JP2008108153A - Information processing system and information processing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an information processing system capable of enhancing conversion accuracy, operability for correction, and convenience, when converting an image of a paper document into a character code, using a character recognition result obtained by recognizing characters from the image of the input paper document, and using an input, from an operator, of characters in an equivalent paper document. <P>SOLUTION: A comparison means of this information processing system compares the first input value by an operation of the operator in an equivalent item with the second input value by character recognition processing, a storage means stores the first input value in the past, the second input value in the past in the equivalent item corresponding to the first input value, and a value after corrected in the equivalent item, while correlated each other, and a correction means executes correction based on the storage means, when the first input value is different from the second input value by the comparison by the comparison means. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an information processing system and an information processing program.

  Paper documents are excellent media for transmitting and recording information, but a space such as a library is required to store them. In addition, when information is recorded and stored in a paper document, if the information recorded in the paper document is needed later, the target information can be selected from many paper documents stored in the archive. I have to find a paper document in which I was recorded.

  Therefore, digitization and storage of paper documents has been performed. Specifically, an image corresponding to each page of a paper document is read by a scanner or the like, and image data corresponding to the image is filed for each paper document and stored in a storage device such as a hard disk. It is coming. Then, character recognition processing is performed on the document image data so that it can be handled as a document as a text code.

  As a technique related to this, for example, Patent Document 1 relates to a recognition result display method for displaying a recognition result of a character or the like that is used in an automatic explanation machine or an automatic announcement machine and is recognized by an optical character reader. Because the image is displayed in correspondence with the keyword in the recognition result while outputting the recognition result, it is possible to see the figure and table on the display along the flow of the text, making it easy to grasp the contents of the document. , A technology that can efficiently convey the contents to the other party is disclosed.

For paper documents such as forms, it is particularly desired to reduce misrecognition due to character recognition. For example, the operator visually observes the character recognition result, and when a misrecognized character is generated, a correction input is made each time to correct the result. As a technique related to this, for example, Patent Document 2 has a problem of processing a form handwritten by many people at a high recognition rate and at a high speed. Correction data that stores the standard character pattern dictionary file stored corresponding to the code, the key information that identifies the person who entered the form, and the character image that was unread during character recognition and the character code that was corrected Refers to the history save file, the correction data history save file that stores the character image that was misread during character recognition and the corrected character code, and the image data by referring to the standard character pattern dictionary file. Recognize handwritten characters from the list, and if the recognition result is unread or misread, the correction data history save file is based on the key information. Rui technique and a image data character recognition unit for correcting or modifying with reference to correction data history storage file is disclosed.
JP-A-2-271470 JP 2005-266925 A

  The present invention has been made in the background of the background art as described above, and has been input as an object that accurate and efficient character code conversion is difficult when converting an image of a paper document into a character code. Using the character recognition result obtained by character recognition from the image of the paper document and the input result of the paper document by the operator, when converting the image of the paper document to the character code, the operability of accurate conversion and correction An object of the present invention is to provide an information processing system and an information processing program that improve convenience.

The gist of the present invention for achieving the object lies in the inventions of the following items.
The invention according to claim 1 of the present application is a comparing means for comparing a first input value by an operator's operation with respect to an equivalent item and a second input value by a character recognition process;
Storage means for associating and storing past first input values, past second input values for equivalent items corresponding to the first input values, and corrected values for equivalent items;
When the first input value and the second input value are different from each other by the comparison by the comparison means, a correction means for correcting based on the contents stored in the storage means is provided. Information processing system.

In the invention according to claim 2 of the present application, when the first input value and the second input value are equal to each other by the comparison by the comparison means, the storage means Memorize the input value,
When the first input value is different from the second input value, the correcting means searches for the second input value stored in the storage means based on the second input value, and is searched. The information processing system according to claim 1, wherein the information is corrected by a corrected value corresponding to the second input value.

The invention according to claim 3 of the present application is characterized in that the correction means corrects based on past corrections in an equivalent document stored in the storage means. It is.

