JPH10261043A - Decoding method, decoder, and bar code processing system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable a decorder to safely and easily decode at high speed, by decoding an encoded character string while combining the method of maximum likelihood estimation and the method of critical distance decoding. SOLUTION: A character string generating part 101 reads a postal code described in the postal code description column of mail and an address indication number through a scanner, performs character recognizing processing based on a provided image signal and extracts character information. At an error correct code (parity) adding part 102, a parity corresponding to a character string outputted from the character string generating part 101 is added and outputted to a bar coding and printing part 103. When the number of erroneous words in the code is less than a number based on the number of words in the parity added to the character string, the code is decoded by performing error correction based on the parity and when the number of erroneous words in the code is more than the number based on the parity added to the character string, the code is decoded by performing the error correction through maximum likelihood estimation so that the encoded character string can be decoded while combining the method of maximum likelihood estimation and the method of critical distance decoding.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a decoding method and a decoding device for a multiplex coding system. Also, the present invention provides a barcode encoding and decoding process for printing a postal code such as a sealed letter, postcard, and the like, and representing a postal code and a numeric part of an address, which are used for classification of each postal mail by destination. And a barcode processing system that performs

[0002]

2. Description of the Related Art In a system for decoding and decoding a given character string, a so-called multiple coding system using a plurality of codes instead of a single code may be used.
By using the multiplex coding method, the decoding algorithm can be simplified by combining simple codes,
High performance can be obtained. When the multiplicity is i,
First, the lowermost i-th stage is encoded, the higher-order encoded symbols are sequentially encoded, and finally the first-stage encoding is performed.

[0003] Decoding in a multiplex coding system is usually performed in the reverse order of coding. For example, when i = 2,
First stage decoding is performed first, and then second stage decoding is performed. At this time, in order to ensure the performance, it is desired to use the decoding information of the first stage in the second stage, but this is generally difficult (Refer to Imai et al., "Decoding method of double encoding method", IEICE A), '82 .12, pp. 1254-1261).

[0004] When the limit distance decoding method is used for the first-stage decoding, the decoded information cannot be used much for the second-stage decoding, so that information loss occurs, so to speak, and the original capability of the double encoding method is not sufficient. There is a problem that it cannot be demonstrated.

There is also a method of using the maximum likelihood estimation method for the first stage decoding and using the result as it is, but the processing tends to be complicated because there are too many candidates for decoding. A barcode processing system used for mail processing is an example of a system that performs encoding and decoding of a character string using a multiplex encoding method.

In recent years, in postal matter processing in which a postal code and an address on a postal matter are read and the postal matter is classified according to the destination based on the information, a new postal code (destination place name information, etc.) different from the present one, The address display number (destination lot number information) is printed as a barcode according to the special barcode system on the address display surface of postcards, postcards, etc., and the barcode is read at the sending office or the delivery office and classified according to the destination of the mail. It has been studied to increase the efficiency of business by automating the sorting of the items.

[0007] It is planned to print a barcode representing a postal code or a numerical part of a destination name on a mail using a special ink such as a fluorescent ink. In the encoding of a postal code consisting of seven digits, the second stage is a Reed-Solomon code (hereinafter referred to as RS).
The first stage uses a constant weight code of weight 3 with a 6-bit configuration as shown in FIG. 2 for each information digit and each 3-digit parity of the RS code.

When reading such a bar code, a one-way error such as dropping of a bar at the time of reading the code often occurs due to the overlap of the address character on the mail and the bar code. In addition, a one-way error of adding a bar occurs due to another cause such as rubbing or scattering of ink.

In a bar code processing system used for mail processing, a constant weight code that generates this directional error is used, so that when decoding, the original character string is RS encoded and the original character string is encoded. Error correction equal to the number of parities added to the character string can be performed, and by utilizing this feature, the original character string is restored from the bar code.

[0010]

In a bar code decoding process for decoding a character string encoded using such a double encoding method, the accuracy of decoding, that is, the accuracy of error correction, is reduced. Improvement is an important issue for preventing code information loss and improving reliability.

