2210 and fed to a microcontroller 2212. Similarly, the fluorescence sensor 2204 outputs an analog signal proportional to the amount of light incident thereon and this signal is amplified by amplifier 2214 and fed to a microcontroller 2212. Analog-to-digital converters 2216 within the microcontroller 2212 convert the signals from the amplifiers 2210, 2214 to digital and these digital signals are processed by the software of the microcontroller 2212. The UV sensor 2202 may be, for example, an ultraviolet enhanced photodiode sensitive to light having a wavelength of about 360 nm and the fluorescence sensor 2204 may be a blue enhanced photodiode sensitive to light having a wavelength of about 450 nm. Such photodiodes are available from, for example, Advanced Photonix, Inc., Massachusetts. The microcontroller 2212 may be, for example, a Motorola 68HC16. The exact characteristics of the sensors 2202, 2204 and the filters 2206, 2208 including the wavelength transmittance ranges of the above filters are not as critical as the prevention of the fluorescence sensor from generating an output signal in response to ultraviolet light, and the prevention of the ultraviolet sensor from generating an output signal in response to visible light. For example, instead of, or in addition to, filters, the authentication system may employ an ultraviolet sensor which is not responsive to light having a wavelength longer than 400 nm and/or a fluorescence sensor which is not responsive to light having a wavelength shorter than 400 nm.
Calibration potentiometers 2218, 2220 permit the gains of amplifiers 2210, 2214 to be adjusted to appropriate levels. Calibration may be performed by positioning a piece of white fluorescent paper on the transport plate 2114 so that it completely covers the rectangular opening 2116. The potentiometers 2218, 2220 may then be adjusted so that the output of the amplifiers 2210, 2214 is 5 volts.
It has been determined that genuine United States cunency reflects a high level of ultraviolet light and does not fluoresce under ultraviolet illumination. It has also been determined that under ultraviolet illumination counterfeit United States cunency exhibits one of the four sets of characteristics listed below:
1) Reflects a low level of ultraviolet light and fluoresces;
2) Reflects a low level of ultraviolet light and does not fluoresce;
3) Reflects a high level of ultraviolet light and fluoresces;
4) Reflects a high level of ultraviolet light and does not fluoresce. Counterfeit bills in categories (1) and (2) may be detected by a cunency authenticator employing an ultraviolet light reflection test. Counterfeit bills in category (3) may be detected by a cunency authenticator employing both an ultraviolet reflection test and a fluorescence test. Only counterfeits in category (4) are not detected by the authenticating methods of the present invention. Fluorescence is determined by any signal that is above the noise floor. Thus, the amplified fluorescent sensor signal 2222 will be approximately 0 volts for genuine U.S. cunency and will vary between approximately 0 and 5 volts for counterfeit bills, depending upon their fluorescence characteristics. Accordingly, an authenticating system will reject bills when signal 2222 exceeds approximately 0 volts.
A high level of reflected UV light ("high UV") is indicated when the amplified UV sensor signal 2224 is above a predetermined threshold. The high/low UV threshold is a function of lamp intensity and reflectance. Lamp intensity can degrade by as much as 50% over the life of the lamp and can be further attenuated by dust accumulation on the lamp and the sensors. The problem of dust accumulation is mitigated by enclosing the lamp and sensors in a housing as discussed above. The authenticating system tracks the intensity of the UV light source and readjusts the high/low threshold accordingly. The degradation of the UV light source may be compensated for by periodically feeding a genuine bill into the system, sampling the output of the UV sensor, and adjusting the threshold accordingly. Altematively, degradation may be compensated for by periodically sampling the output of the UV sensor when no bill is present in the rectangular opening 2116 of the transport plate 2114. It is noted that a certain amount of UV light is always reflected off the acrylic window 2118. By periodically sampling the output of the UV sensor when no bill is present, the system can compensate for light source degradation. Furthermore, such sampling can also be used to indicate when the ultraviolet light source has burned out or otherwise requires replacement. This may be accomplished, for example, by means of a display reading or an illuminated light emitting diode ("LED"). The amplified ultraviolet sensor signal 2224 will initially vary between 1.0 and 5.0 volts depending upon the UV reflectance characteristics of the document being scanned and will slowly drift downward as the light source degrades. Altematively, the sampling of the UV sensor output may be used to adjust the gain of the amplifier 2210, thereby maintaining the output of the amplifier 2210 at its initial levels.
It has been found that the voltage ratio between counterfeit and genuine U.S. bills varies from a discemable 2-to-l ratio to a non-discernable ratio. Thus, a 2-to-l ratio is used to discriminate between genuine and counterfeit bills. For example, if a genuine U.S. bill generates an amplified UV output sensor signal 2224 of 4.0 volts, documents generating an amplified UV output sensor signal 2224 of 2.0 volts or less will be rejected as counterfeit. As described above, this threshold of 2.0 volts may either be lowered as the light source degrades or the gain of the amplifier 2210 may be adjusted so that 2.0 volts remains an appropriate threshold value.
The determination of whether the level of UV reflected off a document is high or low is made by sampling the output of the UV sensor at a number of intervals, averaging the readings, and comparing the average level with the predetermined high/low threshold. Altematively, a comparison may be made by measuring the amount of UV light reflected at a number of locations on the bill and comparing these measurements with those obtained from genuine bills. Altematively, the output of one or more UV sensors may be processed to generate one or more patte s of reflected UV light and these patte s may be compared to the pattems generated by genuine bills.