In the invention according to claim 4, the character recognition process outputs a character recognition result and a certainty factor by the character recognition process,
The storage means further stores a certainty factor by the character recognition process,
The information processing system according to claim 1, wherein the correction unit corrects according to a certainty factor stored in the storage unit.

The invention according to claim 5 is an image input means for inputting an image;
Character recognition processing means for recognizing characters of the image input by the image input means;
A character input means for inputting characters in the image by an operation of an operator;
Comparing means for comparing the first input value by the character input means for the equivalent item with the second input value by the character recognition processing means;
Storage means for associating and storing past first input values, past second input values for equivalent items corresponding to the first input values, and corrected values for equivalent items;
When the first input value and the second input value are different by the comparison by the comparison means, a first correction means for correcting based on the storage means;
An information processing system comprising: second correction means for correcting a result of correction by the first correction means by an operation of an operator.

The invention according to claim 6 of the present application provides a computer,
A comparison means for comparing a first input value obtained by an operator's operation on an equivalent item with a second input value obtained by a character recognition process;
Storage means for associating and storing past first input values, past second input values for equivalent items corresponding to the first input values, and corrected values for equivalent items;
When the first input value and the second input value are different by the comparison by the comparison unit, the comparison unit functions as a correction unit that corrects the content based on the contents stored in the storage unit. An information processing program.

In the invention according to claim 7 of the present application, when the first input value and the second input value are equal to each other by the comparison by the comparison unit, the storage means Memorize the input value,
When the first input value is different from the second input value, the correcting means searches for the second input value stored in the storage means based on the second input value, and is searched. The information processing program according to claim 6, wherein the information is corrected by a corrected value corresponding to the second input value.

The invention according to claim 8 is the information processing program according to claim 6, wherein the correction means corrects based on past corrections in an equivalent document stored in the storage means. It is.

In the invention according to claim 9 of the present application, the character recognition process also outputs a character recognition result and a certainty factor by the character recognition process,
The storage means further stores a certainty factor by the character recognition process,
The information processing program according to claim 6, wherein the correcting unit corrects according to a certainty factor stored in the storage unit.

  According to the information processing system of claim 1, compared to the case where the present configuration is not provided, the character recognition result obtained by character recognition from the input paper document image and the input result of the paper document by the operator When converting the image of the paper document into a character code using the above, accurate character code conversion can be performed, and the operability and convenience of correction can be improved.

  According to the information processing system of claim 2, it is possible to use past results in a wider range, more accurate character code conversion, and operability of correction as compared with the case where this configuration is not provided. Convenience can be further improved.

  According to the information processing system of claim 3, the probability that the correction for the same erroneous recognition result can be used is higher than in the case where the present configuration is not provided, and more accurate character code conversion can be performed. And convenience can be further improved.

  According to the information processing system of claim 4, the probability that the correction for the same erroneous recognition result among the past character recognition results can be used is higher than in the case where the present configuration is not provided, and more accurate characters are obtained. Code conversion can be performed, and the operability and convenience of correction can be further improved.

  According to the information processing system of claim 5, compared with the case where the present configuration is not provided, the character recognition result obtained by character recognition from the input paper document image and the input result of the paper document by the operator When converting the image of the paper document into a character code using the above, accurate character code conversion can be performed, and the operability and convenience of correction can be improved.

  According to the information processing program of claim 6, compared with the case where the present configuration is not provided, the character recognition result obtained by character recognition from the input paper document image and the input result of the paper document by the operator When converting the image of the paper document into a character code using the above, accurate character code conversion can be performed, and the operability and convenience of correction can be improved.

  According to the information processing program of the seventh aspect, compared with the case where the present configuration is not provided, past results can be used in a wider range, more accurate character code conversion can be performed, and correction operability can be achieved. Convenience can be further improved.

  According to the information processing program of claim 8, there is a higher probability that the correction for the same erroneous recognition result can be used compared to the case where the present configuration is not provided, and more accurate character code conversion can be performed. And convenience can be further improved.

