DE19504912C2 - Method and device for recognizing barcode symbols - Google Patents

Description

The invention relates to a method and a device for recognizing Barcode symbols according to the preambles of claims 1 and 10.

Devices for recognizing barcode symbols are different Embodiments in industrial use for years. In the simplest case periodically a barcode symbol with the transmitted light beam of the device scanned and decoded in the device. This is the case with each scan Barcode symbol completely captured by the transmitted light beam, so that in the emp catch signal, which is registered in the device, the complete informa tion via the barcode symbol.

However, the barcode symbol to be detected is oriented so that it is from the Transmitting light beam is only partially detected in the individual scans, so ent each received signal only holds the information of a fragment of the barcode. In order to decode the entire barcode symbol from this partial information are the signal sequences that contain the information about each fragment contain, assembled in a suitable manner so that there is a signal sequence that contains all the information of the barcode symbol. This signal sequence can be decoded in a known manner in the device.

Such a device is known from US 4 916 297. For every The analog receive signals are digitized and sampled in one Computer unit decoded. For this is the pattern of the barcode symbol specified and stored in the computer unit as a string from. This Character string is created with the character string obtained in the individual scans compared part of the barcode symbol in the computer unit. If one of the characters of the barcode symbol is recognized, the corresponding one Characters are stored in a predetermined memory location of the computing unit. Certain characters of the barcode symbol are used for this seen memory locations assigned. The individual scans are  evaluated until the complete string of the barcode symbol is recognized.

An omnidirectional barcode reader is known from EP 0 449 634 A2, which is a sensor array for recording a two-dimensional image of a Has barcode symbol. The image information is in the form of a two-dim sional field of pixels. To evaluate the image information is to first the rough alignment of a barcode is determined. Then the barco de symbol scanned several times in the roughly determined longitudinal direction, d. H. the Grayscale levels of the pixels determined as a function of the location. From the individual Locus curves of the grayscale levels become the cross-correlation function numerically NEN formed, on the one hand, the exact orientation of the barcode symbol can be determined and on the other hand the structure of the barcode sym  bols can be recognized. This method of recognizing barcode sym bols is extremely complex and requires a lot of effort in computing power.

EP 0 436 072 A2 describes a method for decoding barcode symbols bolen from signal sequences of several samples that only the barcode symbol incomplete, known. The barcode symbols each consist of a sequence of characters, for example letters, which in turn consist of a Sequence of light and dark line elements are constructed. The barcode sym bol is periodically in front of the device's transmitted light beam given direction scanned, the transmitted light beam over the Barco de symbol moves so that the barcode symbol with a number of Scans is fully captured.

The analog received signals are evaluated for each scan decoded. Thus, with each scan, a string of a portion of the Get the barcode symbol. Then two strings of on consecutive scans against each other relatively as long as each other moved until as large a part of the strings as possible matches. The strings so matched will result in one put the string together. After that, this string with the next following string concatenated in the same way. This method is repeated until the barcode symbol consists of the individual characters follow is reconstructed.

A disadvantage of this method is that the scanning direction is almost the same as the The longitudinal axis of the barcode symbol must match, so that each scan a sufficiently large number of characters is detected. Otherwise there is Risk of incorrect decoding. In addition, a secure chaining of the individual strings are not guaranteed because the number of overlapping Character is very small.  

EP 0 043 124 A2 describes a further method for decoding Barco de-symbols from several scans that make the barcode symbol incomplete capture, known. For each sample, the received signals are converted to the Widths of the detected line elements determined. The classification of the stroke elements is created by forming quotients from two values of neighboring values Line elements. A sequence of numbers is thus obtained from each scan, the corresponds to a sequence of line elements with certain widths. The Re The barcode symbol is constructed by putting the numbers together follow, which are obtained from the individual samples.

