JPS6411995B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for determining the type of securities paper sheets, such as banknotes, checks, bills, etc., printed in multiple colors.

Hereinafter, the present invention will be described using banknote type determination (the type of banknote is also referred to as denomination) as an example, but the present invention is not limited thereto.

[Conventional technology]

A conventional banknote type determination method will be described below.

Generally, this type of determination requires a means (referred to as a detection means or a sensor) for converting the shading amount of a banknote into an electrical quantity (also referred to as shading amount detection); A means for relatively moving (also referred to as conveyance) the detection means and the banknote for the purpose of detecting the amount of shading (referred to as conveyance means).
and a means (referred to as processing means) for processing the amount of electricity corresponding to the density of the banknote thus obtained and determining the type of banknote.

The detection means includes, for example, an illuminator and a photoelectric converter. For example, an LED is used as the illuminator, and a phototransistor, for example, is used as the photoelectric converter. For example, when detecting the amount of light transmitted through a banknote as the amount of darkness of the banknote, an illuminator and a photoelectric converter are installed so that they each sandwich the banknote.

If this detection means is installed on the path (referred to as a conveyance path) on the conveyance means along which banknotes are conveyed, and a conveyance means such as rollers that conveys the banknotes in one direction is used, the banknotes will be detected by the detector. The amount of electricity corresponding to the amount of shading for the portion passing through is obtained as a function of position.

The amount of electricity corresponding to the amount of darkness of the banknote to be judged obtained in this way (or the amount obtained by adding some processing to this as necessary; the same applies hereinafter) and each type of electricity obtained in the same way and set in advance. The amount of electricity corresponding to the amount of light and shade of a standard banknote is compared, and the type of standard banknote for which the difference is smaller than a predetermined value is determined to be the type of banknote to be determined.

As a conventional technique in line with the above-mentioned general method, an example of a method for determining the quality of a banknote by determining the amount of electricity corresponding to the amount of darkness of the banknote (see Japanese Patent Laid-Open No. 50-107998, Apparatus for determining the quality of paper sheet). ) is shown below.

This method involves arranging multiple photoelectric conversion elements perpendicular to the conveyance direction to detect the amount of darkness in the paper sheet, and integrating the amplified voltages of the photoelectric conversion elements located symmetrically with respect to the center line of the conveyance path by averaging them. By comparing the integrated value with a standard value, it is determined whether the banknote to be determined is torn or soiled.

[Problem to be solved by the invention]

This conventional technique has the following problems.

In the above-mentioned prior art, the means for detecting the amount of darkness of the banknote is fixedly arranged at a position symmetrical to the center line of the conveyance path. Therefore, when a banknote is conveyed with a deviation in the direction orthogonal to the conveyance direction, that is, when the center point of the banknote is conveyed with a deviation from the center line of the conveyance path, there is no means for detecting the amount of darkness of the banknote. The area on the banknote detected by the system is no longer symmetrical with respect to the center point of the banknote, and as a result, a shifted area is detected.

Further, the area detected by the detection means differs on the banknote depending on whether the banknote is reversed in the direction perpendicular to the conveyance direction or not.

In the above, a case has been described in which the transport means is capable of transporting the largest banknote (in the direction perpendicular to the transport direction) among the banknotes handled, but one end (in the direction perpendicular to the transport direction) of the banknote to be transported is Even in the case of a conveyance method (generally referred to as one-sided conveyance) in which the banknotes are regulated (in the direction perpendicular to the conveyance direction), if the dimensions of the banknotes (in the direction perpendicular to the conveyance direction) are different, there will be The center point of the banknote will shift, resulting in the same situation as above.

If the type of banknote is known in advance, it is possible to avoid this problem by, for example, adjusting the width of the conveyance path to match the size of the banknote type each time, but if multiple types of banknotes are mixed together, In the case of so-called banknote type determination, in which the type of banknote is determined from information obtained by a means for detecting the density of the banknote, this becomes an unavoidable problem and a serious drawback.

Apart from the above, when determining the type of banknote, it is effective to detect and judge only areas that clearly represent the characteristics related to the banknote type, such as watermarks, portraits, and characters. In the above-mentioned conventional technology, which uses the integral value of the section from one end to the other end in the conveyance direction for determination, it is not possible to specifically extract and detect such a remarkable characteristic part.

