DE102005046550C5 - Method and device for sorting mailpieces - Google Patents

Abstract

a) Method of sorting mailpieces,
b) wherein the mailpieces are sorted in consideration of data signals of a machine-readable bar code located on the mail pieces, which is formed from a number of bars,
c) wherein, when printing one of the mailpieces, the machine-readable code is generated such that it has one or more data coils,
d) at least one of the data tracks containing a reference clock,
e) at least one of the data tracks contains sorting information,
f) wherein a reading unit generates a graphical image of the machine-readable code,
g) wherein the graphic image is stored in a memory of a data processing unit,
h) wherein the stored image is evaluated so that differences between signal intensities are used to determine clock signals of the reference clock,
i) wherein the data signals are evaluated by assignment to the clock signals of the determined reference clock, that thereby the sorting information is determined and
j) whereby a sorting of the mailpieces in ...

Description

The The invention relates to a method for sorting mailpieces, the mailpieces taking into account data signals a sorted on the mail machine-readable codes sorted become.

From the German translation DE 696 19 147 T2 the European patent specification. EP 0 732 666 B1 For example, a bar code encoding system is known for use in a mailing machine. For this purpose, this document proposes to apply a bar code in a post office, wherein the bar code contains information about a destination and an address number. In another post office, the bar code is read to recognize the information regarding the destination and the address number. On the basis of the recognized information, the mailpieces are sorted according to the addresses.

task The invention is to perform the sorting of mail so that a fast and reliable Sorting of the mailings allows becomes.

According to the invention this Problem solved by a method according to claim 1.

The The method provides for a sorting of mail items a machine-readable code. This is the machine-readable Code generated so that it has one or more data tracks.

At least contains one of the data tracks a reference clock.

Further is intended to use a reading unit which is a graphical Image of the machine-readable code created.

Further a deployment of a data processing unit is provided which equipped so that it contains a memory in which the graphical image of the machine-readable code is stored.

The saved image is evaluated so that differences between signal intensities used to determine clock signals of the reference clock become.

The Data signals are assigned by assignment to the clock signals of the determined Reference clock evaluated.

Preferably this is done by having the data processing unit equipped is that they determine the differences between the signal intensities and an assignment of the data signals to the thus determined clock signals can make.

By the assignment of the data signals to the clock signals of the determined Reference clock, it is possible determine what value a data signal at a predeterminable Data signal position and has the data signals so particularly simple and reliable Way to evaluate.

The Evaluation can be done in several ways, at least a reference clock is used.

In an embodiment of the method at least one of the data tracks has both information about the Reference clock as well over Sorting information.

In an embodiment of the method one of the data tracks the reference clock, and at least one other Contains data track the sorting information.

at The sorting information is information that is used for sorting the mail can be used.

Basically, this can different types of sorting information are used.

A preferred embodiment of the sorting information is address information of a recipient of the mailpiece. The address details may vary depending on the intended sorting purpose be.

In simple embodiments For example, the address information may be a postal code act.

There there is a need, one possible To enable extensive sorting of the mail, it is expedient to further To include information in the address information and thus this to use as sorting information.

Especially can so street names, road sections and / or house numbers or house number ranges as sorting information be used.

The Sorting information can additionally or in place of the information mentioned further information contain.

there It may also be an identification, in particular a Identification number, act.

The Invention includes several advantages.

Especially the code used is small and secure.

Besides that is to apply the code safely and quickly. Furthermore, he can also be read quickly and safely.

By the use of the reference clock, it is possible to correct the code error embody.

• a. Auslöschungen (Falten, verwischte Signale),
• b. stochastische Fehler (Verschmutzung),
• c. systematische Fehler (Ausfall eines Dots, periodisch).

In particular, this can correct the following errors:

• a. Erasures (wrinkles, blurred signals),
• b. stochastic defects (pollution),
• c. systematic errors (failure of a dot, periodic).

A expedient embodiment the method according to the invention is characterized by the fact that a machine-readable code with at least two data tracks is generated, one of the data tracks being the reference clock and one of the other data tracks contains sorting information.

A particularly preferred embodiment the method according to the invention is characterized by the fact that the machine-readable code is generated in this way will be that it contains at least three data tracks.

in this connection it is particularly appropriate that the data track containing the reference clock is between other data tracks located.

Basically it is possible that this data track, which contains the reference clock, at any Place located between the other data tracks.

It However, it is particularly advantageous to generate the code so that the data track containing the reference clock is essentially centered located between the other data tracks.

Of the Term "essentially centered " different embodiments.

Especially preferred embodiments are where there is an odd number of data tracks and the Data track containing the reference clock is arranged so that Join in on her two sides the same number of other data tracks Sorting information is located.

A However, substantially central arrangement is also present when yourself - for example with an even number of data tracks - the data track that is the reference clock contains near the center of the data tracks with sorting information is located.

One An essential advantage of the invention is that the codes shown yourself for both fast printing of mail as well as fast Reading the codes are suitable.

Especially it is also possible the codes according to the invention to be used with direct printing by means of ink jet. This is despite the resulting in the printing process tolerances possible. This represents a significant advantage over known bar-no-bar codes represents.

A expedient embodiment the method according to the invention is characterized by the fact that the generation of the machine-readable Codes are made using interleaving.

According to the invention Method performed so that at least individual components of the code are inverted.