Furthermore, in view of processing a large amount of mail, higher speed processing is required. Therefore, the present invention decodes a character string coded by combining the maximum estimating method and the limit distance decoding method to obtain a correctable error number corresponding to the number of parities guaranteed by the limit distance decoding method. It is an object of the present invention to provide a decoding method and a decoding device capable of preventing loss of information of a code even in a case where the number of errors exceeds the above, and performing safe, easy and high-speed decoding, and a barcode processing system using the same.

[0012]

According to a decoding method of the present invention, in a decoding method for decoding a code obtained by adding an error correction code to a character string and decoding the code, the number of error words of the code is added to the character string. When the number is smaller than the number based on the number of words of the added error correction code, the code is decoded by performing error correction based on the error correction code, and the number of error words of the code is the number of words of the error correction code added to the character string. When the number is larger than the number based on the minimum distance, the maximum distance estimation is performed, error correction is performed, and the code is decoded. Even when there are many errors exceeding the number of correctable errors corresponding to the number of parities guaranteed by the decoding method, loss of code information can be prevented, and decoding can be performed safely, easily and at high speed.

The decoding method according to the present invention is a decoding method for decoding a code obtained by adding an error correction code to a character string and decoding the code. When the number is smaller than the number based on the number of words of the correction code, the error is corrected based on the error correction code to decode the code, and the number of error words of the code is larger than the number based on the number of words of the error correction code added to the character string. When generating an error correction code for each of the character string candidates of the original character string extracted based on the characteristics of the code, extracting the error correction code for each of the generated character string candidates and extracting the error correction code based on the characteristics of the code By comparing the error correction code candidates of the original character string and detecting the original character string based on the comparison result, the character string encoded by combining the foremost estimation method and the limit distance decoding method Restore By prevent information loss of a code even if an error with many such exceed the correctable number of errors corresponding to guaranteed number of parity bounded distance decoding method, safely and easily yet can be decoded at high speed.

[0014] The decoding apparatus according to the present invention is a decoding apparatus for decoding a code obtained by adding an error correction code to a character string and encoding the error correction code in which the number of error words of the code is added to the character string. First decoding means for performing error correction based on the error correction code and decoding the code when the number is smaller than the number based on the number of words of the error correction code; When the number is larger than the number based on the number of words, second decoding means for estimating the error and performing error correction and decoding the code is provided. By decoding the character string, even if there are many errors that exceed the number of correctable errors corresponding to the number of parities guaranteed by the marginal distance decoding method, loss of code information can be prevented and safe and easy. It can be decoded at high speed.

Further, the decoding device of the present invention decodes a code obtained by adding an error correction code to a character string and decoding the code. When the number is smaller than the number based on the number of words of the correction code, a decoding unit that performs error correction based on the error correction code and decodes the code, and the number of error words of the code is based on the number of words of the error correction code added to the character string. Generating means for generating an error correction code for each of the character string candidates of the original character string extracted based on the characteristics of the code when the number is larger than the number, and an error correction code for each of the character string candidates generated by the generation means. And detecting means for comparing error correction code candidates of the original character string extracted based on the characteristics of the code and detecting the original character string based on the comparison result. Push Decodes a character string coded using a combination of the limit distance decoding method and the limit distance decoding method, so that even if there are many errors that exceed the number of correctable errors corresponding to the number of parity guaranteed by the limit distance decoding method, Can be safely and easily and quickly decoded.

A bar code processing system according to the present invention converts a code obtained by adding an error correction code to a character string into a bar code, prints the bar code on a predetermined print target, and prints the bar code from the print target. In a bar code processing system for reading a bar code and restoring an original character string, a conversion means for converting a bar code read from the printing object into a code word, and the number of error words of the code word converted by the conversion means When the number is smaller than the number based on the number of words of the error correction code added to the character string, first decoding means for performing error correction based on the error correction code and decoding the code, When the number of erroneous code words is larger than the number based on the number of error correction codes added to the character string, second decoding means for estimating the error and correcting the error and decoding the code, With the provision, the number of correctable errors corresponding to the number of parities guaranteed by the limit distance decoding method is exceeded by decoding the encoded character string by combining the maximum length estimation method and the limit distance decoding method. Even in the case of such a large number of errors, loss of code information is prevented, and decoding can be performed safely, easily, and at high speed.