[Means to solve the problem]

The method of the present invention is aimed at determining the type of paper leaf, and is intended to solve the problems of the prior art described above.
The main points are as follows.

a first detection means for detecting the presence or absence of a paper sheet and the amount of movement; and a second detection means for obtaining an electric signal according to the density of the paper sheet along the passing direction of the paper sheet. each placed at a predetermined position on the conveyance path of the paper sheet, and determining the center of a line segment on the paper sheet when the paper sheet passes through the second detection means based on the output of the first detection means; At least one pair of partial sections are set at symmetrical positions on the line segment and at a predetermined distance from the center point along the passing direction of the paper sheet, and for each of the set partial sections. Find the integral value of the output of the second detection means, perform a calculation to find the absolute value of the sum or difference for each pair, and check whether the absolute value of the sum or difference found by the calculation is within the set range. The type of the paper sheet is determined by checking whether or not the paper sheet is present.

In addition, some supplementary explanations regarding the above will be provided below.

(1) The amount of shading of a paper sheet refers to the amount of transmitted light, amount of reflected light, amount of magnetism, etc.

(2) Each subsection corresponds to an area that is particularly characteristic of the type of paper, such as a watermark, a portrait, or a text area, and the length is also determined by the distance from the center of the paper. be selected and set in advance.

(3) The reason why the pair of partial sections are arranged symmetrically with respect to the center on the line segment is that when inserting paper sheets into the conveyance means, the front and rear directions are reversed (the paper sheets are reversed in the conveyance direction). This is to ensure that the same calculation result is obtained even if the input is made.

[Effect]

According to the present invention, the first detection means for detecting the presence or absence of a paper sheet and the amount of movement of the paper sheet detects that each sheet of paper actually passes through the second detection means. Since the center point on the line segment is found, the center point can be automatically found each time even for paper sheets of different sizes. Therefore, there is no deviation of the partial section set at the target position with respect to the center point, and the type of paper sheet can be accurately determined based on the calculation result based on the line segment value of this partial section.

〔Example〕

Before describing specific embodiments of the present invention, the principle of the present invention will be explained.

The characteristics of the patterns of the current denominations of 10,000 yen, 5,000 yen, and 1,000 yen notes are roughly as follows, as shown in Figure 1.

(b) 10,000 yen note: There is a dark frame-like pattern around the periphery, a portrait on the right, a watermark and Chinese numerals in the center, and numbers on the left. (See Figure 1 b.) (b) 5,000 yen: There is a dark frame-like pattern around the periphery, a watermark on the right, a portrait in the center, and Chinese numerals on the left. (See Figure 1 d.) (c) Thousand yen note: There is no dark frame-like pattern around it, a portrait on the right side, Chinese numerals in the center, and a watermark on the left side. (See Figure 1 f.) In this example, the denomination is determined using the above-mentioned pattern characteristics of each denomination.

To do this, we focus on the positions where the characteristics of each denomination exist, and determine the patterns of picture frames, watermarks, portraits, and Chinese numerals based on the integral value of the pattern (meaning the amount of light and shade (corresponding amount of electricity)) in the vicinity. do. That is, the first
For example, as shown in Figure a, there are five partial sections symmetrically along the longitudinal direction of the banknote with respect to the center of the banknote.
_T1 to _T5 are provided to integrate the amount of electricity corresponding to the amount of light and shade at predetermined sample points such as the transmitted light pattern, reflected light pattern, or magnetic pattern of each denomination for each section. For example, T ₁ and T ₅ are the dark pattern parts around the 10,000 yen and 5,000 yen notes (i.e., the part corresponding to the frame), and T ₂ and T ₄ are the numbers, portraits, and 5,000 yen of the 10,000 yen note. T ₃ should be placed in the center of each denomination on the accompanying Chinese numerals, watermarks, watermarks and portraits of 1,000 yen notes. For example, the distribution of the amount of electricity corresponding to the shading amount of the pattern along the longitudinal direction in the central part of a 10,000 yen note, a 5,000 yen note, and a 1,000 yen note is as shown in FIG. 1 c, e, and g, respectively. In addition, in the figure, the shaded area indicates the detected amount in the interval from T ₁ to _{T 5} described above. Next, when determining the type of banknote, the direction in which banknotes are usually read is from the front left, from the front right, from the back left, and from the back right (or from the top of the front, from the bottom of the front, from the top of the back, and from the bottom of the back). Four directions can be considered, and the characteristic parameters to be used for type determination must be the same regardless of the direction and use different amounts depending on the denomination. For example, let the integral values of the pattern be S ₁ to _{S 5} for the integral regions T ₁ to _{T 5} in FIG. 1, respectively, and introduce the following characteristic parameters.