A particularly preferred embodiment the method according to the invention is characterized by the fact that additional information is introduced into the code which allow error correction.

in this connection it is particularly appropriate that a Reed-Solomon error correction method is used.

The inventive design the codes lets Both as a further development of known codes as well as in the Use new development of codes.

The Invention allows the use of a basically arbitrarily long code.

It is thereby possible the code length fast and reliable changing adapt postal requirements.

This is for example advantageous if a change between different Sorting should be done. For example, it is possible at a transition from a sorting to delivery districts to a sequence sorting the code length adjust accordingly.

Either for short codes as well as long codes, the reference clock is used the assignment of the respective signals to the associated signal positions.

hereby the recognition accuracy of the signals is fulfilled.

following is exemplified how a 4-state code according to the invention further developed can.

Around More information was encoded at high speeds the 4-state code developed, which encrypts two bits in a stroke and thus has four states.

The Invention leaves however, alike for any Use n-state codes. Where n is the number of states.

A important strength of the inventively designed Codes is that it can be printed very flexibly and yet is readable. In particular, the data density is over one-dimensional Barcodes higher.

following is a particularly preferred use of the invention in the processing of mail in letter centers or freight centers shown. Here are usually several 10,000 postal items sorted within one hour.

Further Advantages, special features and expedient developments of the invention emerge from the dependent claims and the following description of preferred embodiments based on the pictures.

From the pictures shows:

1 Basic structure of a code

2 Basic principle of the representation of the codes in the form of bars;

3 Representation of the code as a 6-bit pattern;

4 Schematic representation of a generalized n-state code;

5 a schematic diagram of a code according to the invention, in which the data line containing the reference clock is designed as a clock line and contains the four data lines;

6 in five partial images, a use of character-wise interleaving in the generation of codes according to the invention;

7 in three fields a use of a barwise inverting;

8th a code with inverted bars;

9 a printed image with maximum inclination.

In The figures illustrate the use of the invention in labeling of mail with a machine-readable code and a subsequent reading of the machine-readable code with extraction sortierrelevanter Information described using the example of a letter sorting system.

Especially the invention is suitable for a use for mass marking of mail as well the likewise mass sorting of mailpieces in one Letter center or a parcel center.

It however, it is also possible perform at least one of the processing steps in a smaller plant.

For example Is it possible, to generate the code with a franking machine.

Further Is it possible, to generate the code with a printer, with the printer on Computer system is connected.

It is particularly advantageous to the code and other information - especially a recipient address in Plain text - in to print a uniform editing process.

As well Is it possible, the sorting of the mailpieces in an imaginary for smaller shipment quantities Perform device, for example an internal mail distribution within a company.

The Invention leaves also use with codes that are designed as modified barcodes are.

Of the Term "bar" refers to one single stroke in a barcode. The bars vary depending on the symbology chosen different widths (eg 2- from 5, code 128) or different lengths (eg 4-state). Symbology is the kind of physical coding designated by pressure.

It however, it is also possible to use the invention for further development of other codes, for example Batch codes (code block, PDF417) where the information in Lines (tracks) arranged one above the other become.

The Invention can also be used in the integration of symbols become.

following the term "symbol" becomes the representation of an element from the set of all representable signs with the selected Symbology shown.

The Set of representable characters is also called alphabet. Every symbol needed in a binary Representation of a fixed number of bits, this is determined by the number of possible Icons set in alphabet.

following An encoding is shown using a 6-bit symbol.

These Symbols then form the basis for the error correction. Ie. not individual bits are corrected, but always whole symbols with 6 bits. The alphabet here includes 64 symbols.

With Track becomes a reading line for a code that is made up of several one above the other arranged rows is called.

As on a tape, the code passes the fixed reading head, so that the scanning is track by track.

Basically it also possible that a movable reading head in the longitudinal direction of the code relative to the Code is moved.

Thereby, that previously recorded images of the machine-readable code evaluated is also a capture of the code in a single step possible. This leaves reach for example by using imaging means, otherwise in other technical fields, such as digital photography or digital copying.

• Kürzeste Variante = A1
• 10 Zeichen aber nur 30 Bit notwendig!
• Bei 6 Bit-Symbolen = 5 Symbole (entspricht UPU serial number)
• Nicht sprechend/keine Fehlerrekonstruktion
• Bevorzugte Variante = A8
• 14 Zeichen mit 56 Bit notwendig
• Bei 6 Bit-Symbolen = 10 Symbole
• Sprechend/keine Fehlerrekonstruktion/keine Synchronisation.

There are a number of variants that make it possible to create a unique identification number. The preferred variant is referred to below as A8. Variant A1 defines a consecutive number and is therefore the shortest possible with 30 bit information.

• Shortest variant = A1
• 10 characters but only 30 bits necessary!
• For 6-bit symbols = 5 symbols (corresponds to UPU serial number)
• Not talking / no error reconstruction
• Preferred variant = A8
• 14 characters with 56 bits necessary
• For 6-bit symbols = 10 symbols
• Speaking / no error reconstruction / no synchronization.

It Further variants were considered included (A12-A14), the date (MMTT or day number) and the consecutive counter (in 1/100 sec since 0:00 or consecutive day counter).

The most preferred version shown here is A14:
Directly 7 · 6 bit symbols
Speaking / no error reconstruction / no synchronization.