  According to the information processing program of claim 9, the probability that the correction for the same erroneous recognition result among the past character recognition results can be used is further increased as compared to the case where the present configuration is not provided, and more accurate characters are obtained. Code conversion can be performed, and the operability and convenience of correction can be further improved.

Hereinafter, a preferred embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a conceptual module configuration diagram of an embodiment of the present invention.
The module generally refers to a component such as software or hardware that can be logically separated. Therefore, the module in the present embodiment indicates not only a module in a program but also a module in a hardware configuration. Therefore, the present embodiment also serves as an explanation of a program, a system, and a method. In addition, the modules correspond almost one-to-one with the functions. However, in mounting, one module may be composed of one program, or a plurality of modules may be composed of one program. A plurality of programs may be used. The plurality of modules may be executed by one computer, or one module may be executed by a plurality of computers in a distributed or parallel environment. Hereinafter, “connection” includes not only physical connection but also logical connection.
The system includes a configuration in which a plurality of computers, hardware, devices, and the like are connected via a network and the like, and includes a case where the system is realized by one computer, hardware, devices, and the like.
Hereinafter, a form will be mainly exemplified and described as an image including a character image to be recognized.
The input value by the operator's operation and the input value by the character recognition process are character codes, and the character code includes a numeric code and a character string code string. Not only Japanese character codes but also English, Chinese, and Korean character codes may be used.

As shown in FIG. 1, the present embodiment includes an image input module 110, an area division module 120, a character recognition module 130, an operator input module 140, a comparison module 150, a correction module 160, and a learning DB (database) 170. ing.
As shown in FIG. 1, the image input module 110 is connected to the area dividing module 120 and inputs a form image. More specifically, a form that is a paper document is input as an image by a scanner.
As shown in FIG. 1, the area division module 120 is connected to the image input module 110 and the character recognition module 130, and divides the area constituting the image from the image input by the image input module 110. The result of the area division is passed to the character recognition module 130.
The division of the area is performed by dividing an area divided by a blank or the like from the image. In this region division processing, for example, an input image by the image input module 110 is binarized and divided by a white pixel region having a certain length (or area) or more. By dividing in this way, a black pixel region (for example, a rectangle) can be divided. Furthermore, the characteristics (for example, area, vertical or horizontal size, shape, location of black pixel block, etc.) of the black pixel area are divided into character areas, graphic areas, table areas, photo areas, etc. Can be determined. Of course, the area may be divided by other existing methods.

As shown in FIG. 1, the character recognition module 130 is connected to the area division module 120 and the comparison module 150, and performs character recognition on the area of the character image among the images divided by the area division module 120. The recognition result is passed to the comparison module 150. The character recognition module 130 generates a character string that is text data by performing character recognition processing on the character area and the table area. However, this character recognition result may include a misrecognition result or the like. Further, a certainty factor representing the certainty of the character recognition result may be calculated and added to each recognition result. Note that the certainty factor is a value calculated as a value indicating the reliability of the character code obtained as a result of the character recognition process. For example, the reliability value of the character described in Japanese Patent No. 2991779 is applied. be able to.
In addition, when the form image input by the image input module 110 is only a character or when an area in which characters are described is set in advance and the character can be recognized according to the setting, the area dividing module 120 is used. The character recognition process may be performed directly on the image input by the image input module 110 without performing the process according to the above.