Further, the bar code processing system of the present invention converts a code obtained by adding an error correction code to a character string into a bar code, prints the bar code on a predetermined print target, and prints the print target. In a bar code processing system for reading a bar code from an object and restoring an original character string, a conversion means for converting the bar code read from the printing object into a code, and the number of error words of the code converted by the conversion means When the number is smaller than the number based on the number of words of the error correction code added to the character string, a decoding unit that performs error correction based on the error correction code to decode the code, and an error of the code converted by the conversion unit. When the number of words is larger than the number based on the number of words of the error correction code added to the character string, each error correction of the character string candidates of the original character string extracted based on the characteristics of the code is performed. Generating means for generating a code, and an error correction code for each character string candidates generated by the generating means,
Detecting means for comparing error correction code candidates of the original character string extracted based on the characteristics of the code, and detecting the original character string based on the comparison result. By decoding a coded character string by combining with the limit distance decoding method, even if there are many errors that exceed the number of correctable errors corresponding to the number of parities guaranteed by the limit distance decoding method, the code Information loss can be prevented, and decryption can be performed safely, easily, and at high speed.

[0018]

Embodiments of the present invention will be described below with reference to the drawings. In the following description,
For example, the encoding / decoding method of the present invention will be described by taking a barcode processing system used for mail processing as an example.

An encoding device of a barcode processing system used for mail processing is configured, for example, as shown in FIG. In FIG. 1, for example, a character string generation unit 101 scans a postal code entered in a postal code entry field of a postal matter (a barcode print target) such as a postcard or a sealed letter, and an address display number in an addressee with a scanner. Character recognition processing is performed based on the read and obtained image signals to extract character information. Then, from the extracted character information, if it is an address display number, the apartment A building 405 of 1-3-2, 1-chome
The room is expanded (changed) according to a system convention that expresses it as "1-2-3-A-405".
Is extended (changed) with a code word that replaces "A" with "C1 + 0". In the case of a postal code, for example, "1"
The postal code "23-4567" is "1234567"
Is extended (changed) on the system.

In the parity adding section 102, a parity corresponding to the character string output from the character string generating section 101 is added, and is output to the bar coding and printing section 103. The bar-coding and printing unit 103 converts the input character string into each code word of a constant weight code on a character basis (see FIG. 2), converts the code string into a bar code pattern, and according to the bar code pattern, A bar code representing a zip code and an address display number is printed at a predetermined position on the address display surface of the mail, together with a bar for detecting start / stop of the bar code, for example, using fluorescent ink.

As described above, in the encoding of a postal code consisting of, for example, seven digits, printed as a bar code on a mail, the second stage is a Reed-Solomon code (hereinafter referred to as an RS code), and the first stage is an information code. Each digit and each 3-digit parity of the RS code have a 6-bit configuration as shown in FIG.
And a multiplexing coding scheme with a multiplicity of “2” using a constant weighting code.

Here, the constant weight code will be described with reference to FIG. FIG. 3 shows a 2-bit 4-state (up, down, up,
For one code composed of three bars (under, no),
This is an example of allocation in the case where the number of code words is 15, to which one character is allocated. In FIG. 3, the barcodes are shown corresponding to barcodes with a constant weight code of weight “3”.

When the error has directionality, the fixed weight code is used to determine the error position, that is, in the form of so-called erasure error correction (erasure correction). There is an advantage that the character error can be corrected.

Here, the character string to be encoded is a new postal code consisting of seven-character numbers, and as shown in FIG. 2, a parity of three characters is added to the new postal code of seven characters to perform encoding. Shall be performed. 2 bits 4
Since one character is represented by three bars in the state, 1
The character is composed of 6 bits, and the parity is also composed of 6 bits.