These is structurally identical to the previously preferred variant A8 and with 42 bit information just a little bit longer than the absolute lower bound from 30 bits. It also provides 8 bit reserve for future use.

following a particularly preferred construction of the payload is shown.

• 2 Symbole für BZ-Nummer und Maschinennummer
• damit sind 4096 Codierer* unterscheidbar
• Reserve 2 Bit
• 1 Symbol für den Tag+
• der Tageszähler beginnt mit 1 am 1.1./1.3./1.5./1.7./1.9./1.11.
• Reserve 2 Bit
• 4 Symbole fortlaufender Zähler°
• an einem Tag
• oder Zeit in 1/100 sec
• seit 0:00 Uhr
• Reserve 4 Bit
• Benötigt werden 7 Symbole 6 Bit, darin ist die Fehlerkorrektur noch nicht enthalten!

Particularly preferred 6-bit symbols are used because they are compatible with Reed-Solomon error correction. The payload is distributed as follows:

• 2 symbols for BZ number and machine number
• so 4096 encoders * are distinguishable
• Reserve 2 bits
• 1 symbol for the day +
• the day counter starts with 1 at 1.1./1.3./1.5./1.7./1.9./1.11.
• Reserve 2 bits
• 4 symbols of consecutive counters °
• in one day
• or time in 1/100 sec
• since 0:00 o'clock
• Reserve 4 bits
• 7 symbols are required 6 bits, the error correction is not included here!

The remaining reserve an integration of future ones Further developments.

• – schlechter Druck
• – Untergrund, der die Lesung behindert
• – nachträgliche Veränderung (Verschmutzung, Beschriftung)
• – Abdeckungen/Auslöschungen z. B. durch Falten
• – usw.

When choosing the code, care should be taken that it fits as well as possible to the expected error structures. The different error situations are for example:

• - bad pressure
• - Underground obstructing the reading
• - subsequent change (soiling, labeling)
• - Covers / cancellations z. B. by folding
• - etc.

It is possible, a complete To achieve error correction of the code according to the invention.

critical is the percentage of payload within the code.

used is preferably a Reed-Solomon error correction on symbols with example 6 bits. It was in the bill start / stop or synchronization mark included as these as well the reading security by adding of redundancy increase. On average, a code rate of preferably at least 20% is used. It is even more advantageous to use a code rate of at least 30%, at a code rate of at least 40% further improvements be achieved. The illustrated embodiments relate to a particularly preferred code rate of about 46%.

One contains particularly preferred code Payload 42-bit / 98-bit code = 42.9%, which is 2 × 2 bits Start / stop included.

• Max. Korrektur fehlerhafter Symbole 4 (3)
• Max Korrektur gelöschter Bars 8 (7)
• Korrektur Bündelfehler Bars 10 (7)
• Korrektur Bündellöschungen Bars 22 (19)

This setting quantifies the correction options as follows:

• Max. Correction of incorrect symbols 4 (3)
• Max Correction of Deleted Bars 8 (7)
• Correction bundle error bars 10 (7)
• Correcting Trunk Deleting Bars 22 (19)

In Clamp the values for 7 error correction symbols.

Especially in the correction of bundle errors and bundle deletions (both burst errors) 3 bars are corrected more. Burst errors, d. H. faulty or coherent Bars are z. B. to expect in wrinkles.

It is possible, an error correction according to the UPU specification with the insert preferred adaptations shown below.

This is done by way of example by the following error correction method: Error correction method: Reed-Solomon Galoir field: GF (64) = GF (2 ⁶ ) Primitive polynomial: p (x) = x ⁶ + x + 1 Generator polynomial: g (x) = Π ⁸ _{i = 1} (x + α ⁱ ) Generator element: α = 000010 = 2

Of the Code structure is systematically analogous to UPU.

It however, it is also possible to use alternative error correction methods.

following such alternative error correction methods are explained: two important code types are the block code and the convolutional code. In the above Chapter was a block code for error correction as required selected.

At the Block code, the input data is divided into blocks of length m (m = number of symbols) and after each block become k redundancy bits added thus amounts to the new block length n = m + k bits. The code rate R is defined as the ratio of Information bits m to the total block length n. Block codes are suitable thus for the correction of symbol errors.

By contrast, the convolution code "blurs" the input data over several output bits. The input data are read into a shift register and the output data through the combination different taps on the register generated. The code rate R is defined here as the quotient of the m bits read in one time to the n bits read out once. By this type of coding, convolutional codes are suitable for correcting individual bit errors.

convolutional codes are binary codes, where the input bits over several output bits are "smeared". In the Endcodierung the input data are read into a shift register and the output data is determined by combinations of taps (Mostly they are Exor links).

With the length S of the shift register results in a memory depth of S times m = 3. The influencing length is against K = (S + 1) times m = 4. The arrangements of the taps in the coders become common indicated by generator polynomials or as an octal number.

A Possibility, In order to increase the efficiency of codes, there are several Concatenate codes together. The first code becomes outer code, called the second inner code.

Becomes for example, as an outer code a block code and a convolutional code selected as the inner code, so For example, the inner code can correct for single bit errors and the outer smaller ones Burst errors. To even longer burst errors to be able to correct you switch between the two coders an interleaver.

