JP2002279352A - Character recognition device and method, and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a character recognition device which provides correct recognition result, without making a user work, when recognizing a URL(uniform resource locator) in a character image. SOLUTION: This character recognition device, connected to the Internet 90, is provided with a character recognition part 20 for recognizing characters in the inputted character image, an internet connection part 40 connected to a connection target expressed in the URL via the Internet 90 and obtaining the result, and a result verification part 30, giving a character column to the internet connection part 40, when the character column recognized by the character recognition part 20 is URL, and verifying whether the character recognition result is right.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character recognition device, a character recognition method, and a recording medium, and more particularly, to verification and correction when erroneous conversion is performed when a character string obtained by character recognition processing is a URL.

[0002]

2. Description of the Related Art Recently, many books and the like include URLs (Uniform Resource Locators) together with articles. When a document such as a book is read by a scanner and character recognition is performed, the
Although L must be recognized, misrecognition of the URL becomes useless at all, so accuracy must be ensured. However, the task of matching the recognition result with the URL in the document is troublesome.
As a technique for automatically recognizing characters constituting a URL described in such a book or the like, Japanese Patent Laid-Open No. 9-274646 discloses a technique.
There is the technology of the publication. In this method, an image for characters and templates (a fixed protocol, several parts such as a head and a type of a domain name are prepared) that can be used for a URL is stored, and after the URL part is extracted, the image is stored. A character string is recognized as a character string by comparing the character image with the extracted partial image. According to this method, although there is relatively no erroneous conversion for the template part, there is a problem that the recognition processing of the other parts is slow in recognition speed and erroneous recognition increases. On the other hand, for access to homepages around the world and collection and transmission of information via the Internet, such as e-mail services,
2. Description of the Related Art Computers and network terminals having an Internet connection function and a browser function (a function of browsing a homepage) have become widespread. JP-A-10-3340
In the technology disclosed in Japanese Patent Publication No. 17, when watching a television program using Internet television, a homepage of the program or a homepage related to the program contents may be introduced. In such a case, when the user instructs to view the introduced homepage, the UR displayed on the television screen is displayed.
The image of the L portion is extracted, and the image can be accessed using the URL obtained as a result of character recognition.

[0003]

However, the URL
If the part is character-recognized, the recognition rate depends only on the performance of the character recognition device. In order to solve this problem, in the above-described related art, when a plurality of candidates as a result of character recognition are detected, the plurality of URLs are sequentially displayed to allow the user to select the URL. in this way,
URL of input image and URL of candidate displayed sequentially
Must be checked one by one, and the user's effort will not be reduced at all. The present invention is intended to solve such a problem, and when recognizing a URL in a character image, a character recognizing device, a character recognizing method, and a recording medium for obtaining a correct recognizing result without labor for a user. The purpose is to provide.

[0004]

According to a first aspect of the present invention, there is provided a character recognition device connected to the Internet, wherein the character recognition unit recognizes characters of an input character image. UR via the Internet
An Internet connection unit that connects to the connection destination indicated by L and obtains the result; and when the character string recognized by the character recognition unit is a URL, the character string is given to the Internet connection unit to verify the recognition result. And a verification unit that verifies whether the result of character recognition is correct. According to a second aspect of the present invention, in the character recognition apparatus according to the first aspect, when the result verification unit recognizes the character string as a URL that does not exist, the result verification unit corrects the character string. The result verification unit may verify again with the character string corrected by the result correction unit. According to a third aspect of the present invention, in the character recognition device according to the second aspect, the character recognition unit obtains a candidate character and its likelihood as a recognition result, and the result correction unit determines the candidate character. It is characterized in that correction is performed from a character with low probability among candidate characters of a character string. According to a fourth aspect of the present invention, in the character recognition device according to the third aspect, the result correction unit sets a character with low probability among candidate characters of the character string of the recognition result as another candidate character. It is characterized by being replaced. According to a fifth aspect of the present invention, in the character recognition device according to the third or fourth aspect, the result correction unit determines a character having a low probability from among candidate characters of the character string of the recognition result as a candidate character. It is characterized by replacing with characters other than. According to a sixth aspect of the present invention, there is provided the character recognition apparatus according to the fifth aspect, further comprising a candidate lattice for holding a character of a recognition result and a character candidate that is likely to be erroneously recognized in association with each other; In the case where a character with low probability among the candidate characters of the character string of the recognition result is a character registered in the candidate lattice, the character is selected from the candidate lattice and replaced to correct a character that is easily misrecognized. It is characterized by doing so.