In a similar manner, the presence of fluorescence may be determined by sampling the output of the fluorescence sensor at a number of intervals. However, a bill is rejected as counterfeit U.S. cunency if any of the sampled outputs rise above the noise floor. The altemative methods discussed above with respect to processing the signal or signals of a UV sensor or sensors may also be employed, especially with respect to cunencies of other countries or other types of documents which may employ as security features certain locations or pattems of fluorescent materials. FIGS. 60-63 illustrate a disc-type coin sorter that uses a coin-driving member having a resilient surface for moving coins along a metal coin-guiding surface of a stationary coin-guiding member. The coin-driving member is a rotating disc, and the coin-guiding member is a stationary sorting head. As can be seen in FIG. 60, a hopper 1510 receives coins of mixed denominations and feeds them through central openings in a housing 1511 and a coin-guiding member in the form of an annular sorting head or guide plate 1512 inside or underneath the housing. As the coins pass through these openings, they are deposited on the top surface of a coin-driving member in the form of a rotatable disc 1513. This disc 1513 is mounted for rotation on a stub shaft (not shown) and driven by an electric motor 1514 mounted to a base plate 1515. The disc 1513 comprises a resilient pad 1516 bonded to the top surface of a solid metal disc 1517.
The top surface of the resilient pad 1516 is preferably spaced from the lower surface of the sorting head 1512 by a gap of about 0.005 inches (0J3 mm). The gap is set around the circumference of the sorting head 1512 by a three point mounting anangement including a pair of rear pivots 1518, 1519 loaded by respective torsion springs 1520 which tend to elevate the forward portion of the sorting head. During normal operation, however, the forward portion of the sorting head 1512 is held in position by a latch 1522 which is pivotally mounted to the frame 1515 by a bolt 1523. The latch 1522 engages a pin 1524 secured to the sorting head. For gaining access to the opposing surfaces of the resilient pad 1516 and the sorting head, the latch is pivoted to disengage the pin 1524, and the forward portion of the sorting head is raised to an upward position (not shown) by the torsion springs 1520. As the disc 1513 is rotated, the coins 1525 deposited on the top surface thereof tend to slide outwardly over the surface of the pad due to centrifugal force. The coins 1525, for example, are initially displaced from the center of the disc 1513 by a cone 1526, and therefore are subjected to sufficient centrifugal force to overcome their static friction with the upper surface of the disc. As the coins move outwardly, those coins which are lying flat on the pad enter the gap between the pad surface and the guide plate 1512 because the underside of the inner periphery of this plate is spaced above the pad 16 by a distance which is about the same as the thickness of the thickest coin. As further described below, the coins are sorted into their respective denominations, and the coins for each denomination issue from a respective exit slot, such as the slots 1527, 1528, 1529, 1530, 1531 and 1532 (see FIGS. 60 and 61) for dimes, pennies, nickels, quarters, dollars, and half-dollars, respectively. In general, the coins for any given cunency are sorted by the variation in diameter for the various denominations. Preferably most of the aligning, referencing, sorting, and ejecting operations are performed when the coins are pressed into engagement with the lower surface of the sorting head 1512. In other words, the distance between the lower surfaces of the sorting head 1512 with the passages conveying the coins and the upper surface of the rotating disc 1513 is less than the thickness of the coins being conveyed. As mentioned above, such positive control permits the coin sorter to be quickly stopped by braking the rotation of the disc 1513 when a preselected number of coins of a selected denomination have been ejected from the sorter. Positive control also permits the sorter to be relatively compact yet operate at high speed. The positive control, for example, permits the single file stream of coins to be relatively dense, and ensures that each coin in this stream can be directed to a respective exit slot. Tuming now to FIG. 61, there is shown a bottom view of the prefened sorting head 1512 including various channels and other means especially designed for high-speed sorting with positive control of the coins, yet avoiding the galling problem. It should be kept in mind that the circulation of the coins, which is clockwise in FIG. 60, appears counterclockwise in FIG. 61 because FIG. 61 is a bottom view. The various means operating upon the circulating coins include an entrance region 1540, means 1541 for stripping "shingled" coins, means 1542 for selecting thick coins, first means 1544 for recirculating coins, first referencing means 1545 including means 1546 for recirculating coins, second referencing means 1547, and the exit means 1527, 1528, 1529, 1530, 1531 and 1532 for six different coin denominations, such as dimes, pennies, nickels, quarters, dollars and half-dollars. The lowermost surface of the sorting head 1512 is indicated by the reference numeral 1550.
Considering first the entrance region 1540. the outwardly moving coins initially enter under a semi-annular region underneath a planar surface 1561 formed in the underside of the guide plate or sorting head 1512. Coin Cl , superimposed on the bottom plan view of the guide plate in FIG. 61 is an example of a coin which has entered the entrance region 1540. Free radial movement of the coins within the entrance region 1540 is terminated when they engage a wall 1562, though the coins continue to move circumferentially along the wall 1562 by the rotational movement of the pad 1516, as indicated by the central aπow in the counterclockwise direction in FIG. 61. To prevent the entrance region 1540 from becoming blocked by shingled coins, the planar region 1561 is provided with an inclined surface 1541 forming a wall or step 1563 for engaging the upper most coin in a shingled pair. In FIG. 61, for example, an upper coin C2 is shingled over a lower coin C3. As further shown in FIG. 62, movement of the upper coin C2 is limited by the wall 1563 so that the upper coin C2 is forced off of the lower coin C3 as the lower coin is moved by the rotating disc 1513.