It is advantageous, the coding in the particular application in dependence to choose from the expected errors.

Accordingly Also, the illustrated Reed-Solomon coding is merely exemplary to understand and can in each individual case by another error correction procedure be replaced.

• 1. Startzeichen 2 Bars
• 2. Nutzinformationen 21 Bars
• 3. Fehlerkorrektur 24 Bars
• 4. Stoppzeichen 2 Bars

A particularly preferred code is constructed as follows:

• 1st start sign 2 bars
• 2. Payload 21 bars
• 3. Error correction 24 bars
• 4th stop sign 2 bars

Out For printing or mechanical reasons, the representation must be mirrored to account for the passage through the machine to wear. The mathematical properties remain unaffected. The Decoding from a camera image (CCD from the OCR scanner) is without restriction possible.

With the determination of the content is still the symbology to use not fixed. Symbology is the kind of physical coding designated by pressure. Known symbologies are z. Eg barcodes.

One Advantage of using modified 4-state codes according to UPU is that makes it interoperable with other postal companies is achieved.

• Full = H = A + D + T = 00 = 0
• Ascender = A + T = 01 = 1
• Descender = D + T = 10 = 2
• Tracker = T = 11 = 3
• Start/Stop bestehen aus 2 Bars Start = AT = 13
• Stop = TA = 32

The bars are set as follows:

• Full = H = A + D + T = 00 = 0
• Ascender = A + T = 01 = 1
• Descender = D + T = 10 = 2
• Tracker = T = 11 = 3
• Start / Stop consist of 2 bars Start = AT = 13
• Stop = TA = 32

each Icon consists of 4 bars (6 bit) arranged as shown in the picture are.

The Example A + F + T = 1 × 16 + 0 · 4 + 3 = 19

following is a use of generalized n-state codes VnSC represented.

The 4-state code can also be interpreted as data transmission by means of three lines. 4 shows the scan of a 4-state code with 10 pressure points on top of each other.

The two data lines correspond to two signal lines and the interpretation is from the clock specified. The clock is represented by 4 points, the two data lines are represented by three dots each. The following describes the lines with track.

It reading is thus a parallel reading of the tracks, whereas the transfer between Transmitter (encoder printer) and receiver (reader decoder) is asynchronous. The synchronization the reader takes over the clock in the middle of the code.

VnSC variant 1

The generalized n state code is defined as follows:
For a given clock in the middle track above and below the data is displayed as a bit sequence. Where n is the number of tracks. All data tracks have the same clock, which must be maintained with a lower tolerance. The individual points do not have to be exactly on top of each other. Ie. in the printing direction, a temporal offset within a bar is allowed.

each Data track doubles the number of possible states of a Bars. This makes the familiar 4-state code identical to V2SC. in the Borderline is for each data track uses only one point.

Thereby The result is an efficient dot code with high data density for high printing speeds independently distortions in the direction of travel.

The Example shows a V4SC encoded with 24 = 16 states in each bar can be.

As In the initial example, each bar is displayed with dots on top of each other. each Track features over here 2 points.

VnSC variant 2 (interleaved)

By Inverting every second bar and pressure into the interstices becomes the data density doubles.

This The procedure is described in detail in Chapter 4-State interleaved.

VnSC variant 3 (autotaktend)

There can write all data tracks with the same clock, the clock be regenerated from the data itself. This will be a self-timed Coding uses z. B. Manchester. In this case, the receiver can the Derive clock rate directly from the data stream.

The Data density is thereby further increased, in the borderline case to the maximum print density of the printer used.

VnSC variant 4 (interleaved autotact)

These Variant is a combination of variants 2 and 3.

For all variants an error correction is possible. This must be adapted to the problem specific to the expected To compensate for errors.

Of the 4 State Interleaved is used as parallel data transmission over 2 lines with clock line construed.

All known methods for clock regeneration and signal quality improvement can be so be used.

Central is the regularity the code that needs to be tuned to the printer (clock regeneration at reading).

Since the printing resolution depends on the conveying speed and the number of dots per bar, depending on the printer used, the parameters for a precise print image must be matched become.

ever leads to ink and underground Run during the drying time for the deterioration of the printed image.

Possibly. should be left free between two bars when printing 1 point.

The Print image must be for optimally adapted to the printer used. The number of Points should be chosen accordingly.

Big pressure points have the advantage that even with overlapping points the bar line remains visible.

little one Pressure points can however, be processed as well.

By different thicknesses / sizes of Changes pressure points only the duty cycle in the bar line.

Differently size Lead points to temporal offset within the tracks (edge uncertainty jitter)

Also here is the correction z. B. with a dPLL possible.

The Frequency remains the same at different thicknesses of the pressure points. She lets z. B. with FFT (Fast Fourier Transformation) from the overall picture reconstruct.

The Data tracks are preferably printed / scanned at twice the frequency.