WO 92/05516 describes a method for decoding barcode symbols bolen known in which the widths of line elements of parts of a Barcode symbols, which are detected during the scans, as code fragments get saved. These code fragments that exist as sequences of numbers are shifted against each other until the maximum match the code fragments is achieved. Then the code fragments are put together set. This process is repeated until the entire bar code symbol is fully captured. Disadvantageous when assembling from Sequences of existing code fragments is that when feh lerable code the correspondence of the individual fragments is not error-free can be accomplished. For example, a code has a wide bright one Line element on which a dark color spot has been sprayed on, so according to this method for a scan that does not hit this spot correctly determined the length of the line element. However, the transmitted light beam strikes in the subsequent scan for this spot, instead of one bright broad line element two narrow light line elements with approximately half width, separated by a narrow dark line element. Like this with are incorrectly ver in the successive scans different stroke patterns found.

The invention has for its object a method and an apparatus of the type mentioned in such a way that a barcode symbol consists of several  incomplete scans safely and with the least possible calculation effort can be decoded.

To solve this problem, the features of claims 1 and 10 are provided seen. Advantageous embodiments and expedient developments of the Invention are described in claims 2-9 and 11-14.

According to the bar code symbols are scanned repeatedly, with fragments of a barcode symbol are recorded in the individual scans and in be stored in a memory. For the reconstruction of a barcode sym bols the fragments of the barcode symbol are shifted against each other and checked for consistency.

This takes place on the basis of binary signals in the form of bit sequences. These binary signals are obtained using an analog-digital converter, to which the analog received signals are fed. One from the device Scanned dark or light bar element of a barcode is at the exit of the analog-to-digital converter as a sequence of "0" or "1" values certain length. The bit sequences forming the individual fragments become shifted against each other so that this is the greatest possible match exhibit. For this purpose, the cross-correlation function of the two signal sequences educated. Then the individual bit sequences are put together until finally the complete pattern of the barcode symbol can be decoded. The fragments of the barcode symbols are processed in one Correlator and a recombiner.

The advantage of the method according to the invention is that in the correction individual bits can be folded. In comparison to Convolution of an analog or a digitized signal of a given amplitude this means a considerable saving in computing time.

Another advantage of the method according to the invention is that  the correlation and subsequent recombination of the bit sequences also then error-free decoding if the barcode symbol has minor errors places such. B. color defects.

Finally, it is advantageous in the processing of binary signal sequences that the individual signal sequences only a relatively small part of the barcode symbol  must contain in order to ensure error-free decoding. This means that the angle between the scanning direction of the transmitted light beam and the longitudinal axis of the bar code symbol can be relatively large.

The invention is explained below with reference to the drawings. It demonstrate:

Fig. 1 shows a basic structure of the device for recognizing barcode symbols

Fig. 2 multiple scanning of a barcode symbol

Fig. 3 evaluation unit of the device according to Fig. 1

• a) First embodiment
• b) second embodiment

Fig. 4 Overview of occurring in the evaluation waveforms:

• a) Line pattern of a barcode symbol
• b) analog received signal
• c) binary signal sequence
• d) quantized signal
• e) classified signal
• f) decoded signal

Fig. 5 example of a recombination of binary signal sequences.

In Fig. 1 the basic structure of an optoelectronic device 1 is shown for detecting bar code symbols. 2 The bar code symbols 2 hen best of a series of black and white line elements 2 a, b DEFINE ter length and width.

The optoelectronic device 1 consists essentially of a sensor element 3 , a receiving element 4 and an evaluation unit 5 . The transmission element 3 consists of a transmitter 6 , preferably a laser diode, and of a transmission optics 7 connected upstream of the transmitter 6 for focusing the transmission light beam 8 . The focused transmitted light beam 8 is deflected from a steering unit 9 , which in the present exemplary embodiment is formed by a rotating polygon mirror wheel, and is guided via the barcode symbol 2 to be detected. The axis of rotation of the polygon mirror wheel is arranged perpendicular to the equatorial plane of the polygon mirror wheel shown in FIG. 1.

The received from the barcode symbol 2 reflected light 10 is guided over the polygon mirror wheel to the receiving element 4 . The receiving element 4 consists of a photodiode 11 , in which the receiving light 10 is converted into an electrical Emp signal, and one of these amplifiers 12th To improve the detection sensitivity, the receiving element 4 is preceded by receiving optics 13 .

The received signal at the output of the receiving element 4 is fed to the evaluation unit 5 and stored as a function of the position of the transmitted light beam 8 . The position of the transmitted light beam 8 can be obtained directly from the received signal if the sampling rate is known. Then there is an unambiguous assignment between the temporal sequence of the received signal and the position of the transmitted light beam 8 .