According to a seventh aspect of the present invention, there is provided a character recognition method for character recognition of an input character image, wherein the input character image is subjected to character recognition, and when the recognized character string is a URL, It connects to the connection destination indicated by the character string via the Internet, and verifies whether or not the recognition result is correct based on the result. According to an eighth aspect of the present invention, in the character recognition method according to the seventh aspect, in the case where the URL does not include the character string, the character string is corrected and verified again. According to a ninth aspect of the present invention, in the character recognition method according to the eighth aspect, a candidate character and its likelihood are obtained as a recognition result of the character recognition. It is characterized in that the candidate character of the character string is corrected from a character with low probability. Claim 10 of the present invention relates to claim 9
In the character recognition method described above, among the candidate characters of the character string of the recognition result, a character with low probability is replaced with another candidate character. According to a eleventh aspect of the present invention, in the character recognition method according to the ninth or tenth aspect, a character having low probability among candidate characters of the character string of the recognition result is replaced with a character other than the candidate character. It is characterized by. Claim 12 of the present invention is:
12. The character recognition method according to claim 11, further comprising a candidate lattice for holding a character of a recognition result and a character candidate that is likely to be erroneously recognized in association with each other, and determining a certainty among the candidate characters of the character string of the recognition result. When the low character is a character registered in the candidate lattice, the character is selected from the candidate lattice and replaced, and a character that is easily misrecognized is corrected. According to a thirteenth aspect of the present invention, there is provided a computer-readable recording medium recording a program for causing a computer to function as a character recognition device that recognizes a character image input from a computer connected to the Internet, A character recognition unit for character recognition of an input character image, an Internet connection unit for connecting to a connection destination indicated by a URL via the Internet and obtaining the result, and a character string recognized by the character recognition unit In the case of a URL, a character verification unit is provided with a result verification unit that verifies the recognition result by giving the character string to the Internet connection unit, and records a program for verifying whether the character recognition result is correct. Also,
According to a fourteenth aspect of the present invention, in the recording medium according to the thirteenth aspect, when the result verification unit recognizes the character string as a URL that does not exist, the result verification unit includes a result correction unit that corrects the character string. The result verification unit is characterized by recording a program for performing verification again with the character string corrected by the result correction unit.

[0006]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The construction and operation of an embodiment of the present invention will be described below in detail with reference to the drawings. Embodiment (1) Configuration FIG. 1 is a block diagram showing a configuration of a character recognition device according to an embodiment of the present invention. The character recognition device of the embodiment is
Image input unit 10, character recognition unit 20, result verification unit 30, Internet connection unit 40, result correction unit 50, display unit 6
0, an image memory 70, a recognition result memory 80, and a network 90. The user captures a document image obtained by reading a document with a scanner or a CCD camera, or an image of a television screen or a file. Image input unit 10
Captures a character image to be recognized from the image. The user designates an area to be recognized with a mouse or the like, or specifies a character image area using a technique such as area identification, and stores the image data in the image memory 70.
The character recognition unit 20 cuts out the image data in the image memory 70 into individual character images, normalizes the size, extracts the feature amount of the normalized character image, and recognizes the feature amount as a standard pattern. Compared with the dictionary, a certain number (for example, 10) of candidate characters are obtained in ascending order of the value of the probability of recognition (hereinafter referred to as the certainty factor), and the candidate characters and their certainty factors are stored in the recognition result memory 80. . Result verification unit 30
Determines whether the document image to be recognized is a character string representing a URL. This determination may be made in advance using a pointing device such as a mouse when a character string corresponding to a URL is displayed on a display or the like before character recognition. Alternatively, a trigger character such as “http” may be detected at the stage of an image, or may be detected from a recognition result.