As shown in FIG. 1, the operator input module 140 is connected to the comparison module 150, and the contents of the form equivalent to the form that has been recognized by the character recognition module 130 are displayed as an operator (operator, user, or For example, using a keyboard. The equivalent includes at least the same case.
As shown in FIG. 1, the comparison module 150 is connected to the character recognition module 130, the operator input module 140, the correction module 160, and the learning DB 170, and the input value by the operator input module 140 for the equivalent item in the equivalent form The input value by the character recognition module 130 is compared.
That is, the contents of the equivalent items of the equivalent form are input by the character recognition module 130 and the operator input module 140 by two different methods of character recognition and input by the operator. In this case, originally, the character codes input from both should match. However, with the current character recognition technology, it is difficult to guarantee 100% of the recognition result, and erroneous recognition results are mixed. On the other hand, the input by the operator can also be an erroneous input. However, since the input is performed by two different methods, it can be said that there is a low probability that both are wrong. Therefore, there is a high possibility that one of them is a correct character code. Moreover, even if both are mistaken, it is thought that the probability of making a mistake similarly is still lower. Therefore, if there is an erroneous recognition due to character recognition or an erroneous input by an operator, the character codes input by the two types of methods should not match.
The comparison module 150 compares the input value from the operator input module 140 with the input value from the character recognition module 130, and if they are different, it determines that an error has occurred and triggers correction.

As shown in FIG. 1, the correction module 160 is connected to the comparison module 150 and the learning DB 170, and the input value from the operator input module 140 and the input value from the character recognition module 130 are different due to the comparison by the comparison module 150. Corrects the input value based on the contents stored in the learning DB 170. That is, erroneous input by the operator input module 140 and erroneous recognition by the character recognition module 130 are often repeated for equivalent characters. The correction module 160 makes corrections by referring to past correction results.
In addition, when the input value from the operator input module 140 and the input value from the character recognition module 130 are different, the correction module 160 has a character recognition module 130 stored in the learning DB 170 based on the input value from the character recognition module 130. It is also possible to search for the input value by and to correct the value after correction corresponding to the input value by the searched character recognition module 130. That is, the character recognition by the character recognition module 130 has good reproducibility, but the input by the operator input module 140 is not constant because it is performed by the operator. Therefore, when searching for past correction results, the character recognition result by the character recognition module 130 may be preferentially searched.
Further, the correction module 160 may make corrections based on past corrections in an equivalent form stored in the learning DB 170 in order to use the immediately preceding result with the highest priority.

Further, the correction module 160 may be configured so that the operator performs final confirmation and second correction after the above-described correction. That is, the correction result by the correction module 160 may be displayed, an operation related to the correction by the operator may be received, and the correction may be performed according to the operation. In that case, the correction result by the operator is stored in the learning DB 170.
As shown in FIG. 1, the learning DB 170 is connected to the comparison module 150 and the correction module 160, and the past values for the equivalent values corresponding to the input values of the past operator input module 140 and the input values of the operator input module 140 are stored. It is a database that stores the input value by the character recognition module 130 and the value after correction by the correction module 160 for equivalent items in association with each other. Further, the learning DB 170 determines that the input value by the operator input module 140 and the input value by the character recognition module 130 are the same as the input value by the operator input module 140 or the input by the character recognition module 130 when the input value by the operator input module 140 is equivalent to the input value by the character recognition module 130. The value may be stored. That is, since the character recognition by the character recognition module 130 has good reproducibility, even when the character recognition result matches the input result by the operator input module 140, the learning data can be used effectively. When an erroneous input is made in the operator input module 140, when the same character has not been erroneously entered in the past (when the character recognition result of the character recognition module 130 matches the input result of the operator input module 140) If there is, try to correct it using that case. Note that the learning DB 170 may be realized as a database and may function as a storage unit.

With reference to FIG. 2, the configuration of the entire system example in the case of realizing the present embodiment will be described.
As a whole configuration, the scan server 210 and the client PC 220 are connected by a communication line 230. The scan server 210 has a function of an image input device (scanner), receives an instruction from another device via the communication line 230, operates the image input device, or is input by the image input device. It has at least a function capable of transmitting image data to another device.
For example, the configuration of the present embodiment shown in FIG. 1 is built in the scan server 210, and an operator inputs from the client PC 220, and the input result is an operator in the scan server 210 via the communication line 230. It may be input to the input module 140. The scan server 210 includes an image input module 110, an area division module 120, and a character recognition module 130. The client PC 220 includes an operator input module 140, a comparison module 150, a correction module 160, and a learning DB 170. May be.
There may be a plurality of scan servers 210, client PCs 220, and communication lines 230.