• 1. Zweidimensionale Codes weisen spezielle Suchmuster auf, die für das Auffinden des Codes im Bild benötigt werden (Bild). Zusätzlich liefern die Suchmuster auch Informationen über die Orientierung und die dem Code zugrunde. liegende Druckmatrix. Dieser Code kann auf Suchmuster verzichten, da Orientierung und Positionierung durch die Fördertechnik und die Anordnung der Codierer bzw. OCR-Scanner vorgegeben sind.
• 2. Gemeinsam ist allen 2-dimensionalen Codes, dass die Anordnung der gedruckten Zeichen in beide Richtungen definiert ist, d. h., der Code muss sowohl in Transportrichtung den Spezifikationen entsprechen als auch senkrecht dazu. In der Praxis bedeutet dies, dass zumindest scharfe senkrechte Striche gedruckt werden müssen. Dieser Code benötigt durch die Spurorientierung keine exakte Ausrichtung senkrecht zur Transportrichtung. Der gemeinsame Takt der Aufzeichnung in allen Spuren ist ausreichend.
• 3. Die Seitenverhältnisse der von 2-D-Codes unterliegen festen Regeln, die obigen Codes z. B. müssen stets quadratisch dargestellt werden. Dies begrenzt automatisch die Menge der codierbaren Daten, da insbesondere beim Inkjet-Druck nur begrenzt Punkte in der Vertikalen zur Verfügung stehen. Es werden hingegen im Wesentlichen parallele Datenspuren bereitgestellt. Die einzelnen Spuren können im Prinzip beliebig lang werden, sofern der Platz für den Druck verfügbar ist.
• 4. Gestapelte zweidimensionale Codes (PDF417, Codablock) sind flexibel in den Seitenverhältnissen und damit in der Datenmenge. Sie nutzen verschiedene Strichstärken (Mehrbreitencodes), die im Druck einzuhalten sind. Bei hohen Druckgeschwindigkeiten ist dies jedoch nur begrenzt möglich.

The generalized n-state code offers many advantages in particular for direct printing at high speeds compared to all known two-dimensional codes. Advantages of this code are:

• 1. Two-dimensional codes have special search patterns needed to find the code in the image (image). In addition, the search patterns also provide information about the orientation and the underlying code. lying pressure matrix. This code can dispense with search patterns, since orientation and positioning are predetermined by the conveyor technology and the arrangement of the encoder or OCR scanner.
• 2. Common to all 2-dimensional codes is that the layout of the printed characters is defined in both directions, ie the code must conform to the specifications both in the direction of transport and perpendicular to it. In practice, this means that at least sharp vertical bars must be printed. Due to the track orientation, this code does not require an exact alignment perpendicular to the transport direction. The common tact of recording in all tracks is sufficient.
• 3. The aspect ratios of 2-D codes are subject to fixed rules, the above codes z. B. must always be displayed square. This automatically limits the amount of encodable data, as there are only limited vertical points available for inkjet printing. In contrast, essentially parallel data tracks are provided. The individual tracks can in principle be arbitrarily long, provided that the space for printing is available.
• 4. Stacked two-dimensional codes (PDF417, Codablock) are flexible in the aspect ratios and thus in the amount of data. They use different line widths (multi-width codes) that must be adhered to in printing. At high printing speeds, however, this is only possible to a limited extent.

A particularly preferred embodiment The invention is characterized in that the machine-readable Code and an address information in plain text in one printing process be applied to the mailpiece. It is appropriate, the inventive codes to design them so that they are only a single set Have width of the bars. As a result, they are of line effect unaffected and are thus suitable for high-speed printing.

At the 4-state code is the code length essentially determined by the number of bits to be coded. The Graphic shows the three variants considered compared to UPU as standard. The minimum area (minimum Height = 4.0 mm / minimum pitch * = 1.14 mm) _the maximum area (max Height = 5.8 mm / maximum pitch = 1.26 mm)

The Codes are still pretty big. A suitable method to significantly reduce the code length is interleaving, that is the compression by nesting the data.

• – Interleaving für ganze Zeichen
• – Interleaving mit einzelnen Bars

Basically, two variants can be considered:

• - Interleaving for whole characters
• - Interleaving with individual bars

Both Variants are shown below.

Sign-wise interleaving

The 7 symbols of payload will start and stop "normal" as 4-state code shown. Together with the 2 bars each for the start and stop sign 25 bars are printed. These are separated by 24 gaps. In the 24 = 8 · 3 gaps printed the 8 symbols of the RS error correction codes.

6 clarifies the process. In the two upper fields of 6 each two sub-images, each containing a symbol shown.

In the middle row of 6 the right symbol is inverted.

In the bottom line of 6 It shows how the two symbols are nested. This comb-like nesting reduces the code by about 50% in length.

simultaneously is achieved by inverting the test marks that the Clock line in the middle of the code is completely preserved.

For the reading only then must the information in the gaps be interpreted if the code alone is unreadable.

The disadvantage of this compression is that a local error immediately damages 2 symbols. This will make the reserve of 4
correctable characters quickly consumed.

The same thing Result regarding Density and tact leaves also reach alternatively.

Barweises inverting

Of the entire code consists of 2 + 7 · 3 + 8 · 3 + 2 bars = 49 bars.

• normaler Bar,
• inverser Bar,
• normaler Bar.

Every second is now displayed inverted and the pitch between the bars is halved. This means that all odd-index symbols are represented as follows:

• normal bar,
• inverse bar,
• normal bar.

• inverser Bar,
• normaler Bar,
• inverser Bar.

All even-numbered symbols are displayed:

• inverse bar,
• normal bar,
• inverse bar.

7 clarifies the effect.

The The result is a compact error-proofed code. The clock in the Middle stays for the reading systems always received. The transition from dark to hell must trigger the inversion during the read process.