X = S ₁ + S ₅ (1) Y = Max (S ₂ , S ₄ ) - Min (S ₂ , S ₄ ) (2) Z = S ₃ (3) X is a feature parameter that emphasizes the picture frame (periphery) It becomes larger than the 5,000 yen note and the 10,000 yen note,
It's almost zero for a 1,000 yen note. Y is T ₂ , T ₄
It is a characteristic parameter representing the magnitude of the difference in the integral value of , which is almost zero for 10,000 yen notes and increases positively for 5,000 yen and 1,000 yen notes. Z has a small kanji numeral in _T3 for a 10,000 yen note, so it becomes zero, and it becomes a positive number for a 5,000 yen note and a 1,000 yen note. These X, Y, and Z characteristic parameters take values within a range unique to each denomination, regardless of the reading direction of the pattern.

Additionally, in the example above, the sensor that detects the amount of shading in the pattern is placed in the center of the banknote, but even if the pattern is read by multiple sensors placed outside the center, the reading direction is the same as above. Feature parameters that are independent of can be introduced. However, even if a banknote is skewed, it is necessary to correct the amount of delay in the pattern in accordance with the amount of skew so that positional deviation between the pattern and the integral section does not occur.

Also, the integral region is T ₁ to T _2N+1 (N=0,
1,... Even if a total of (2N+1) regions are set up to .

Examples of the present invention will be described in detail below.

Fig. 2 shows an overview of the mechanism for sending banknotes (conveying means) that is subject to type determination according to the present invention, a passage detection sensor serving as the first detection means for detecting the passage of banknotes, and the amount of shading of banknotes. An example of the arrangement of a pattern sensor serving as a second detection means for detecting is shown. FIG. 2a is a side view, and FIG. 2b is a front view.

In FIG. 2, banknotes 21 are sent in the direction of the arrow by a roller 22 that is rotationally driven by a motor or the like (not shown).

Passage detection sensor 23- when the presence or absence of banknotes is 1 or more
1, 23-2, and the light and shade pattern on the banknote is detected by the pattern sensor 24-1 placed near the center line _C1 in the longitudinal direction of the banknote, or one or more patterns placed at other predetermined positions. Sensor 24-2
~24-L (L=2, 3, . . . ), etc. are detected.

At this time, the amount of movement of the banknote 21 is detected by an encoder 28 driven by a gear 26 attached to the rotating shaft 25 of the roller 22 via another gear 27 .

The encoder 28 generates an electric signal (pulse) every time the banknote moves a predetermined distance (for example, 1 mm), and thereby samples the shading pattern on the banknote at predetermined intervals regardless of the speed of movement of the banknote. can do.

As this encoder 28, a step-by-step rotary encoder may be used.

The passage detection sensors 23-1 and 23-2 include, for example, a light emitting diode (LED) and a photodiode, or an LED on the light emitting side and the light receiving side, respectively.
The sensor may be of either a reflective type or a transmissive type in which a phototransistor and a phototransistor are arranged in pairs.
Furthermore, as the pattern sensors 24-1 to 24-L, etc., sensors of the same type as the passage detection sensor may be used, and magnetic sensors that detect changes in reactance (see Japanese Patent Application No. 1987-71875). It is also possible to use

FIG. 3 shows a block diagram of a circuit showing one embodiment of the present invention.

The output signals of the pattern sensors 24-1 to 24-L are sent to the acquisition control circuit 32 through amplifier circuits 31-1 to 31-L, respectively. The acquisition control circuit 32 includes an analog multiplexer 321 and an analog-digital converter (hereinafter referred to as ADC) 32.
2, a predetermined analog input signal is selected by a control signal given from a discrimination control circuit (kind determination control circuit; hereinafter referred to as RCC) 33, and it is converted into a digital signal and then sent to the RAM.
34.

In RCC33, passage detection sensor 23-
The encoder 28 after detecting the leading edge of the banknote from the signals obtained by digitizing the output signals 1 and 23-2 in the binarization circuits 35 and 36, respectively.
Control is performed to use the output pulse as the write timing signal for the RAM 34. Passage detection sensor 23
Digital signals corresponding to pattern sensor output values at predetermined sampling intervals in a predetermined area are stored in the RAM 34 from the output signals of -1 and 23-2 until the trailing edge of the banknote is detected.