Retuming to FIG. 61 , the circulating coins in the entrance region 1540, such as the coin Cl, are next directed to the means 1542 for selecting thick coins. This means 1542 includes a surface 1564 recessed into the sorting head 1512 at a depth of 0.070 inches (1.78 mm) from the lowermost surface 1550 of the sorting head.
Therefore, a step or wall 1565 is formed between the surface 1561 of the entrance region 1540 and the surface 1564. The distance between the surface 1564 and the upper surface of the disc 1513 is therefore about 0.075 inches so that relatively thick coins between the surface 1564 and the disc 1513 are held by pad pressure. To initially engage such thick coins, an initial portion of the surface 1564 is formed with a ramp 1566 located adjacent to the wall 1562. Therefore, as the disc 1513 rotates, thick coins in the entrance region that are next to the wall 1562 are engaged by the ramp 1566 and thereafter their radial position is fixed by pressure between the disc and the surface 1564. Thick coins which fail to initially engage the ramp 1566, however, engage the wall 1565 and are therefore recirculated back within the central region of the sorting head. This is illustrated, for example, in ?IG. 63 for the coin C4. This initial selecting and positioning of the thick coins prevents misaligned thick coins from hindering the flow of coins to the first referencing means 1545.
Retuming now to FIG. 61 , the ramp 1566 in the means 1542 for selecting the thick coins can also engage a pair or stack of thin coins. Such a stack or pair of thin coins will be carried under pad pressure between the surface 1564 and the rotating disc 1513. In the same manner as a thick coin, such a pair of stacked coins will have its radial position fixed and will be carried toward the first referencing means 1545. The first means 1545 for referencing the coins obtains a single-file stream of coins directed against the outer wall 1562 and leading up to a ramp 1573.
Coins are introduced into the referencing means 1545 by the thinner coins moving radially outward via centrifugal force, or by the thicker coin(s) C52a following concentricity via pad pressure. The stacked coins C58a and C50a are separated at the inner wall 1582 such that the lower coin C58a is canied against surface 1572a. The progression of the lower coin C58a is depicted by its positions at C58b, C58c, C58d, and C58e. More specifically, the lower coin C58 becomes engaged between the rotating disc 1513 and the surface 1572 in order to carry the lower coin to the first recirculating means 1544, where it is recirculated by the wall 1575 at positions C58d and C58e. At the beginning of the wall 1582, a ramp 1590 is used to recycle coins not fully between the outer and inner walls 1562 and 1582 and under the sorting head 1512. As shown in FIG. 61, no other means is needed to provide a proper introduction of the coins into the referencing means 1545. The referencing means 1545 is further recessed over a region 1591 of sufficient length to allow the coins C54 of the widest denomination to move to the outer wall 1562 by centrifugal force. This allows coins C54 of the widest denomination to move freely into the referencing means 1545 toward its outer wall 1562 without being pressed between the resilient pad 1516 and the sorting head 1512 at the ramp 1590. The inner wall 1582 is preferably constructed to follow the contour of the recess ceiling. The region 1591 of the referencing recess 1545 is raised into the head 1512 by ramps 1593 and 1594, and the consistent contour at the inner wall 1582 is provided by a ramp 1595.
The first referencing means 1545 is sufficiently deep to allow coins C50 having a lesser thickness to be guided along the outer wall 1562 by centrifugal force, but sufficiently shallow to permit coins C52, C54 having a greater thickness to be pressed between the pad 1516 and the sorting head 1512, so that they are guided along the inner wall 1582 as they move through the referencing means 1545. The referencing recess 1545 includes a section 1596 which bends such that coins C52, which are sufficiently thick to be guided by the inner wall 1582 but have a width which is less than the width of the referencing recess 1545, are carried away from the inner wall 1582 from a maximum radial location 1583 on the inner wall toward the ramp 1573.
This configuration in the sorting head 1512 allows the coins of all denominations to converge at a nanow ramped finger 1573a on the ramp 1573, with coins C54 having the largest width being carried between the inner and outer walls via the surface 1596 to the ramped finger 1573a so as to bring the outer edges of all coins to a generally common radial location. By directing the coins C50 radially inward along the latter portion of the outer wall 1562, the probability of coins being offset from the outer wall 1562 by adjacent coins and being led onto the ramped finger 1573a is significantly reduced. Any coins C50 which are slightiy offset from the outer wall 1562 while being led onto the ramp finger 1573a may be accommodated by moving the edge 1551 of exit slot 1527 radially inward, enough to increase the width of the slot 1527 to capture offset coins C50 but to prevent the capture of coins of the larger denominations. For sorting Dutch coins, the width of the ramp finger 1573a may be about 0J40 inch. At the terminal end of the ramp
1573, the coins become firmly pressed into the pad 16 and are carried forward to the second referencing means 1547. A coin such as the coin C50c will be carried forward to the seconci referencing means 1547 so long as a portion of the coin is engaged by the nanow ramped finger 1573a on the ramp 1573. If a coin is not sufficiently close to the wall 1562 so as to be engaged by this ramped finger 1573a, then the coin strikes a wall 1574 defined by the second recirculating means 1546, and that coin is recirculated back to the entrance region 1540.