The resulting overall image with inverted bars is in 8th shown.

6 barcode characteristic & reading

used becomes the UPU specification. This is now modified to that the distance (pitch) of 1.2 mm refers to the distance between two "normal" and two "inverted" bars. Of the Distance between "normal" and "inverted" is 0.6 mm.

Central is the regularity of the code that needs to be tuned to the printer (clock generation at reading). Because the print resolution depending on the printer used by the conveyor speed and the Number of points per bar depends is, must the parameters for a precise print image are matched to one another.

ever leads to ink and underground Run during the drying time for the deterioration of the printed image. Possibly should be left free between two bars when printing 1 point.

Around the code should have a quiet zone of 2.5 mm. This equals ½ UPU. The comparison with the Australian solution suggests that the quiet zone may be lower above and below. It should be 1 mm not be fallen below.

This Code can be sent directly (for example via inkjet) on the shipment or on a label, which is stuck on a consignment (machine and / or manually). The use of a label is with problematic background (imprint, geometry of the shipment, ...), because they are pre-produced in high quality (on roll) can be.

• minimal 295,2 mm²
• maximal 385,6 mm²

The height of the bars is according to UPU between 4 mm and 5.8 mm. A14 with inversion thus requires including quiet zones

• minimal 295.2 mm ²
• maximum 385.6 mm ²

reading

Around Reading 2-D codes requires CCD scanners or CCD cameras with it a picture for the evaluation can be created. Alternatively, the picture may look like many scan lines are composed when the scan is fast enough done. Since the code can also be skewed / skewed (max 5 °), must the reading at full transport speed at least twice resolution have like the code (sampling theorem).

There the code also oblique / wrong may be present (maximum 5 °), the reading at full transport speed must be at least the double resolution have like the code (sampling theorem). Known offers point to the OCR reading at least one resolution from 200 dote-per-inch. Ie. per millimeter will be more than 7.5 Points resolved. A bar minimum width of 0.3 mm will be significantly more than 2 times detected.

9 shows a printed image with maximum inclination.

Left the bars are shown with minimum width and maximum distance. That's how the gap is between the bars maximum.

in the right part is shown the opposite case. This will overlap the bars are a bit small. A CCD line with the width as in the picture can resolve both cases.

There the code through the compression is much smaller, it must be Pay attention that even shipments of variable thickness are always in focus of the reading system are (depth of focus).

In In a simple test, the 4-state code was scanned at 200 DPI.

Even though the printed image is very irregular, Each pressure point was detected several times by the scanner.

The Distortions of individual points are different Attributed raster between print and scan. This effect will too occur in the application and is to be expected back.

at The present print is a point between the bars kept free Service. In this gap will be In variant A14 the inverted bars are printed.

at the reading or decoding of a video image occurs depending on the compression increasingly disturbances which can be addressed by appropriate technical measures. These are shown here briefly.

The Scanned image shows the central problem. The information above and below the clock line are not synchronized with to the beat.

to Correction can be a digital "phase Locked loop "used become.

The literal but true translation of "Phase Locked Loop" (PLL) would be:
"Phase-locking loop". The term "loop" stands for "control loop". The PLL is thus a control circuit in which the control is performed by a constant ("latched") phase.

Basic Tasks of the PLL are:

retiming

In all manifestations is for to regenerate the decoding of the clock from the picture. Given is a reference clock, of which the nominal frequency is known, but only temporarily available. In the reference signal can So partially flanks fail. For further processing, these must be missing Flanks added again so that a continuous cycle is available.

clock synchronization

given Here is a reference clock to which generates a clock of the same frequency shall be. This in itself strange Task arises if the reference clock has properties which exclude a direct use, see above.

So could the reference clock is a very unfavorable one or even fluctuating duty cycle exhibit.

To could a strong short-term flank uncertainty (jitter) come. Of the In contrast, the derived clock should have a constant duty cycle and have no or only low edge uncertainty.

Horizontal scan

It is particularly advantageous to make a horizontal scan.

This has several advantages.

Especially this makes a multiple reading of the code possible.

Further are line scanners used.

The Invention allows also a solution other resulting from the generation or processing of the codes Problems.

For example can be through an oblique extending baseline a horizontal offset arise.

One another reason for the emergence of a possible horizontal offset is a crooked print.

The resulting phase shift (skewing) of the data tracks leaves to the tactile trace Avoid one or more of the following ways: One particularly preferred correction is done with dPLL (digital phase locked LooP).

To Individual level changes are required in the data tracks. Singular unfavorable codes are therefore not used.

• – Finden des Codes;
• – Geometriebestimmung;
• – Suchen der Taktzeile;
• – Decodierung der Zeichen durch Bildverarbeitung;
• – Korrekturen an Störungen im Bild;
• – Fehlerkorrektur/Variieren des Codes;
• – Anpassung lediglich der Decodierung der Zeichen an den neuen Code.

The basic procedure preferably consists of the following steps:

• - finding the code;
• - geometry determination;
• - search the clock line;
• - decoding the characters by image processing;
• - corrections of disturbances in the picture;
• Error correction / variation of the code;
• - Adjusting only the decoding of the characters to the new code.

These Coding has several advantages.

Especially Is it possible, UPU specifications.