In FIG. 2, an example of a speed v moving in a predetermined direction with a Ge bar code symbol 2 shown. The barcode symbol 2 has a color defect in the middle. For example, the barcode symbol 2 can be applied to a package that is transported on an assembly line. The barcode symbol 2 is periodically scanned by the optoelectronic device 1 at an angle α to the longitudinal axis. Due to the movement of the bar code symbol 2 , the transmitted light beam 8 is guided successively over the bar code symbol 2 . The individual lines A identify the successive scans. The start is marked with "START" and the end point of a scan with "END". With each scan, the transmitted light beam 8 sweeps over an area Δx.

As can be seen from FIG. 2, bar code symbol 2 is mostly only partially detected with one scan. The received signals, which are registered during a scan, thus only contain fragments of the bar code symbol 2 . The evaluation unit 5 is designed such that, despite the partial scanning and despite the color error in the bar pattern of the barcode symbol 2, decoding of the barcode symbol 2 is possible.

Fig. 3a, b show two embodiments of the evaluation units 5 according to the invention. The signal sequences that are generated or processed in the individual components of the evaluation units 5 are shown in FIG. 4.

The inventive method will first be explained with reference to the exemplary embodiment according to FIG. 3a. The standing at the output of the receiving element 4 analog received signal ( Fig. 4b) is converted in an analog-digital converter 14 in a binary signal sequence. An example of such a signal sequence is shown in Fig. 4c. Bright line elements 2 b of the barcode symbol correspond to a sequence of "1" values, the lengths of the sequences being proportional to the width of the line elements 2 b. Dark line elements 2 a correspond to a sequence of "0" values. The binary signal sequences that are present in the samples n, n + 1, n + 2. . . are detected, are stored in a scan memory 15 sorted by the number of the scan.

The bit sequences of successive scans are fed to a correlator 16 from this scan memory 15 . The cross correlation function of the two bit sequences is determined there. Since the signal sequences contain individual bits, no multi-plications need to be carried out to form the cross-correlation function. Only gate links and counter functions are required to form the cross-correlation function. The formation of the cross correlation can thus take place without great computation effort. With the cross correlation function, the position of the best match of the two bit sequences and thus also the relative displacement of the bit sequences as a result of the movement of the bar code symbol 2 are determined.

The maximum of the cross correlation function determines the relative position of the Bit strings for which the maximum match is achieved.

If several maxima occur, the most likely maximum is selected. This most probable maximum can be determined by continuously determining and storing the mean shift of two successive bit sequences. At a constant speed of the barcode symbol 2 , the maximum of the cross-correlation function is the most probable which comes closest to the mean shift of the bit sequences.

The bit sequences of the individual scans brought into the position of the best match by means of the correlator 16 are fed to a recombiner 17 . The recombiner 17 consists essentially of a computer unit in which bit strings of individual samples are put together. The mode of operation of the recombiner 17 is illustrated in FIG. 5.

In the recombiner 17 , the bit sequences of three successive scans n, n + 1, n + 2 are compared bit by bit in the position of their best match. In the present example, the position of the best match is achieved when a neighboring bit sequence is shifted by one bit at a time.

The recombination takes place by a majority decision for each bit position p. The value "1" is obtained for each of the three bit sequences at bit position p ₁ . Accordingly, the value "1" is adopted as the result in the resulting bit sequence E. With the majority decision method, errors in the bit sequences, which are based, for example, on local disturbances of the barcode symbol 2 , can be efficiently eliminated. Such faulty bits are underlined in the bit sequence n + 1 at location p ₂ and in the bit sequence n + 2 at location p ₃ . Since the value "0" is contained at the location p ₂ in the bit sequences n, n + 2 and the value "0" is also contained at the location p ₃ in the bit sequences n, n + 1, p ₂ and p ₃ in the resulting bit sequence E assume the correct values "0", which consequently ignores errors.

The values in the edge zone of a single bit sequence in which there are no overlaps with other bit sequences, such as e.g. B. at the locations p ₆ , p ₇ , are taken directly into the resulting bit sequence E.