The first recirculating means 1544, the second recirculating means 1546 and the second referencing means 1547 are defined at successive positions in the sorting head 1512. It should be apparent that the first recirculating means 1544, as well as the second recirculating means 1546, recirculate the coins under positive control of pad pressure. The second referencing means 1547 also uses positive control of the coins to align the outer most edge of the coins with a gaging wall 1577. For this purpose, the second referencing means 1547 includes a surface 1576, for example, at 0J 10 inches (1.27 mm) from the bottom surface of the sorting head 1512, and a ramp 1578 which engages the inner edge portions of the coins, such as the coin C50d.
As best shown in FIG. 61, the initial portion of the gaging wall 1577 is along a spiral path with respect to the center of the sorting head 1512 and the sorting disc 1513, so that as the coins are positively driven in the circumferential direction by the rotating disc 1513, the outer edges of the coins engage the gaging wall 1577 and are forced slightly radially inward to a precise gaging radius, as shown for the coin C16 in FIG. 62. FIG. 62 further shows a coin C17 having been ejected from the second recirculating means 1546.
Referring back to FIG. 61, the second referencing means 1547 terminates with a slight ramp 1580 causing the coins to be firmly pressed into the pad 1516 on the rotating disc with their outer most edges aligned with the gaging radius provided by the gaging wall 1577. At the terminal end of the ramp 1580 the coins are gripped between the guide plate 1512 and the resilient pad 1516 with the maximum compressive force. This ensures that the coins are held securely in the new radial position determined by the wall 1577 of the second referencing means 1547.
The sorting head 1512 further includes sorting means comprising a series of ejection recesses 1527, 1528, 1529, 1530, 1531 and 1532 spaced circumferentially around the outer periphery of the plate, with the innermost edges of successive slots located progressively farther away from the common radial location of the outer edges o: all the coins for receiving and ejecting coins in order of increasing diameter. The width of each ejection recess is slightly larger than the diameter of the coin to be received and ejected by that particular recess, and the surface of the guide plate adjacent the radially outer edge of each ejection recess presses the outer portions of the coins received by that recess into the resilient pad so that the inner edges of those coins are tilted upwardly into the recess. The ejection recesses extend outwardly to the periphery of the guide plate so that the inner edges of these recesses guide the tilted coins outwardly and eventually eject those coins from between the guide plate 1512 and the resilient pad 1516.
The innermost edges of the ejection recesses are positioned so that the inner edge of a coin of only one particular denomination can enter each recess; the coins of all other remaining denominations extend inwardly beyond the innermost edge of that particular recess so that the inner edges of those coins cannot enter the recess. For example, the first ejection recess 1527 is intended to discharge only dimes, and thus the innermost edge 1551 of this recess is located at a radius that is spaced inwardly from the radius of the gaging wall 1577 by a distance that is only slightly greater than the diameter of a dime. Consequently, only dimes can enter the recess 1527. Because the outer edges of all denominations of coins are located at the same radial position when they leave the second referencing means 1547, the inner edges of the pennies, nickels, quarters, dollars and half dollars all extend inwardly beyond the innermost edge of the recess 1527, thereby preventing these coins from entering that particular recess. At recess 1528, the inner edges of only pennies are located close enough to the periphery of the sorting head 1512 to enter the recess. The inner edges of all the larger coins extend inwardly beyond the innermost edge 1552 of the recess 1528 so that they remain gripped between the guide plate and the resilient pad. Consequently, all the coins except the pennies continue to be rotated past the recess 1528. Similarly, only nickels enter the ejection recess 1529, only the quarters enter the recess 1530, only the dollars enter the recess 1531, and only the half dollars enter the recess 1532.
Because each coin is gripped between the sorting head 1512 and the resilient pad 16 throughout its movement through the ejection recess, the coins are under positive control at all times. Thus, any coin can be stopped at any point along the length of its ejection recess, even when the coin is already partially projecting beyond the outer periphery of the guide plate. Consequently, no matter when the rotating disc is stopped (e.g., in response to the counting of a preselected number of coins of a particular denomination), those coins which are already within the various ejection recesses can be retained within the sorting head until the disc is re-started for the next counting operation.
One of six proximity sensors S S6 is mounted along the outboard edge of each of the six exit channels 1527-1532 in the sorting head for sensing and counting coins passing through the respective exit channels. By locating the sensors S,-S6 in the exit channels, each sensor is dedicated to one particular denomination of coin, and thus it is not necessary to process the sensor output signals to determine the coin denomination. The effective fields of the sensors S,-S6 are all located just outboard of the radius at which the outer edges of all coin denominations are gaged before they reach the exit channels 1527-1532, so that each sensor detects only the coins which enter its exit channel and does not detect the coins which bypass that exit channel. Only the largest coin denomination (e.g., U.S. half dollars) reaches the sixth exit channel 1532, and thus the location of the sensor in this exit channel is not as critical as in the other exit channels 1527-1531. In addition to the proximity sensors S1-S6, each of the exit channels 1527-1532 also includes one of six coin discrimination sensors D1-D6. These sensors D1-D6 are the eddy cunent sensors, and will be described in more detail below in connection with FIGS. 64-67 of the drawings.
When one of the discrimination sensors detects a coin material that is not the proper material for coins in that exit channel, the disc may be stopped by de-energizing or disengaging the drive motor and energizing a brake. The suspect coin may then be discharged by jogging the drive motor with one or more electrical pulses until the trailing edge of the suspect coin clears the exit edge of its exit channel. The exact disc movement required to move the trailing edge of a coin from its sensor to the exit edge of its exit channel, can be empirically determined for each coin denomination and then stored in the memory of the control system. An encoder on the sorter disc can then be used to measure the actual disc movement following the sensmg of the suspect coin, so that the disc can be stopped at the precise position where the suspect coin clears the exit edge of its exit channel, thereby ensuring that no coins following the suspect coin are discharged.