Note that the skew amount (θ) is the amount of bill movement calculated from the number of output pulses of the encoder 28 until the tip of the bill passes one sensor, with D ₀ being the distance between the two passage detection sensors. When, θ=tan ⁻¹ (l/D ₀ ).

Digital pattern information for one banknote is stored in RAM
34 and the type determination process after storage according to the flowchart shown in FIG.
It is executed in RCC33.

FIG. 5 shows the block configuration of a circuit showing one embodiment of the RCC 33, including the acquisition control circuit 32 and the RAM.
34 is also shown.

Hereinafter, the block configuration of a circuit for implementing each step in FIG. 4 will be explained using FIG. 5.

(i) Step 41 The monostable multivibrator 504 is triggered by the rising edge of the output signal of the OR gate 503, which takes the logical sum of the signals 501 and 502 obtained by digitizing the output signals of the passage detection sensors 23-1 and 23-2, and its output Address counter 506 is reset by the rise of signal 505.

The subsequent output pulses from the encoder 28 pass through an AND gate 507 and become input signals to an address counter 506, which is counted up one by one. The contents of address counter 506 are selected by multiplexer 508 and become the address signal for RAM 34. In sync with this, ANDGATE 5
The output signal 07 is selected by the multiplexer 509 and becomes a write designation signal for the RAM 34.
As a result, the pattern sensors 24-1 to 24 at the sample points on the banknote corresponding to the time point when the encoder output pulse is generated immediately after the leading edge is detected.
-L output signal is selected in the acquisition control circuit 32 and further digitized, and then the RAM 34
is written to address 1. Thereafter, in the same manner, the address counter 50 is
6 is counted up by 1, and 2 of RAM 34 is counted up.
The digital signals at designated sample points of the banknote are sequentially stored at address 3, etc.
When the trailing edge of the banknote passes through the passage detection sensor, the output of the OR gate 503 becomes logic "0" and the address counter 506 stops counting up. If the contents of the address counter 506 at this point are A, then
Digital pattern information for one banknote is stored in RAM3
4, from address 1 to address A.

(ii) Step 42 The contents A of the address counter are shifted to the right by one bit by the shifter 511.

This corresponds to performing the following calculation, and the RAM 34 where the central part of the banknote is stored
address is requested.

(iii) Step 43 Use the output of shifter 511 as an address signal.
The contents of the ROM 512 are read out and stored in the buffer register 513.

However, in the ROM 512, (2N+1) integral regions corresponding to each value are stored at various addresses within a predetermined range from the standard value of [A/2] ([ ] is the Gauss symbol). It is assumed that the starting address of each area has been written in advance.

(iv) Step 44 The number of integration regions (2N+1) is set in the number of regions counter 515 by the output pulse of the monostable multivibrator 514 triggered by the fall of the output signal of the OR gate 503 indicating that the trailing edge of the banknote has been detected. Preset.

Based on the output of this counter 515, (2N+1) is stored in the buffer register 513.
The multiplexer 516 selects one of the top addresses of N 0 , sets this in the area address counter 517 , and selects a predetermined number N _{0 .}
(Number of addresses in each integral area in RAM34)
Pulse generator 518 is activated to generate pulses of .

The contents of the in-area address counter 517 are selected by the multiplexer 508 and the pulse generator 5
Eighteen output pulses are selected by the multiplexer 509 and serve as the address signal and read designation signal for the RAM 34, respectively, and the read contents of the RAM 34 are added to the accumulator 521.

Thereafter, the area address counter 517 is counted up by 1 for each output pulse of the pulse generator 518, and the contents of the RAM 34 at the specified address are sequentially read out and added to the accumulator 521, and the integral calculation is performed. It will be executed. The content of the counter 519 that calculates the number of output pulses of the pulse generator 518 is the predetermined number N ₀
When this is reached, this is detected by decoder 520 and counter 519 is reset. Further, the area number counter 515 is counted down by 1, the start address of the corresponding integral area is set in the area address counter 517, and the pulse generator 518 is restarted to perform integral calculations in the same manner as described above. is executed.

Thereafter, a total of (2N+1) integral values are obtained in the same manner until the content of the area number counter 515 becomes 0.