An example of a data structure for handling the processing results obtained by the area dividing module 120 and the character recognition module 130 will be described with reference to FIG.
As shown in FIG. 3, the area division / character recognition result table 300 includes an ID column 302, an upper left coordinate column 304, a lower right coordinate column 306, a type column 308, a title column 310, an attribute value column 312, and a certainty factor column 314. Have. The ID column 302 to the type column 308 store the processing results by the area dividing module 120, and the title column 310, the attribute value column 312, and the certainty factor column 314 store the processing results by the character recognition module 130.

The ID column 302 stores a code (identifier) that can uniquely identify an area extracted in the form image. This area corresponds to an item in the form.
In the upper left coordinate column 304 and the lower right coordinate column 306, the positions of the areas divided by the area dividing module 120 in the form image are stored. For example, when the divided area is a rectangle, the image has a plane space with the vertical axis as the X axis and the horizontal axis as the Y axis, the upper left corner coordinates of the extraction area in the image are stored in the upper left coordinate column 304, and the lower right corner coordinates Is stored in the lower right coordinate field 306, the position of the extraction region in the image can be specified.
The type field 308 stores the type of the divided image (character area, graphic area, table area, photo area, etc.). The character area and the table area are to be processed after the character recognition module 130.

  The title column 310 includes item titles (title, title, title, heading, etc., and subtitles) from among the character recognition results by the character recognition module 130 for the area specified by the upper left coordinate column 304 and the lower right coordinate column 306. ) Is extracted as a character string. For example, titles include “payee”, “account item”, “summary”, “amount”, “total”, and the like. As processing, for example, a table (table, storage means) that stores titles of predetermined form items is prepared, and matching of the titles in the table and the character recognition result by the character recognition module 130 is performed. You may make it take. At the time of matching, when several characters (including one character) are different, it may be recognized as a title by determining whether or not the different characters are similar in character form.

The attribute value column 312 stores an attribute value corresponding to the title stored in the title column 310. For example, a specific company name corresponding to “payee” as a title. In addition, as processing, for example, a table storing titles of predetermined form items is prepared, and a character string that does not match a title in the table and a character recognition result by the character recognition module 130 is used as an attribute value. Also good.
In the certainty factor column 314, a value indicating the certainty factor (reliability) of the character code obtained as a result of the character recognition processing by the character recognition module 130 is stored. For example, as a method for obtaining the certainty factor, a method described in Japanese Patent No. 2991779 can be applied.

An example of a data structure for handling an input result by an operator by the operator input module 140 will be described with reference to FIG.
As shown in FIG. 4, the operator input table 400 has an ID column 402, an upper left coordinate range column 404, a lower right coordinate range column 406, a title column 408, and an attribute value column 410. ID column 402 to title column 408 are set in advance according to the form. The character string input by the operator input module 140 is stored in the attribute value column 410.
The ID column 402 stores a code (identifier) that can uniquely identify an item in the form.
In the upper left coordinate range column 404 and the lower right coordinate range column 406, coordinates that can specify the position of the item in the form are stored. The coordinate system here uses the same coordinate system as the upper left coordinate column 304 and the lower right coordinate column 306 of the region division / character recognition result table 300. However, the coordinates are values having a range, and this is for matching with the upper left coordinate column 304 and the lower right coordinate column 306 of the region division / character recognition result table 300. That is, if the coordinates of the upper left coordinate field 304 and the lower right coordinate field 306 of the region division / character recognition result table 300 are included in the upper left coordinate range field 404 and the lower right coordinate range field 406, they are equivalent items. to decide.
The title column 408 stores the titles of items in the form.
The attribute value column 410 stores attribute values input by the operator using the keyboard.