If erasure correction is performed by using a constant weight code, one character error correction can be performed with one character parity. Next, an encoding method of the barcode processing system used for mail processing will be described.

When reading a bar code, a one-way error such as dropping of a bar often occurs at the time of reading the code due to the overlap of the address character on the mail and the bar code. In addition, a one-way error of adding a bar occurs due to another cause such as rubbing or scattering of ink.

When such directional errors occur, if the constituent characters of the barcode are composed of a constant weight code, an erasure error (as long as the error position can be specified) will occur as the error form of the first stage. In the second stage, erasure correction (correction of erasure error) can be performed within the range of the number of error characters corresponding to the number of parities guaranteed by the limit distance decoding method. This is a correction method according to the conventional method, assuming that error character restoration target candidates are all code words such as 10 digits.

Generally, a code for detecting a one-way error is AU
ED code (all-unidirectional-e
rror-detecting code). As an example, the present embodiment will be described by taking a constant weight code as an example.

In general, for example, in a decoding method of an encoding method with a multiplicity of “2”, likelihood (the likelihood of an input as viewed from the output) is added to the candidates for restoration in the first stage, and It is also used for decryption. Further improvements include:
Processing can be simplified by decoding only with candidates having a high likelihood. In this case, it is more effective to select a code that produces a large likelihood difference.

For example, in the case of a character in which one of the three bars is missing (for example, when one bit of the code word is missing), one of the bars is missing and any of the candidates for the missing position is one of four. It is assumed that the likelihood is larger than that in the case where two are dropped and one is added. In this case, only a maximum of four characters generated from the four drop-out position candidates remain as candidates with high likelihood. Specifically, when the first bar of the symbol “0” in FIG. 2 is dropped, only the character candidates “0” or “C2” remain, and when the second bar is dropped, the character candidates are “0” and “6”. , "8" and "9" remain.

As described above, the constant weight code can be said to be a code capable of performing decoding suitable for the maximum likelihood estimation method because the original character string candidates can be narrowed down. Here, the decoding method by the maximum likelihood estimation method is a method of recognizing and decoding the most probable input from the viewpoint of output.
For example, when a constant weight code having a weight of “3” is used, a candidate of a code word having an error is limited based on the directionality of an error (dropping or addition of bits of a code word) and the feature of a weight of “3” ( At that time, the degree of brilliance may be obtained), and a character string that is most likely to be the original character string among the limited candidates is output as a decoding result.

The above is a case where the decoding of the upper stage is performed only by the maximum likelihood estimation method. When the error is small, for example, the erasure error which can be easily counted is the number of parity (here,
In the case of 3 digits or less, it is possible to perform more reliable processing by using the marginal distance decoding method and in other cases by using the maximum likelihood estimation method. This is because when there are few errors, the limit distance decoding method guarantees unique decoding.

The present invention provides a decoding method of the double encoding system from such a viewpoint. In a system where there are many errors other than erasure errors, the maximum likelihood estimation method is used if the number of erasure errors is larger than the number obtained by adding an appropriate positive / negative bias to the number of parities, and the limit distance decoding method is used for other cases. I just need.

FIG. 4 shows an example of the configuration of a decoding apparatus used in a bar code processing system or the like, for example, using the decoding method of the duplex decoding method of the present invention. A bar code reading unit (not shown) reads the bar code printed on the mail using a scanner, and the error number determination unit 1 converts the bar code information into a code for each character based on the read bar code information. When a dropout of a bar (dropout of a bit of a code word) is recognized, the dropped bar (bit) is grasped (for example, the position and the number of the dropped bar are grasped). The original character string candidates are extracted based on the characteristics of (1). At this time,
When the number of errors is large, the character string candidate and likelihood information 12 are input to the decoding unit 2 using the maximum likelihood estimation method. When the number of errors is small, the character string and erasure information 13 are input to the decoding unit 3 using the limit distance decoding method. Is entered.