Tuming now to FIGS. 64-67, one embodiment of the present invention employs an eddy cunent sensor 1710 to perform as the coin handling system's coin discrimination sensors D1-D6. The eddy cunent sensor 1710 includes an excitation coil 1712 for generating an alternating magnetic field used to induce eddy cunents in a coin 1714. The excitation coil 1712 has a start end 1716 and a finish end 1718. An embodiment an a-c. excitation coil voltage Veχ, e.g. , a sinusoidal signal of 250 KHz and 10 volts peak-to-peak, is applied across the start end 1716 and the finish end 1718 of the excitation coil 1712. The alternating voltage V£χ produces a conesponding cunent in the excitation coil 1712 which in turn produces a conesponding alternating magnetic field. The alternating magnetic field exists within and around the excitation coil 1712 and extends outwardly to the coin 1714. The magnetic field penetrates the coin 1714 as the coin is moving in close proximity to the excitation coil 1712, and eddy cunents are induced in the coin 1714 as the coin moves through the alternating magnetic field. The strength of the eddy cunents flowing in the coin 1714 is dependent on the material composition of the coin, and particularly the electrical resistance of that material. Resistance affects how much cunent will flow in the coin 1614 according to Ohm's Law (voltage = cunent * resistance).
The eddy cunents themselves also produce a conesponding magnetic field. A proximal detector coil 1722 and a distal coil 1724 are disposed above the coin 1714 so that the eddy cunent-generated magnetic field induces voltages upon the coils 1722, 1724. The distal detector coil 1724 is positioned above the coin 1714, and the proximal detector coil 1722 is positioned between the distal detector coil 1724 and the passing coin 1714. In one embodiment, the excitation coil 1712, the proximal detector coil 1722 and the distal detector coil 1724 are all wound in the same direction (either clockwise or counterclockwise). The proximal detection coil 1722 and the distal detector coil 1724 are wound in the same direction so that the voltages induced on these coils by the eddy cunents are properly oriented.
The proximal detection coil 1722 has a starting end 1726 and a finish end 1728. Similarly, the distal coil 1724 has a starting end 1730 and a finish end 1632. In order of increasing distance from the coin 1614, the detector coils 1722, 1724 are positioned as follows: finish end 1728 of the proximal detector coil 1722, start end 1726 of the proximal detector coil 1722, finish end 1732 of the distal detector coil 1724 and start end 1730 of the distal detector coil 1724. The finish end 1728 of the proximal detection coil 1722 is connected to the finish end 1732 of the distal detector coil 1724 via a conductive wire 1734. It will be appreciated by those skilled in the art that other detector coil 1722, 1724 combinations are possible. For example, in an altemative embodiment the proximal detection coil 1722 is wound in the opposite direction of the distal detection coil 1724. In this case the start end 1726 of the proximal coil 1722 is connected to the finish end 1732 of the distal coil 1724.
Eddy cunents in fhe coin 1714 induce voltages V_roχ and V^j. respectively on the detector coils 1722, 1724. Likewise, the excitation coil 1712 also induces a common-mode voltage Vcom on each of the detector coils 1722, 1724. The common-mode voltage Vcom is effectively the same on each detector coil due to the symmetry of the detector coils' physical aπangement within the excitation coil 1712. Because the detector coils 1722, 1724 are wound and physically oriented in the same direction and connected at their finish ends 1728, 1732, the common-mode voltage Vcom induced by the excitation coil 1712 is subtracted out, leaving only a difference voltage Vjjjf coπesponding to the eddy cunents in the coin 1714. This eliminates the need for additional circuitry to subtract out the common-mode voltage Vcom. The common-mode voltage Vcom is effectively subtracted out because both the distal detection coil 1724 and the proximal detection coil 1722 receive the same level of induced voltage Vcom from the excitation coil 1712.
Unlike the common-mode voltage, the voltages induced by the eddy cunent in the detector coils are not effectively the same. This is because the proximal detector coil 1722 is purposely positioned closer to the passing coin than the distal detector coil 1724. Thus, the voltage induced in the proximal detector coil 1722 is significantly stronger, i.e. has greater amplitude, than the voltage induced in the distal detector coil 1724. Although the present invention subtracts the eddy cunent- induced voltage on the distal coil 1724 from the eddy cunent-induced voltage on the proximal coil 1722, the voltage amplitude difference is sufficientiy great to permit detailed resolution of the eddy cunent response.
As seen in FIG. 64, the excitation coil 1712 is radially sunounded by a magnetic shield 1734. The magnet shield 1734 has a high level of magnetic permeability in order to help contain the magnetic field sunounding the excitation coil 1712. The magnetic shield 1734 has the advantage of preventing stray magnetic field from interfering with other nearby eddy cunent sensors. The magnetic shield is itself radially sunounded by a steel outer case 1736.