When a character string representing a URL is detected, the character string is passed to the Internet connection unit 40 to connect to an address corresponding to the URL character string via the Internet 90. As a result, if it is found that the connection is normally made from the Internet connection unit 40, it is determined that the URL character string of the recognition result is correctly recognized, and the recognition result memory 80 is updated with this character string. However, if it is found that the connection has not been made normally, the UR of this recognition result
In order to correct the L character string, the URL character string is passed to the result correction unit 50, and the corrected URL character string is passed again to the Internet connection unit 40 and verified again. This is repeated until it is correct, or if the correction is not correct even after repeating the predetermined number of times, the user is made to correct it. The Internet connection unit 40 includes software for Internet connection and a WWW browser, connects to the Internet, and performs processing related to browsing homepages on a WWW server. U from the result verification unit 30
When the RL is passed, it accesses the WWW server having the URL. At this time, if the URL character string is a correct address, an HTML document is sent from the destination WWW server. However, if the address is not correct, "Not
An error code like "Found" is returned. The result, that is, whether the connection to the WWW server of the given URL has been normally made is returned to the result verification unit 40.

For example, as shown in FIG. 2, for an input image “http://www.ricoh.co.jp”, “http: //www.r
Suppose that the recognition result of "ic0h.co.jp" was obtained. Also,
Assuming that there is no domain name "ric0h", even if you connect to the Internet as it is, this address does not exist and an error code will be returned. The result correction unit 50 corrects and returns the URL character string of the recognition result passed from the result verification unit 30. This corrects a character with a low degree of certainty in the URL character string of the passed recognition result. For example, referring to FIG. 2, the UR of the recognition result is input to the input image
The confidence of the L string "http://www.ric0h.co.jp"
Characters other than “0” have a certainty of 99, and only “0” has a certainty of 50. In such a case, the following correction is performed while paying attention to the point “0” having the lowest certainty factor. Any one of these modifications may be performed sequentially. (a) Replace with the second or lower candidate character of the correction target character.
In the case of the above example, the URL is formed again using the second and subsequent candidate characters of “0”. For example, the second candidate is replaced with O (capital letter O), and if that is not the case, the third candidate is replaced with o (lowercase letter O), and so on. (b) For a character that is likely to be erroneously recognized, the original character is stored in advance as a candidate character lattice in a table or the like, and when any of the characters appears in the recognition result, the correction target character (See FIG. 3). o (lowercase o), 0 (zero), O (uppercase o)
Characters such as, 1 (Ichi), l (L), etc. are stored in a table or the like as candidate character lattices in advance for characters that are easily misrecognized, and when the characters in them are to be Are sequentially replaced. (c) Register a character pattern in which an error in character segmentation is likely to occur, and when these characters appear, replace them with another registered character pattern (see FIG. 5).

For example, referring to FIG. 4, an input image "ht
tp: //www.ricoh.co.jp ”and the recognition result“ http: // w
ww.ncoh.co.jp ". This occurs because “ri” was mistakenly regarded as one character when extracting characters from the input image. In such a case, a cutout candidate lattice is stored in a table or the like in advance for a character that is likely to make a cutout error as shown in FIG. In this case, too, the certainty at the time of recognizing “n” should be low, so candidate characters for “n” are sequentially extracted and replaced with reference to the extraction candidate lattice. In this case, replace "n" with the pattern "ri" and use "http: //www.ricoh.co.j
Generate a modified UTL string "p". (d) The URL includes a protocol, a domain name representing an address, a path, or a file name. These characters generally consist of letters, numbers and symbols. Therefore, the character to be corrected is replaced with any character that can constitute this URL. In the replacement as described above, an increase in the number of combinations can be suppressed by setting a threshold value in advance for the certainty factor for correcting the character recognition result. The display unit 60
An image input by a scanner or the like is displayed on a display device such as a display, or used when a user selects a character image to be recognized from among the displayed images. And display it. The image memory 70 is a storage unit (for example, a storage device such as a memory or a magnetic disk) that stores an input image. This image is an image to be recognized by the character recognition unit 20 and is displayed in comparison with the recognition result.
It is used when correcting the recognition result. The recognition result memory 80 stores a predetermined number (for example, 1) of candidate characters having a high degree of certainty for all the character images to be recognized.
0), at least the character code and the certainty factor are stored. The network 90 is a transmission path for connecting the character recognition device of the present invention and a WWW (World Wide Web) server, and is generally realized by a cable, and uses TCP / IP as a communication protocol. However, the transmission path is not limited to the cable, and may use a wireless LAN or the like as long as the communication protocol between them is the same.