The decoding result 1 of the decoding unit 2 using the maximum likelihood estimation method
4 and the decoding result 15 of the decoding unit 3 by the limit distance decoding method
Is input to the output unit 4, converted into an original character string, and output as a decoding result 16.

The flowchart shown in FIG. 5 schematically shows the overall processing operation of the decoding device shown in FIG.
The error number determination unit 1 compares the number of code errors with a value based on the number of parities (for example, the number of parities “3” itself) (step S1). If not, the decoding process is performed by the limit distance decoding method based on the character string and the erasure information 13 (step S2). In other cases, a decoding process using the maximum likelihood estimation method is performed based on the character string candidates and the likelihood information 12 (step S3).

As an example of the method of using the maximum likelihood estimation method, there is a method in which if there is a decoded character string candidate having a likelihood higher than a certain threshold, it is treated as a decoding result.

The code in the upper stage must be a code that can be decoded by the maximum likelihood estimation method. Further, in a system having many errors and thus many processes by the maximum likelihood estimation method, it is relatively effective to select and use a high-performance code in the maximum likelihood estimation method rather than a high-performance code in the limit distance decoding method. is there.

Next, a description will be given of a method of using the likelihood at the lower stage when decoding is performed by the maximum likelihood estimation method, that is, a decoding method in cooperation with the lower stage when there are a plurality of candidates having high likelihood. Here, when performing decoding by the maximum likelihood estimation method, ordinary decoding such as limit distance decoding represented by decoding of a lower-stage RS code for a candidate having a high likelihood generally exceeds the capability of parity. This will not be done because it will cause erroneous correction. Instead, a decoding method is adopted in which a candidate having a high likelihood is coded and only the one that satisfies the parity condition of the lower stage is left as a candidate. If there is a single candidate, that is, the candidate is adopted as having been decoded, and if it is not a single candidate, it is rejected.

FIG. 6 shows a configuration example of the decoding unit 2 using the maximum estimating method for performing such processing. Less than,
6 will be described with reference to the flowchart shown in FIG.

Of the character string candidates having a high likelihood, candidates for the code information part (see FIG. 2) are temporarily stored in the information part candidate setting section 51 (step S10), and each of these candidates 61 is encoded. To generate a parity part candidate 62 (step S11).

On the other hand, among the character string candidates with a high likelihood, candidates for the parity part of the code (see FIG. 2) are temporarily stored in the parity part candidate setting unit 53 (step S12). The comparing unit 54 compares the parity part candidate 63 temporarily stored in the parity part candidate setting unit 53 with the parity generated by the encoding unit 52, and when they match, sends a match signal 64 to the information part candidate saving unit 55. Output (Step S13).

When the match signal 64 is input, the information part candidate saving section 55 saves the corresponding character string candidate 61 as one of the correct answer candidates (step S14). After the encoding of all the character string candidates set in the information part candidate setting part 51 and the comparison between the parity generated thereby and the parity candidate are completed (steps S15 and S16), the information part candidate saving part 55 If there is only one candidate saved in (1) (step S17), the
4 is output, and if not a single result is output.

Hereinafter, the processing operation of the decoding unit 2 using the maximum estimating method will be specifically described. For example, in the case where the first bit of the fourth “0” and the second bit of the fifth “0” from the beginning are dropped in the constant weight code as shown in FIG. 3 of the postal code “2790001”, For example, there are 2 × 4 = 8 character string candidates for the entire 7 digits.

When each is coded, it is generally coded into a parity of a value that can be represented by eight different three digits (steps S10 to S11 in FIG. 7). Here,
For example, the parity value of “2790001” is “C17
0 ”.

On the other hand, when the second bit of the last “0” of the parity “C170” is dropped, “C17
0 ”,“ C176 ”,“ C178 ”,“ C179 ”and 4
There are street candidates. Each of them is compared with a parity value obtained by encoding eight different character string candidates (step S13 in FIG. 7).

First, the parity value of the encoding result of “2790001” is “C17” which is one of the parity candidates.
Since it matches "0", "2790001" is saved in the information part candidate saving part 55 (step S14 in FIG. 7).
At this time, comparison with the remaining parity candidates is omitted.