In one embodiment the excitation coil utilizes a cylindrical ceramic (e.g. , alumina) core 1738. Alumina has the advantages of being impervious to humidity and providing a good wear surface. It is desirable that the core 1748 be able to withstand wear because it may come into frictional contact with the coin 1714. Alumina withstands frictional contact well because of its high degree of hardness, i.e. , approximately 9 on mohs scale. To form the eddy cunent sensor 1510, the detection coils 1722, 1724 are wound on a coil form (not shown). A prefened form is a cylinder having a length of 0.5 inch, a maximum diameter of 0.2620 inch, a minimum diameter of 0J660 inch, and two grooves of 0.060 inch width spaced apart by 0.060 inch and spaced from one end of the form by 0.03 inch. Both the proximal detection coil 1722 and the distal detector coil 1724 have 350 turns of #44 AWG enamel covered magnet wire layer wound to generally uniformly fill the available space in the grooves. Each of the detector coils 1722, 1724 are wound in the same direction with the finish ends 1728, 1732 being connected together by the conductive wire 1734. The start ends 1726, 1730 of the detector coils 1722, 1724 are connected to separately identified wires in a connecting cable.
The excitation coil 1712 is a generally uniformly layer wound on a cylindrical alumina ceramic coil form having a length of 0.5 inch, an outside diameter of 0.2750 inch, and a wall thickness of 0.03125 inch. The excitation coil 1712 is wound with 135 turns of #42 AWG enamel covered magnet wire in the same direction as the detector coils 1722, 1724. The excitation coil voltage Vgχ is applied across the start end 1716 and the finish end 1718. After the excitation coil 1712 and detector coils 1722, 1724 are wound, the excitation coil 1712 is slipped over the detector coils 1722, 1724 around a common center axis. At this time the sensor 1710 is connected to a test oscillator (not shown) which applies the excitation voltage V£χ to the excitation coil 1712. The excitation coil's position is adjusted along the axis of the coil to give a null response from the detector coils 1722, 1724 on an a-c. voltmeter with no metal near the coil windings. Then the magnetic shield 1644 is the slipped over the excitation coil 1712 and adjusted to again give a null response from the detector coils 1722, 1724.
The magnetic shield 1744 and coils 1712, 1722, 1724 within the magnetic shield 1744 are then placed in the steel outer case 1746 and encapsulated with a polymer resin (not shown) to "freeze" the position of the magnetic shield 1744 and coils 1712, 1722, 1724.
After curing the resin, an end of the eddy cunent sensor 1710 nearest the proximal detector coil 1722 is sanded and lapped to produce a flat and smooth surface with the coils 1712, 1722 slightly recessed within the resin. In order to detect the effect of the coin 1714 on the voltages induced upon the detector coils 1722, 1724, it is prefened to use a combination of phase and amplitude analysis of the detected voltage. This type of analysis minimizes the effects of variations in coin surface geometry and in the distance between the coin and the coils. The voltage applied to the excitation coil 1712 causes cuπent to flow in the coil 1712 which lags behind the voltage 1720. For example, the cunent may lag the voltage 1720 by 90 degrees in a superconductive coil. In effect, the coin's 1714 eddy cunents impose a resistive loss on the cunent in the excitation coil 1712. Therefore, the initial phase difference between the voltage and cunent in the excitation coil 1712 is decreased by the presence of the coin 1714. Thus, when the detector coils 1724, 1726 have a voltage induced upon them, the phase difference between the voltage applied to the excitation coil 1712 and that of the detector coils is reduced due to the jm eddy cunent effect in the coin. The amount of reduction in the phase difference is proportional to the electrical and magnetic characteristics of the coin and thus the composition of the coin. By analyzing both the phase difference and the maximum amplitude, an accurate assessment of the composition of the coin is achieved. FIGS. 67A and 67B illustrate a prefened phase-sensitive detector 1750 for sampling the differential output signal V^^ from the two detector coils 1722, 1724. The differential output signal
Figure imgf000124_0001
is passed through a buffer amplifier 252 to a switch 1754, where the buffered V^j^ is sampled once per cycle by momentarily closing the switch 1754. The switch 1754 is controlled by a series of reference pulses produced from the V£χ signal, one pulse per cycle. The reference pulses 1758 are synchronized with excitation voltage Veχ. so that the amplitude of the differential output signal
Figure imgf000124_0002
during the sampling interval is a function not only of the amplitude of the detector coil voltages 1736, 1738, but also of the phase difference between the signals in excitation coil 1712 and the detection coils 1736, 1738. The pulses derived from V£χ are delayed by an "offset angX' which can be adjusted to minimize the sensitivity of V^^ to variations in the gap between the proximal face of the sensor 1710 and the surface of the coin 1714 being sensed. The value of the offset angle for any given coin can be determined empirically by moving a standard metal disc, made of the same material as the coin 1714, from a position where it contacts the sensor face, to a position where it is spaced about 0.001 to 0.020 inch from the sensor face. The signal sample from the detector 1750 is measured at both positions, and the difference between the two measurements is noted. This process is repeated at several different offset angles to determine the offset angle which produces the minimum difference between the two measurements. Each time buffered
Figure imgf000124_0003
is sampled, the resulting sample is passed through a second buffer amplifier 1756 to an analog-to-digital converter (not shown). The resulting digital value is supplied to a microprocessor (not shown) which compares that value with several different ranges of values stored in a lookup table (not shown). Each stored range of values conesponds to a particular coin material, and thus the coin material represented by any given sample value is determined by the particular stored range into which the sample value falls. The stored ranges of values can be determined empirically by simply measuring a batch of coins of each denomination and storing the resulting range of values measured for each denomination.
If desired, the coin sorting and counting module 8 may be replaced with a coin discriminating module which does not sort the coins. Such a module would align the coins of all denominations in a single file and guide them past a single coin discrimination sensor to deteπnine whether the coins are genuine. The coins of all denominations would then be discharged into a single storage receptacle and sorted at a later time. Coins that are detected to be non-genuine would be diverted and retumed to the customer at the coin retum station 4.