In such a configuration, the character recognition device of the embodiment operates as follows. The user inputs a document to be recognized by a scanner or a CCD camera, or inputs an image from a file. The image input unit 10 displays the input image and stores an image of an area where character recognition is desired in the image memory 70. Next, character recognition means 2
0 indicates that the image data in the image memory 70 is cut out into individual character images, the size is normalized, the feature amount of the normalized character image is extracted, and the feature amount is compared with the recognition dictionary to perform recognition. A certain number of candidate characters are obtained in the order of increasing likelihood (certainty) of the candidate characters, and the candidate characters and the certainty are stored in the recognition result memory 80 as recognition result data. Next, the result verification unit 30 is called. Result verification unit 30
When a character string representing a URL is detected in a document image to be recognized, the character string is
To connect to the address corresponding to the URL character string via the Internet 90. Upon receiving the URL from the result verification unit 30, the Internet connection unit 40 accesses the WWW server having the URL, and
It monitors whether an HTML document or an error code is sent from the destination WWW server, and returns to the result verification unit 40 whether or not the connection was normally established. This result is received by the result verification unit 40, and if it is determined that the connection has been made normally, it is determined that the URL character string of the recognition result has been correctly recognized, the recognition result memory 80 is updated, and the verification ends. But,
If the connection is not successful, the UR of this recognition result
In order to correct the L character string, the URL character string is passed to the result correction unit 50, and the corrected URL character string is passed again to the Internet connection unit 40 and verified again. This is repeated until it is correct, or if the correction is not correct even after repeating the predetermined number of times, the user is made to correct it. The result correction unit 50 corrects and returns the URL character string of the recognition result passed from the result verification unit 30. In this case, the following correction is sequentially performed on characters having a low degree of certainty in the URL character string of the passed recognition result. (a) Replace with the second or lower candidate character of the correction target character. (b) For a character that is likely to be erroneously recognized, the original character is stored in advance as a candidate character lattice in a table or the like, and when any of the characters appears in the recognition result, the character to be corrected is extracted from the candidate character lattice. Replace with (c) A character pattern in which an error in character segmentation is likely to be registered is registered, and when these characters appear, the character pattern is replaced with another registered character pattern. (d) The character to be corrected is replaced with any character that can constitute a URL. Finally, the display unit 60 stores the recognition result thus verified in the recognition result memory 80.
And displays it on a display device such as a display in comparison with the input image.

With the above-described structure, when a URL character string is present in a document image, the recognition of the URL character string can be performed accurately and the user's labor can be reduced. Also, the act of actually connecting to the other party in an uncertain state is annoying to the wrong party by telephone or e-mail.
Regarding the WW service, although there is an increase in network traffic, verification can be performed without direct trouble.

(2) Processing Flow FIG. 6 is a flowchart for explaining the processing flow of the character recognition device of this embodiment. The user scans a document to be recognized by a scanner or C
Input from a CD camera or input an image from a file. The input image is displayed, and the image of the area where character recognition is desired is stored in the image memory 70 (step S100). The image data in the image memory 70 is cut out into individual character images, the size is normalized, the feature amount of the normalized character image is extracted, and the feature amount is compared with a recognition dictionary to confirm the recognition accuracy. A certain number of candidate characters are obtained in descending order of (certainty), and the candidate characters and the certainty are stored in the recognition result memory 80 as recognition result data (step S110). A character string representing a URL is detected in the document image to be recognized, and this character string is called to the partner WWW server via the Internet 90.
It accesses and monitors whether an HTML document or an error code is sent from the destination WWW server (step S120). When a response is returned from the WWW server, it is determined whether or not the connection has been made normally (step S130). When the connection has been made normally, the recognition result memory 80 is updated, and this verification ends, and step S15 is performed.
Go to 0. However, if the connection is not successful,
The process proceeds to step S140 to correct the URL character string of the recognition result. In step S140, the recognition result U
Modify the RL string and return. This is because the following corrections are sequentially performed on characters having a low degree of certainty in the URL character string of the passed recognition result, and the process returns to step S120, where the corrected URL character string is again transmitted to the Internet. To verify again. (a) Replace with the second or lower candidate character of the correction target character. (b) When any of the characters registered in the candidate character lattice previously stored as a table with respect to the character that is likely to be erroneously recognized appears in the recognition result, the character is replaced with the candidate character of the candidate character lattice. (c) A character pattern in which an error in character segmentation is likely to be registered is registered, and when these characters appear, the character pattern is replaced with another registered character pattern. (d) The character to be corrected is replaced with any character that can constitute a URL. Finally, in step S150, the recognition result that has been verified in this way is retrieved from the recognition result memory 80, and displayed on a display device such as a display in comparison with the input image.