Assuming that the next character string candidate set in the information part candidate setting part 51 is "2790601", this is encoded, compared with four kinds of parity candidates, and saved if they match. If none of the four matches, save is not performed.

When this processing is completed with the eight character string candidates set in the information part candidate setting part 51 (steps S15 and S16 in FIG. 7), the information part candidates saved in the information part candidate saving part 55 Is confirmed, and if there is only one, the candidate is determined as a decoding result, and if there are a plurality of candidates, the result is rejected (step S17 in FIG. 7).

Further, as shown in FIG. 8, when a database 71 in which rules and knowledge concerning information sources are stored can be used, a plurality of information section candidates of the rejected decoding result (the information section in FIG. 6) The information stored in the database 71, for example, the information portion of the zip code is 10 digits, and the knowledge of the zip code book is used for the information stored in the candidate save section 55). Similarly, the candidates are narrowed down, and if the result is a single candidate, the candidate is adopted as that having been decoded, and if it is not a single candidate, it may be rejected.

The order of application of each process of decoding by decoding, rules, and knowledge may be changed according to each processing time, performance, that is, the degree of narrowing down candidates. Further, the processing is not limited to the barcode processing system used for mail processing, and the processing may be appropriately terminated depending on the number of candidates and the processing elapsed time according to the application object, and a result of rejection may be output.

The likelihood information may be used not only for the lower stage but also for decoding characters at different positions in the same stage. The decoding method of the multiplex decoding method described above is not limited to the barcode processing system.

[0053]

By decoding a character string coded by combining the maximum estimating method and the limit distance decoding method, the number of correctable errors corresponding to the number of parities guaranteed by the limit distance decoding method is exceeded. Even in the case of such a large number of errors, loss of code information is prevented, and decoding can be performed safely, easily, and at high speed. Further, when the maximum likelihood estimation method is employed, normal decoding is not performed in the lower stage, so that erroneous correction can be suppressed and decoding can be performed safely and easily.

[Brief description of the drawings]

FIG. 1 is a diagram schematically illustrating a configuration example of an encoding device of a barcode processing system according to an embodiment of the present invention.

FIG. 2 is a diagram showing a data configuration example when a character string composed of 7-digit characters and a parity composed of 3-digit characters are encoded.

FIG. 3 is a diagram for explaining a specific example of a constant weight code having a weight of 3; one code composed of three bars of two bits and four states (up, down, up, down, no) for one character; Is an example of allocation in the case where the number of codewords (codes) is 15.

FIG. 4 is a diagram schematically illustrating a configuration example of a decoding device according to an embodiment of the present invention.

FIG. 5 is a flowchart for explaining the outline of the processing operation of the decoding device in FIG. 4;

6 is a diagram schematically illustrating a configuration example of a decoding unit using the maximum estimating method of the decoding device in FIG. 4;

FIG. 7 is a flowchart for explaining the outline of the processing operation of the decoding unit in FIG. 6;

FIG. 8 is a diagram schematically illustrating another configuration example of a decoding device.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Error number determination part 2 ... Decoding part using the maximum estimating method 3 ... Decoding part by limiting distance decoding method 4 ... Output part 51 ... Information part candidate setting part 52 ... Encoding part 53 ... Parity part candidate setting part 54 … Comparing part 55… Information part candidate saving part

Claims (12)