When an invalid coin is detected by one of the discriminating sensors described above, the invalid coin is separated from the valid coins and retumed to the customer. In the illustrative module 8, this separation is effected outside the sorting disc by the shunting device illustrated in FIGS. 68-71. The curved exit chute 1800 includes two slots 1802, 1804 separated by an intemal partition 1806. The intemal partition 1806 is pivotally mounted to a stationary base 1808 so that the intemal partition 1806 may be moved, perpendicular to the plane of the coins, by an actuator 1810 between an up position (FIG. 70) and a down position (FIG. 69). The exit chute 1800 is positioned adjacent an exit channel of the coin sorter such that coins exiting the coin sorter are guided into the slot 1802 when the intemal partition 1806 is in the down position (FIG. 69). When an invalid coin is detected by the discriminating sensor D, the actuator 1810 moves the intemal partition 1806 to the up position (FIG. 66) so that the invalid coin now enters the slot 1804 of the exit chute 1800. Coins entering the slot 1804 are discharged into the tube 9 that conveys those coins to the coin-return slot 4 at the front of the ATM . While FIGS. 67-70 illustrate only a single exit chute, it will be apparent that a similar exit chute is provided at each of the six coin exit locations around the circumference of the sorting disc.
The actuator 1810 moves the intemal partition 1806 between the up and down positions in response to detection of invalid and valid coins. Thus, if the intemal partition 1806 is in the down position and an invalid coin is detected, the partition 1806 is moved to the up position so that the invalid coin will be diverted into the slot 1804. Altematively, an invalid coin may be separated from the valid coins by use of inboard actuators in the sorting head, activated by signals derived from one or more sensors mounted in the sorting head upstream of the actuators. Such an anangement is described in U.S. Patent No. 5,299,977, which is incorporated herein by reference.

Claims

CLAIMS: l. A cunency processing station for receiving and dispensing cash and substantially immediately furnishing an associated accounting system with data, including the value of the cunency processed, for each transaction, said station comprising a bill dispenser including a bill storage device and controllable transport means for dispensing selected numbers of bills from said storage device, a bill receptacle for receiving stacks of bills to be deposited, a bill counter and scanner for rapidly removing the bills one at a time from said receptacle and counting the bills while determining the denomination of each bill, said counter and scanner including means for generating data representing the denomination of each bill, and the number of bills of each denomination, passed through said counter and scanner, a memory for receiving and storing data representing the number of bills of each denomination passed through said counter and scanner in each transaction, and data representing the total value of the bills passed through said counter and scanner in each transaction, and control means for transferring data from said memory to an associated cash accounting system so that the deposits and withdrawals executed at said cunency processing station are entered in said accounting system substantially immediately after the execution of those transactions.
2. The cunency processing station of claim 1 wherein said counter and scanner and said transport means process bills at a rate in excess of about 350 bills per minute.
3. The cunency processing station of claim 1 wherein said counter and scanner includes an optical scanhead which illuminates each bill and detects light reflected from the bill and produces coπesponding electrical signals, and signal processing means for receiving said electrical signals and determining the denomination of the bill from which the light was reflected.
4. The cunency processing station of claim 3 which includes at least two optical scanheads, located on opposite sides of the bills being scanned.
5. The cunency processing station of claim 1 which includes a coin sorter for receiving successive batches of coins of mixed denominations, sorting the coins by denomination, and generating electrical signals representing the number of coins of each denomination in each batch of coins that is processed.
6. The cunency processing station of claim 1 which includes authenticating means for checking the genuineness of each bill and coin that is counted, and producing a control signal in response to the detection of a non-genuine bill or coin.
7. The cunency processing station of claim 6 which includes means responsive to said control signal for altering the processing of the bill or coin detected to be non-genuine.
8. The cunency processing station of claim 3 which includes a memory for storing master characteristic pattems of a plurality of denominations of genuine bills, and said signal processing means includes means for comparing the signals from said scanhead with the stored pattems in said memory to determine the denomination of each scanned bill.