<Example by Computer> Further, the present invention is not limited to only the above-described embodiment. For example, the character recognition device of FIG. 1 can be realized by the configuration of a computer device as shown in FIG. 7 includes an input device 1, an output device 2, a CPU (Central Processing Unit) 3, a memory 4, a storage device 5, a medium drive device 6, and a network connection device 7. Are connected to each other. The input device 1 includes a keyboard, a mouse, a touch panel, a scanner, and the like, and is used to input text and image information. Output device 2
Is a device for outputting various output information, information input from the input device 1, and the like, and is a display device or a printer device. The CPU 3 operates various programs. The memory 4 holds the image memory 70, the recognition result memory 80, and the program itself, and also holds information temporarily created when the program is executed by the CPU 3. The storage device 5 stores a program, temporary information at the time of executing the program, and the like. Image memory 70 and recognition result memory 80 handled by the character recognition device of the present invention
May be stored in the storage device 5 and taken out to the memory 4 when the program is executed. The medium drive device 6 is used to mount a recording medium storing programs, data, and the like, read them, and store them in the memory 4 or the storage device 5. Further, it may be used for directly inputting / outputting data or executing a program. As a recording medium, a semiconductor medium (for example, ROM, IC memory card, etc.), an optical medium (for example, DVD, MO, MD, CD-R, etc.), a magnetic medium (for example, magnetic tape, flexible disk, etc.)
Any of these may be used. The network connection device 7
The computer device is connected to the WWW server via the network 90.

Each function constituting the character recognition apparatus shown in FIG. 1 is programmed and written in advance on a recording medium such as a CD-ROM.
By installing these programs in a computer equipped with a medium drive device 6 such as an OM drive, and storing these programs in the memory 4 or the storage device 5 of each computer device, and executing them, A similar function can be realized. Further, not only the functions of the above-described embodiments are realized by executing the loaded program, but also the operating system or the like performs part or all of the actual processing based on the instructions of the program, and the processing performs The case where the functions of the above-described embodiments are realized is also included. Further, when the program for realizing the above-described embodiment is a semiconductor recording medium such as a ROM, the program is directly loaded into the memory 4 and executed not from the medium driving device 6.
Further, a program for realizing the functions of the above-described embodiment is loaded into the memory provided on the function expansion board or the function expansion unit, and the CPU or the like provided on the function expansion board or the function expansion unit is actually executed based on the instructions of the program. And a part of the entire process is performed, and the function of the above-described embodiment is realized by the process.

<Operation in Network Environment of the Present Invention> FIG. 8 shows a configuration of a mode in which the present invention is connected to a network and operated. A user terminal 110 composed of a computer device as shown in FIG.
Receives the character recognition program from the server 120 holding the character recognition program, inputs an image, causes the terminal 110 to execute the received character recognition program, and outputs the execution result. This has the advantage that the latest character recognition program can always be used, and the recognition dictionary can be unique to the user. The server 120 and the terminal 1 as shown in FIG.
10 can be connected via a network 90, and a character recognition program for realizing the functions of the present invention can be stored in a storage device such as a magnetic disk of the server 120 and distributed in a form such as download. Further, the program for realizing the functions of the present invention may be provided by distributing it by broadcast waves.

[0016]

As described above, according to the present invention,
When there is a URL character string in a document image, the recognition can be performed accurately and with less user effort.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a character recognition device according to an embodiment of the present invention.

FIG. 2 is a diagram for explaining an image and a recognition result when erroneous recognition is performed.

FIG. 3 is a diagram of a candidate character lattice relating to characters that are easily misrecognized.

FIG. 4 is a diagram for explaining an image and a recognition result when a cutout error occurs.

FIG. 5 is a diagram for explaining a candidate lattice in the case of a cutout error.