[Claims] 1. A decoding method for decoding a code obtained by adding an error correction code to a character string and encoding the code, wherein the number of error words of the code is a number based on the number of error correction codes added to the character string. When the number is smaller, the code is decoded by performing error correction based on the error correction code, and when the number of erroneous words of the code is larger than the number based on the number of words of the error correction code added to the character string, the maximum is estimated. A decoding method for performing error correction on the code and decoding the code. 2. A decoding method for decoding a code obtained by adding an error correction code to a character string and encoding the code, wherein the number of error words of the code is a number based on the number of error correction codes added to the character string. When the number is smaller, the code is decoded by performing error correction based on the error correction code, and when the number of error words of the code is larger than the number based on the number of error correction codes added to the character string, the characteristic of the code is An error correction code is generated for each of the character string candidates of the original character string extracted based on the error correction code of each of the generated character string candidates and the original character string extracted based on the characteristics of the code. A decoding method comprising comparing error correction code candidates and detecting an original character string based on the comparison result. 3. A method according to claim 1, wherein only one of the generated error correction codes of each character string candidate is any one of the error correction code candidates of the original character string extracted based on the characteristics of the code.
3. The decoding method according to claim 2, wherein when the numbers match, a character string candidate corresponding to the error correction code is detected as an original character string. 4. The decoding method according to claim 1, wherein the code is a constant weight code for detecting a directional error. 5. A decoding apparatus for decoding a code obtained by adding an error correction code to a character string and encoding the code, wherein the number of error words of the code is a number based on the number of error correction codes added to the character string. When the number is smaller, first decoding means for performing error correction based on the error correction code and decoding the code, the number of error words of the code being greater than the number based on the number of error correction codes added to the character string. And a second decoding means for performing error correction by estimating the maximum value when the value is larger and decoding the code. 6. A decoding device for decoding a code obtained by adding an error correction code to a character string and encoding the code, wherein the number of error words of the code is a number based on the number of error correction codes added to the character string. When the number is smaller, decoding means for performing error correction based on the error correction code and decoding the code, and when the number of error words of the code is larger than the number based on the number of error correction codes added to the character string, the code Generating means for generating an error correction code for each of the character string candidates of the original character string extracted based on the characteristics of: an error correction code for each character string candidate generated by the generation means; Detecting means for comparing the error correction code candidates of the original character string extracted based on the extracted character string, and detecting the original character string based on the comparison result. 7. A method according to claim 1, wherein only one of the error correction codes of each character string candidate generated by said generating means is one of the error correction code candidates of the original character string extracted based on the characteristics of said code. 7. The decoding device according to claim 6, wherein, when one of them matches, a character string candidate corresponding to the error correction code is detected as an original character string. 8. The decoding apparatus according to claim 5, wherein the code is a constant weight code for detecting a directional error. 9. A code obtained by adding an error correction code to a character string and encoding it, converts the code into a barcode, prints it on a predetermined print target, reads the barcode from the print target, and reads the original barcode. In a bar code processing system for restoring a character string, conversion means for converting a bar code read from the printing object into a code word, and the number of error words of the code word converted by the conversion means is added to the character string. First decoding means for performing error correction based on the error correction code to decode the code when the number is smaller than the number based on the number of error correction codes, Second decoding means for estimating the maximum error and correcting the error and decoding the code when the number is larger than the number based on the number of words of the error correction code added to the character string. Over-de-processing system. 10. A code obtained by adding an error correction code to a character string and converting the code to a barcode, printing the barcode on a predetermined print target, reading the barcode from the print target, and reading the original barcode. In a bar code processing system for restoring a character string, conversion means for converting a bar code read from the printing object into a code word, and the number of error words of the code word converted by the conversion means is added to the character string. Decoding means for performing error correction based on the error correction code and decoding the code when the number is smaller than the number based on the number of error correction codes, and the number of error words of the code word converted by the conversion means is included in the character string. Generating means for generating each error correction code of a character string candidate of the original character string extracted based on the characteristics of the code, when the number is larger than the number based on the number of words of the added error correction code, The error correction code of each character string candidate generated by the generation means is compared with the error correction code candidate of the original character string extracted based on the characteristics of the code, and the original character string is determined based on the comparison result. A bar code processing system comprising: a detection unit configured to detect a bar code. 11. A method according to claim 1, wherein only one of the error correction codes of each character string candidate generated by said generating means is one of the error correction code candidates of the original character string extracted based on the characteristics of said code. 11. The barcode processing system according to claim 10, wherein when one of them matches, a character string candidate corresponding to the error correction code is detected as an original character string. 12. The bar code processing system according to claim 9, wherein the code is a constant weight code for detecting a directional error.