At locations p ₄ , p ₅ , only the bit sequences n + 1 and n + 2 overlap. If the values in the two bit sequences match, they are adopted directly into the resulting bit sequence E.

However, if the bit values differ at the locations p ₄ or p ₅ , a value determined at random is entered in the corresponding position of the resulting bit sequence. Since the recombination usually has to be carried out for a large number of samples, such random quantities are eliminated by averaging.

After recombination has taken place, the resulting bit sequence is stored in a result memory 18 .

Then the recombination is repeated with the bit sequences of the subsequent scans, the resulting bit sequence E of the previous recombination being read into the recombiner 17 and used as the bit sequence for the current recombination. This process is repeated continuously.

The resulting bit sequences E stored in the result memory 18 thus continue to grow until finally the entire information of a bar code symbol 2 is present in the last resulting bit sequence E.

Compared to other methods of recombination, such as. B. Averaging This method requires very little computing time. Because the recombina bit by bit, there is almost no loss of information.

The resulting bit sequences E are continuously read into a quantizer 19 . There the binary signal sequence ( FIG. 4c) is converted into a sequence of numbers as shown in FIG. 4d. The numbers forming the individual elements represent the lengths of the bit sequences with the values "0" or "1" and thus correspond to the widths of the individual line elements 2 a, b of the bar code symbol 2 . These sequences of numbers are stored in an element memory 20 , which is preferably designed as a FIFO ("first in first out") memory, and are read from there into a decoder 21 .

In the decoder 21 the sequences of numbers are examined to determine whether the individual elements are dark or light elements of different widths, ie whether a narrow line "s", a wide line "S", a narrow gap "l" or one there is a wide gap "L". This is shown in Fig. 4e. Subsequently, this sequence of bar patterns representing the barcode symbol 2 is decoded, ie individual elements become certain characters Z1, Z2, Z3. . . too ordered. The characters can be formed by letters or numbers ( Fig. 4f).

The decoding is carried out continuously after the individual scans. Decoding is complete when a complete barcode symbol 2 has been identified. The barcode symbols 2 have a defined start and end zone at their ends. As soon as these are recognized, the barcode symbol 2 is regarded as identified. For this purpose, the various patterns of the barcode symbols 2 are expediently stored in the decoder 21 as a reference pattern.

The method described is particularly suitable if only one and the same barcode symbol 2 is detected in each of the individual scans. However, two different barcode symbols 2 can also be scanned in one scan.

For such applications, the evaluation unit 5 has means by means of which the bit sequences for the individual fragments can be extracted, sorted and then decoded separately. An embodiment of such an evaluation unit 5 is shown in Fig. 3b.

The analog received signal present at the receiving element 4 is in turn converted into the binary signal sequence in the analog-digital converter 14 . The binary signal sequence is converted into a number sequence in the quantizer 19 ′ and stored in an element memory 20 ′ designed as a FIFO memory. From there, the number sequences are read into an extractor 22 . The ex tractor 22 is formed by a software module. In the extractor 22 , the partial sequences of a number sequence that belong to different bar code symbols 2 are separated. In the extractor 22 , for example, the end or start zones of a barcode symbol 2 are identified by recognizing a corresponding partial sequence of a sequence of numbers. In addition, two partial sequences for two barcode symbols 2 can be separated and differentiated in the extractor 22 by detecting a quiet zone between two number sequences corresponding to fragments of the barcode symbols 2 , ie a longer partial sequence with the same bit values "0" or "1".

The individual partial sequences are read into an administrator 23 . This is also a software module that combines partial sequences for the same barcode symbol 2 . For this purpose, the position of the transmitted light steel 8 is evaluated. At a known speed v at which the bar code symbol 2 is moved, the individual sequences of numbers can be assigned to specific bar code symbols 2 . The movement of the barcode symbol 2 can be determined from the mean shift of the individual sequences of numbers.

The output of the administrator 23 is fed to the correlator 16 and the recombiner 17 connected downstream. In the administrator 23 , the partial sequences present as number sequences are converted into binary signal sequences for a specific barcode symbol 2 and are successively inserted into the recombiner 17 . In this way it is ensured that when combining the binary signal sequences registered in different scans, partial sequences which correspond to the same bar code symbols 2 are processed.