FIG. 6 is a flowchart illustrating a process according to the embodiment;

FIG. 7 is a diagram illustrating a hardware configuration when the image processing apparatus of the present invention is implemented by a computer.

FIG. 8 is a diagram showing a configuration of a mode in which the present invention is operated in a network.

[Explanation of symbols]

 Reference Signs List 1 input device 2 output device 3 CPU 4 memory 5 storage device 6 medium drive device 7 network connection device 8 bus 10 image input portion 20 character recognition portion 30 result verification portion 40 internet connection portion 50 result correction portion 60 display portion 70 image memory 80 Recognition result memory 90 Internet 110 Terminal 120 Server

Claims (14)

[Claims] 1. A character recognition device connected to the Internet, a character recognition unit for character recognition of an input character image, and an Internet connected to a connection destination indicated by a URL via the Internet to obtain a result. A connection unit, and a result verification unit that, when the character string recognized by the character recognition unit is a URL, gives the character string to the Internet connection unit and verifies the recognition result, and the character recognition result is correct. A character recognition device for verifying whether or not a character is recognized. 2. The character recognition device according to claim 1, wherein the character string of the recognition result in the result verification unit does not exist.
A character recognition device comprising: a result correction unit that corrects a character string in the case of RL, wherein the result verification unit performs verification again with the character string corrected by the result correction unit. 3. The character recognition device according to claim 2, wherein the character recognition unit obtains a candidate character and its certainty as a recognition result, and wherein the result correction unit obtains a candidate character of the character string of the recognition result. A character recognition device characterized by correcting characters having low probability among the characters. 4. The character recognition device according to claim 3, wherein the result correction unit replaces a character with low probability among candidate characters of the character string of the recognition result with another candidate character. Character recognition device. 5. The character recognition device according to claim 3, wherein the result correction unit replaces a character with low probability among candidate characters of the character string of the recognition result with a character other than the candidate character. A character recognition device characterized in that: 6. The character recognition device according to claim 5, further comprising: a candidate lattice that holds a character of a recognition result and a character candidate that is likely to be erroneously recognized in association with each other; In the case where a character with low probability among the candidate characters of the character string is a character registered in the candidate lattice, the character is selected from the candidate lattice and replaced, and a character that is easily misrecognized is corrected. Character recognition device. 7. A character recognition method for recognizing a character from an input character image, comprising the steps of: (a) recognizing a character of an input character image; A character recognition method comprising: connecting to a connection destination indicated by, and verifying whether or not the recognition result is correct based on the result. 8. The character recognition method according to claim 7, wherein in the case of a URL where the character string does not exist, the character string is corrected and verified again. 9. The character recognition method according to claim 8, wherein a candidate character and its certainty are obtained as a recognition result of the character recognition, and in the case of a URL where the character string does not exist, a candidate for the character string of the recognition result is obtained. A character recognition method characterized by correcting characters having low certainty among characters. 10. The character recognition method according to claim 9, wherein among the candidate characters of the character string of the recognition result, a character with low probability is replaced with another candidate character. 11. The character recognition method according to claim 9, wherein a character with low probability among candidate characters of the character string of the recognition result is replaced with a character other than the candidate character. Recognition method. 12. The character recognition method according to claim 11, further comprising a candidate lattice for holding a character of a recognition result and a character candidate that is likely to be erroneously recognized in association with each other, wherein the candidate character of the character string of the recognition result is provided. A character having a low likelihood is a character registered in the candidate lattice, the character is selected from the candidate lattice and replaced, and a character which is easily misrecognized is corrected. . 13. A computer-readable recording medium recording a program for causing a character recognition device to function as a character recognition device for character images input from a computer connected to the Internet, comprising: A character recognition unit that recognizes a character, an Internet connection unit that connects to a connection destination indicated by a URL via the Internet and obtains a result, and a character string recognized by the character recognition unit is a URL. A computer-readable recording medium, comprising: a result verification unit that verifies a recognition result by providing the character string to an Internet connection unit, and records a program that verifies whether the character recognition result is correct. . 14. The recording medium according to claim 13, wherein the character string of the recognition result in the result verification unit does not exist.
A computer that has a result correction unit that corrects the character string in the case of the RL, wherein the result verification unit records a program that verifies again with the character string corrected by the result correction unit; A readable recording medium.