Every time the integral calculation in each integral region is completed,
The contents of the area number counter 515 and the decoder 51
9 output signals are respectively sent to the multiplexer 52.
2,523 becomes the address signal and write signal for the RAM 524, and the contents of the accumulator 521 are stored in the RAM 524.

(v) Step 45 Specified by the feature parameter calculation unit 525
Multiplexers 522 and 523 whose address signal and continuation signal of the RAM 524 are respectively controlled by an output signal of a decoder (not shown) indicating that the content of the area number counter 515 is 0.
is selected, the integral value in the desired region is read out, and the calculations (1) to (3) above are executed to calculate the characteristic parameters X, Y, and Z.

In other words, Si
(i=1, 2, . . . 5), and based on this, find X using a normal adder, find Y using a normal comparator and adder, and use Si as Z.

(vi) Step 46 X calculated by the feature parameter calculation unit 525,
The collation unit 527 collates Y, Z with the characteristic parameters (Xs, Ys, Zs) of standard banknotes stored in the ROM 526, and outputs the type determination result (discrimination result in the figure) according to the collation result. do.

If the type cannot be determined (indistinguishable from the diagram),
Just output the code that means this.

In the matching unit 527, for example, the following calculation is performed, and this is set as a predetermined threshold value (T).
You can compare it with.

C = |

(1) First method for detecting the presence or absence of paper sheets and the amount of movement
The second detection means determines the center point on the line segment where the paper sheet actually passes through the second detection means for each sheet being conveyed, so it is difficult to detect paper sheets of different sizes. can also automatically find the center point each time. Therefore, there is almost no deviation between the line segment set at the target position with respect to the center point, and the calculation result based on the line segment value of this line segment indicates the exact type of paper sheet. Judgment can be made.

(In the above conventional technology, deviations in the detection area occur due to misalignment of the paper sheet during conveyance and differences in the size of the paper sheet.) (2) Each partial section consists of watermarks, portraits, characters, Since the length or the distance from the center point of the paper sheet is selected and set in advance to correspond to the area that clearly represents the characteristics related to the type of paper sheet, such as the number of paper sheets, it is possible to effectively detect the paper sheet with high accuracy. It is possible to determine the type of leaves.

(In the above-mentioned conventional technology, since the determination is made based on the integral value over the entire section of the paper sheet, it is not possible to utilize the remarkable features related to the type of paper sheet.) (3) In the unit configuration of the present invention, the second detection means is Since there is only one, the device cost can be reduced.

(The unit configuration of the prior art described above requires two second detection means.) (4) Since the length of the partial section and the distance from the center point of the sheet can be set or changed by a program, It can easily respond to changes in conditions such as paper sheets to be judged.

(In the above prior art, the function corresponding to the partial section is a pair of detection means, which are permanently installed at a predetermined position on the conveyance path, so there is no flexibility as described above.)

[Brief explanation of drawings]

FIG. 1 is a diagram showing the output waveform of a pattern sensor corresponding to a typical banknote pattern and a light/dark pattern on the banknote, and FIG. FIG. 3 is a diagram showing a circuit block configuration of an embodiment of the present invention, FIG. 4 is a flowchart of the type determination processing procedure according to the present invention, and FIG. 5 is a diagram showing the discrimination in FIG. 3. 1 is a diagram showing a block configuration of a circuit showing one embodiment of a control circuit; FIG. 21...Banknote, 22...Banknote moving roller, 2
8...Encoder, 32...Intake control circuit, 33
...Discrimination control circuit.

Claims (1)

[Claims] 1 First for detecting the presence or absence of paper sheets and the amount of movement
and a second detection means for obtaining an electric signal corresponding to the density of the paper sheet along the passing direction of the paper sheet, respectively, are arranged at predetermined positions on the conveyance path of the paper sheet, and the above-mentioned Based on the output of the first detection means, the center point of the line segment on the paper sheet when the paper sheet passes through the second detection means is determined, and the At least one at a symmetrical position at a predetermined distance along the direction of leaf passage.
Setting a pair of subintervals, determining the integral value of the output of the second detection means for each of the set subintervals, and performing an operation to determine the absolute value of the sum or difference for each pair, A method for determining the type of paper leaf, characterized in that the type of the paper leaf is determined by checking whether the absolute value of the sum or difference determined by the calculation is within a preset range.