The international organization Universal Postal Union (UPU) has on the Internet at the Internet address
http://www.upu.int/standards/en/lists/upu_code_lists.shtml
published in English a list of codes issued by her.

• 0 UPU Master Code List
• 1 UPU Standard numbers
• 100 Consignment Event Codes
• 101 Consignment Status Codes
• 102 Container Identifier Assignment Agency Codes
• 103 Container Event Codes
• 104 Dangerous Goods Type Code List Maint. Agency Codes
• 105 EMSEVT Event Codes
• 106 Event Reason Codes
• 107 Handling Class Codes
• 108 International Mail Processing Centre (IMPC) Codes
• 109 S31 Issuer codes
• 110 Item Customer Charging Arranging Codes
• 111 Item Customs Retention Codes
• 112 Item Non-Delivery Reason Codes
• 113 Item Non-Delivery Action Taken Codes
• 114 Location Type/Function Codes
• 115 Mail Category Codes
• 116 Mail Class Codes
• 117 Mail Sub-Class Codes
• 118 Organisation Identifiers
• 119 Receptacle Event Codes
• 120 Format of Contents Codes
• 121 Receptacle Type Codes
• 122 Receptacle Weight Type Codes
• 123 Responsible Agency Codes
• 124 Service Indicator Codes
• 125 TRAKIT Event Codes
• 126 Transport Mode Codes
• 127 Accounting Method Codes
• 128 Date/Time Status Indicator
• 129 Container Disposition Codes
• 130 Representation Symbology Codes
• 131 Level of Presence Codes
• 132 Item Dutiable Indicator Codes
• 133 Item Duty Charging Arrangements codes
• 134 Proof of Delivery Level Codes
• 135 Reason Codes
• 136 Item Nature Indication Codes
• 137 Item Non-Delivery Disposition Codes
• 138 Payment On Delivery Method Codes
• 139 Location Codes
• 140 Payment an Delivery Validation Method Codes
• 141 Service Data Codes
• 142 Entity Type Codes
• 143 Entity Condition Codes
• 144 Event Report (EVTRPT) Event Codes
• 145 Event Status Codes
• 146 Location Function Codes
• 148 Entity Status Codes
• 149 S35 Party-ID Category Codes
• 150 UPU EDIFACT Data Element 9164 Group Codes
• 151 Money order product types
• 152 Money order product qualifier codes
• 153 Party role
• 154 Entity type code
• 155 Location function codes
• 156 S37 formst codes
• 157 ISO/IEC 15459 identifier prefix codes
• 158 Container type
• 159 Interchange network identifier codes
• 160 Interchange network party addresses
• 161 Aggregate mailstream segregation codes
• 163 Basic service codes for letter mail
• 164 Supplementary service codes for letter mail
• 165 Basic service codes for parcels
• 166 Supplementary service codes for parcels
• 167 Basic service codes for EMS
• 168 Supplementary service codes for EMS items
• 169 Basic service codes for empty receptacles
• 170 Supplementary service codes for empty receptacles
• 171 Defining authority codes
• 172 Service code process data type
• 173 Size units codes
• 174 Weight units codes.

The codes published on the Internet at the time of filing of the present application have the following English names:

• 0 UPU Master Code List
• 1 UPU standard numbers
• 100 Consignment Event Codes
• 101 Consignment Status Codes
• 102 Container Identifier Assignment Agency Codes
• 103 container event codes
• 104 Dangerous Goods Type Code List Maint. Agency Codes
• 105 EMSEVT event codes
• 106 event reason codes
• 107 Handling Class Codes
• 108 International Mail Processing Center (IMPC) codes
• 109 S31 issuer codes
• 110 Item Customer Charging Arranging Codes
• 111 Item Customs retention codes
• 112 Item Non-Delivery Reason Codes
• 113 Item Non-Delivery Action Taken Codes
• 114 Location Type / Function Codes
• 115 Mail Category Codes
• 116 mail class codes
• 117 Mail sub-class codes
• 118 Organization Identifiers
• 119 Receptacle Event Codes
• 120 Format of Contents Codes
• 121 Receptacle Type Codes
• 122 Receptacle Weight Type codes
• 123 Responsible Agency Codes
• 124 service indicator codes
• 125 TRAKIT Event Codes
• 126 transport mode codes
• 127 Accounting Method Codes
• 128 Date / Time Status Indicator
• 129 Container Disposition Codes
• 130 Representation Symbology Codes
• 131 Level of Presence Codes
• 132 Item Dutiable Indicator Codes
• 133 Item Duty Charging Arrangements codes
• 134 Proof of Delivery Level Codes
• 135 reason codes
• 136 Item Nature Indication Codes
• 137 Item Non-Delivery Disposition Codes
• 138 Payment On Delivery Method Codes
• 139 Location Codes
• 140 Payment to Delivery Validation Method Codes
• 141 Service Data Codes
• 142 entity type codes
• 143 entity condition codes
• 144 Event Report (EVTRPT) Event Codes
• 145 event status codes
• 146 Location Function Codes
• 148 entity status codes
• 149 S35 Party ID Category Codes
• 150 UPU EDIFACT Data Element 9164 Group Codes
• 151 Money order product types
• 152 Money order product qualifier codes
• 153 party role
• 154 entity type code
• 155 Location function codes
• 156 S37 molding codes
• 157 ISO / IEC 15459 identifier prefix codes
• 158 container type
• 159 Interchange network identifier codes
• 160 Interchange network party addresses
• 161 Aggregate mailstream segregation codes
• 163 Basic service codes for letter mail
• 164 Supplementary service codes for letter mail
• 165 Basic service codes for parcels
• 166 Supplementary service codes for parcels
• 167 Basic service codes for EMS
• 168 Supplementary service codes for EMS items
• 169 Basic service codes for empty receptacles
• 170 Supplementary service codes for empty receptacles
• 171 Defining authority codes
• 172 Service code process data type
• 173 size units codes
• 174 weight units codes.