PCT/US1996/006257 1995-05-02 1996-05-02 Automatic currency processing system WO1996036933A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP96915489A EP0830653A4 (en) 1995-05-02 1996-05-02 Automatic currency processing system
AU57251/96A AU5725196A (en) 1995-05-02 1996-05-02 Automatic currency processing system

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US43392095A 1995-05-02 1995-05-02
US08/433,920 1995-05-02

Publications (2)

Publication Number Publication Date
WO1996036933A1 WO1996036933A1 (en) 1996-11-21
WO1996036933A9 true WO1996036933A9 (en) 1997-02-20

Family

ID=23722079

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US1996/006257 WO1996036933A1 (en) 1995-05-02 1996-05-02 Automatic currency processing system

Country Status (3)

Country Link
EP (1) EP0830653A4 (en)
AU (1) AU5725196A (en)
WO (1) WO1996036933A1 (en)

Families Citing this family (33)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5982918A (en) * 1995-05-02 1999-11-09 Cummins-Allison, Corp. Automatic funds processing system
US6363164B1 (en) 1996-05-13 2002-03-26 Cummins-Allison Corp. Automated document processing system using full image scanning
US8950566B2 (en) 1996-05-13 2015-02-10 Cummins Allison Corp. Apparatus, system and method for coin exchange
US7187795B2 (en) 2001-09-27 2007-03-06 Cummins-Allison Corp. Document processing system using full image scanning
US8162125B1 (en) 1996-05-29 2012-04-24 Cummins-Allison Corp. Apparatus and system for imaging currency bills and financial documents and method for using the same
US7753189B2 (en) 2003-08-01 2010-07-13 Cummins-Allison Corp. Currency processing device, method and system
US20050276458A1 (en) 2004-05-25 2005-12-15 Cummins-Allison Corp. Automated document processing system and method using image scanning
EP1480177B1 (en) * 1997-04-14 2007-11-07 Cummins-Allison Corporation Method of updating accounts
US8701857B2 (en) 2000-02-11 2014-04-22 Cummins-Allison Corp. System and method for processing currency bills and tickets
JP2001256534A (en) * 2000-03-10 2001-09-21 Fujitsu Ltd Proceeds reception device for shop
FR2824410B1 (en) 2001-05-03 2003-08-08 Michel Pernot FACIAL VALUE INDICATOR FOR BANK NOTES
US7647275B2 (en) 2001-07-05 2010-01-12 Cummins-Allison Corp. Automated payment system and method
US8944234B1 (en) 2001-09-27 2015-02-03 Cummins-Allison Corp. Apparatus and system for imaging currency bills and financial documents and method for using the same
US8437529B1 (en) 2001-09-27 2013-05-07 Cummins-Allison Corp. Apparatus and system for imaging currency bills and financial documents and method for using the same
US8437530B1 (en) 2001-09-27 2013-05-07 Cummins-Allison Corp. Apparatus and system for imaging currency bills and financial documents and method for using the same
US8428332B1 (en) 2001-09-27 2013-04-23 Cummins-Allison Corp. Apparatus and system for imaging currency bills and financial documents and method for using the same
US7066335B2 (en) 2001-12-19 2006-06-27 Pretech As Apparatus for receiving and distributing cash
US6896118B2 (en) 2002-01-10 2005-05-24 Cummins-Allison Corp. Coin redemption system
US8627939B1 (en) 2002-09-25 2014-01-14 Cummins-Allison Corp. Apparatus and system for imaging currency bills and financial documents and method for using the same
WO2007044570A2 (en) 2005-10-05 2007-04-19 Cummins-Allison Corp. Currency processing system with fitness detection
US7762380B2 (en) 2006-03-09 2010-07-27 Cummins-Allison Corp. Currency discrimination system and method
WO2007143128A2 (en) 2006-06-01 2007-12-13 Cummins-Allison Corp. Angled currency processing system
US8417017B1 (en) 2007-03-09 2013-04-09 Cummins-Allison Corp. Apparatus and system for imaging currency bills and financial documents and method for using the same
GB2486832A (en) 2007-03-09 2012-06-27 Cummins Allison Corp Document processing system using blind balancing
US8401268B1 (en) 2007-03-09 2013-03-19 Cummins-Allison Corp. Optical imaging sensor for a document processing device
US8929640B1 (en) 2009-04-15 2015-01-06 Cummins-Allison Corp. Apparatus and system for imaging currency bills and financial documents and method for using the same
US8437528B1 (en) 2009-04-15 2013-05-07 Cummins-Allison Corp. Apparatus and system for imaging currency bills and financial documents and method for using the same
US8391583B1 (en) 2009-04-15 2013-03-05 Cummins-Allison Corp. Apparatus and system for imaging currency bills and financial documents and method for using the same
CA2786231C (en) 2010-03-03 2016-05-03 Cummins-Allison Corp. Currency bill processing device and method
US9141876B1 (en) 2013-02-22 2015-09-22 Cummins-Allison Corp. Apparatus and system for processing currency bills and financial documents and method for using the same
US8973817B1 (en) 2013-03-15 2015-03-10 Cummins-Allison Corp. Apparatus, method, and system for loading currency bills into a currency processing device
US11734983B1 (en) 2018-12-18 2023-08-22 Cummins-Allison Corp. Banknote transport mechanisms and methods
US11417178B1 (en) 2021-06-16 2022-08-16 Wells Fargo Bank, N.A. No-downtime universal utility cassette for recirculation in multiple systems

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS5674790A (en) * 1979-11-22 1981-06-20 Tokyo Shibaura Electric Co Bill discriminator
JPS5896389A (en) * 1981-12-02 1983-06-08 グローリー工業株式会社 Currency settling method
GB2135494B (en) * 1983-01-26 1986-05-21 Tokyo Shibaura Electric Co Automatic bank note transaction apparatus
JPS63177289A (en) * 1987-01-19 1988-07-21 株式会社東芝 Sheet paper processor
US5295196A (en) * 1990-02-05 1994-03-15 Cummins-Allison Corp. Method and apparatus for currency discrimination and counting
US5141443A (en) * 1990-05-14 1992-08-25 Cummins-Allison Corp. Coin sorter with automatic bag-switching or stopping

Similar Documents

Publication Publication Date Title
US6128402A (en) Automatic currency processing system
US5905810A (en) Automatic currency processing system
US8023715B2 (en) Automatic currency processing system having ticket redemption module
US5982918A (en) Automatic funds processing system
US6068194A (en) Software loading system for an automatic funds processing system
WO1996036933A9 (en) Automatic currency processing system
EP0805408B1 (en) Method and apparatus for discriminating and counting documents
EP0830653A1 (en) Automatic currency processing system
US5912982A (en) Method and apparatus for discriminating and counting documents
CA2215886C (en) Method and apparatus for currency discrimination
CA2357502C (en) Currency identification device and method