The output of the extractor 22 is passed to a converter 24 , which converts the number sequences back into binary signal sequences, which are stored in a downstream scan memory 15 . A result memory 18 is connected to the correlator 16 and the recombiner 17 in the same way as in the exemplary embodiment according to FIG. 3a. The recombiner 17 is in turn connected to the quantizer 19 , which is led to the decoder 21 via the element memory 20 .

The operation of the scan memory 15 , the result memory 18 , the correlator 16 , the recombiner 17 , the quantizer 19 , the element memory 20 and the decoder 21 is the same as in the embodiment according to FIG. 3a.

Claims (14)

1. A method for recognizing bar code symbols consisting of a sequence of light and dark bar elements, wherein a bar code symbol is scanned repeatedly and thereby fragments of the bar code symbol are captured and stored, and wherein the fragments of the bar code symbol against each other are shifted and checked for their correspondence and put together, the barcode symbols ( 2 ) being in the form of bit sequences forming binary signals and the bits being shifted relative to one another, thereby forming the cross-correlation function, the maximum of which corresponds to the position of the best match of the two bit sequences defined. 2. The method according to claim 1, characterized in that a bar code symbol ( 2 ) is scanned repeatedly in a predetermined direction. 3. The method according to claim 1 or 2, characterized in that at Auf several maxima of the cross-correlation function emerge the maximum whichever is the most likely shift of the two Corresponds to bit sequences. 4. The method according to claim 3, characterized in that the probability Most shift by averaging the previously determined ver shifts between two successive bit sequences is determined. 5. The method according to any one of claims 1-4, characterized in that for assembling the fragments of a bar code symbol ( 2 ) little least three consecutive bit sequences in the positions of their best matches are compared bit by bit and as a result of the bit value occurring in each case in a resulting bit sequence (E) is adopted. 6. The method according to claim 5, characterized in that the together setting the fragments is repeated continuously, each time the previous The resulting bit sequence (E) determined, at least with the two successive ones the bit sequences are compared. 7. The method according to claim 5 or 6, characterized in that the over the remaining bit sequences protruding parts of a bit sequence into the resul bit sequence (E). 8. The method according to any one of claims 5-7, characterized in that the assembly of the fragments is continued until the complete bar code symbol ( 2 ) is present in the resulting bit sequence (E). 9. The method according to any one of claims 1-8, characterized in that the signal sequences determined during a scan are examined to determine whether fragments of several bar code symbols ( 2 ) are detected, and that the bit sequences for the fragments of the various NEN from the signal sequences Barcode symbols ( 2 ) are extracted, after which the fragments of the different barcode symbols ( 2 ) are decoded separately. 10. Device for recognizing bar code symbols with a deflection unit which periodically deflects a transmitted light beam in a predetermined direction and repeatedly leads over the bar code symbols to be detected, the received analog signals received in a receiving element in an analog-to-digital converter binary signals forming a bit sequence are converted and stored in a memory, and the bits follow two successive scans are fed to a correlator ( 16 ) and successive bit sequences which are matched in the correlator ( 16 ) are obtained in a recombiner ( 17 ) assembled and stored in a result memory ( 20 ). 11. The device according to claim 10, characterized in that in the result memory ( 20 ) stored bit sequences as intermediate results are supplied to the recombiner ( 17 ). 12. The apparatus of claim 10 or 11, characterized in that the bit sequence forming the end result of the recombination of the result memory ( 18 ) a quantizer ( 19 ) for determining the width of the line elements ( 2 a, b) of a bar code symbol ( 2nd ) is supplied and then transferred to a decoder ( 21 ) for the classification of the bar code symbol ( 2 ). 13. Device according to one of claims 10-12, characterized in that the correlator (16), an extractor (22) and an administrator (23) is connected upstream, wherein the extractor (22) of different fragments in a bit sequence barcode symbols (2 ) are separated and sorted in Admini stator (23) and are stored, and wherein the administrator (23) in each case a bar code symbol (2) transmits fragments in the form of bit sequences to the correlator (16) for evaluation. 14. The apparatus according to claim 13, characterized in that the fragments of different barcode symbols ( 2 ) are identified in the extractor ( 22 ) on the basis of their defined end and start zones, their bit pattern and on the basis of their position within a bit sequence.