Even though it basically possible is to generate and process codes according to other standards the illustrated UPU codes are particularly advantageous because they provide compatibility for editing of mail through several postal companies.

Further a clear identification of the codes is possible.

Further can running counters to get integrated.

The Benefits An error correction with Reed-Solomon on GF (64) and 8 characters have already been explained above.

Further can Startup icons for the recognition of code beginnings to get integrated.

Further is a use of information and error correction symbols possible.

The Advantages of compaction by alternating inversion of the bars were also already presented.

following becomes a preferred embodiment the sorting of mailpieces using the determined sorting information, from which a barcode is generated, shown.

The consignments placed in the branches and at the mailboxes are collected by the postal company and sent to a sorting center in the region of the place of dispatch. In this so-called dispatch center (BZA), the shipments are sorted according to destination regions and in particular according to the entry letter centers (BZE) located in the areas of the destination regions of the mailpieces. This sorting takes place in a sorting process with one or more passes / runs in a sorting device, which detects the delivery addresses applied to the mailpieces. From the collected data, a barcode is generated and applied to the shipments. With the help of this barcode, the items are fed to the terminals of the sorting devices and dropped there into containers which are associated with an applied to the mailing delivery address. Shipments that can not be sorted by machine, eg. B., because the applied barcode can not be read, are sorted manually. After the BZA sorting, the shipments are transported to the inbound letter center. There, the items are sorted by sorting on delivery districts and in a preferred embodiment - by gear sequence. If the code can not be read, a manual sorting is done here as well. After that, the consignments are fed to delivery points and PO Boxes and arrive at the delivery or in the mailboxes.

to Compliance with the quality specification E + 1 (delivery of goods the day after delivery) and to ensure economic processing, it is important that the machining process throughout the machining process can be ensured and that a sorting depth is achieved, the preferably includes a sorting by order of the shipments, because this is the time-consuming manual sorting by gear sequence canceled by the deliverer. The deliverer needed there is no time left for that Preparation of delivery, but can take place immediately after takeover of shipments begin with the delivery process.

To For this purpose it is necessary that the barcode applied - in the specified narrow time window, which is available for processing, - recorded as well as completely and can be read correctly. Can not read the code so the shipment is manual with corresponding time and cost to sort. If the code can not be read completely, it will not the desired one Sorting depth achieved, d. H. it then z. B. a sequence sorting. If the code is read but misinterpreted, this will result to a misdirection of the consignment concerned, the additional Sorting and transport costs and late delivery has the consequence.

These embodiment of the code bears help in that - at Adherence to the given narrow time window - the reading rate increased and one full and correct interpretation of the delivery information. In order to the invention makes a significant contribution to quality and efficiency when processing mail in the letter centers.

Claims (11)

a) method for sorting mailpieces, b) wherein the mailpieces are sorted taking into account data signals of a machine-readable barcode located on the mailpieces, which is formed from a number of bars, c) wherein, when printing one of the mailpieces, the machine-readable code is generated e) wherein at least one of the data tracks contains sorting information; f) wherein a reading unit generates a graphical image of the machine-readable code, g) wherein the graphical image h) wherein the stored image is evaluated so that differences between signal intensities are used for determining clock signals of the reference clock, i) wherein the data signals are evaluated by assignment to the clock signals of the determined reference clock in that thereby the sorting information is determined and j) wherein sorting of the mailpieces takes place in dependence on the sorting information, characterized in that k) that the machine-readable code is inverted characterwise or barwise, such that in each case a non-inverted bar is followed by an inverted bar so that a pitch between the bars can be reduced and a clock for reading systems is maintained. Method according to claim 1, characterized in that that the machine-readable code is generated so that it at least contains two data tracks, wherein one of the data tracks is the reference clock and one of the others Data tracks contains information about the sorting of the mailpieces. Method according to claim 2, characterized in that that the machine-readable code is generated so that it at least contains three data tracks Method according to claim 3, characterized that the machine-readable code is generated so that the data track, which contains the reference clock, is between other data tracks. Method according to one or both of claims 3 or 4, characterized in that the machine-readable code is generated will cause the data track containing the reference clock to become essentially located midway between the other data tracks. Method according to one or more of the preceding Claims, characterized in that the generation of the machine-readable Codes are made using interleaving. Method according to one or more of the preceding Claims, characterized in that in the machine-readable code additional Information is introduced, which allow an error correction. Method according to claim 7, characterized in that that a Reed-Solomon error correction method is used. Method according to claim 7 or 8, characterized that block codes are used. Method according to one of claims 7 to 9, characterized that convolutional codes are used. Method according to one or more of the preceding Claims, characterized in that the machine-readable code and a Address information in plain text in a printing process on the mailpiece be applied.