An example of a data structure that is stored in the learning DB 170 and handles processing results by the comparison module 150 and the correction module 160 will be described with reference to FIG.
As shown in FIG. 5, the learning table 500 includes an ID column 502, a title column 504, a pre-correction character recognition attribute value column 506, a device ID column 508, a pre-correction input attribute value column 510, an operator ID column 512, a post-correction attribute. A value column 514 and a corrector ID column 516 are provided.
The ID column 502 stores a code (identifier) that can uniquely identify the corrected character string.
The title column 504 stores the title of the item that has been corrected.
In the pre-correction character recognition attribute value field 506, the character string in the item recognized by the character recognition module 130 is stored.
The device ID column 508 stores an identifier that can uniquely identify the character recognition module 130. For example, an identifier such as a character recognition device that performs processing by the character recognition module 130 or a scanner is stored.
The pre-correction input attribute value column 510 stores the attribute value in the item input by the operator input module 140.
The operator ID column 512 stores an identifier that can uniquely identify the operator who has input through the operator input module 140.
The corrected attribute value column 514 stores the attribute value corrected by the correction module 160.
In the corrector ID column 516, when the correction is made by the correction module 160, the operator ID is stored if the correction is made by the operator. When the correction is made without the intervention of the operator, an identifier for uniquely identifying the correction module 160 is stored. Is done.

Next, the function and operation (operation) will be described.
A processing example by the image input module 110, the region division module 120, and the character recognition module 130 will be described with reference to FIG.
In step S602, a form, which is a paper document targeted by the image input module 110, is read as image data. Specifically, it may be read by a scanner, may be input via a communication line such as FAX, or may be read from a database storing form images.
In step S604, the region division module 120 performs region division on the image input in step S602. As a result, values are stored in the ID column 302 to the type column 308 of the area division / character recognition result table 300.
In step S606, the character recognition module 130 performs character recognition processing on the character area and the table area among the areas divided in step S604.
In step S608, the result of character recognition performed by the character recognition module 130 in step S606 is stored as a value in the title column 310 to the certainty factor column 314 of the area division / character recognition result table 300.
On the other hand, the operator input module 140 inputs the attribute value of each item, which is the contents of the form input by the image input module 110 by the operator, and stores it in the attribute value column 410 of the operator input table 400.

A processing example by the comparison module 150 and the correction module 160 will be described with reference to FIGS.
In step S <b> 702, the comparison module 150 inputs a character recognition result by the character recognition module 130. That is, the character recognition module 130 passes the area division / character recognition result table 300 to the comparison module 150.
In step S <b> 704, the comparison module 150 inputs the input result from the operator input module 140. That is, the operator input module 140 passes the operator input table 400 to the comparison module 150.
In step S706, the comparison module 150 determines whether or not the attribute values for the equivalent items match. If they match (if correction is not necessary), the process proceeds to step S716. If not (correction is required), the process proceeds to step S708.
In step S708, the correction module 160 displays the corresponding attribute value input in steps S702 and S704 on the screen. That is, as shown in FIG. 8, the character recognition result in step S702 is displayed in the character recognition result column 802, and the operator input result column 804 displayed side by side with the character recognition result column 802 is displayed in step S704. Displays the input value by the operator. For example, when the character recognition result is “link” and the input value by the operator is “contact”, “link” is displayed in the character recognition result column 802, and “contact” is displayed in the operator input result column 804. ”Is displayed. However, nothing is displayed in the result column 810 at the time of step S708.

In step S <b> 710, the comparison module 150 displays a correction value using the learning DB 170 in the result column 810. Here, the correction value may be displayed based only on the past correction history in the equivalent form stored in the learning DB 170. That is, there is a high possibility that the same error recognition and erroneous input in one form are repeated, and corrections made in another form may not be a reference for correction. For example, if the “link” is corrected as “contact” in the form and stored in the learning DB 170, the “link” in the character recognition result is not adopted in the above example. The “contact” that is the correction value stored in the learning DB 170 may be adopted and displayed in the result column 810.
In step S712, the operator determines whether or not the attribute value displayed in the result column 810 is correct while looking at the form that is a paper document. If the attribute value is correct, the operator depresses the registration button 812. The attribute value displayed in the result column 810 is corrected using the keyboard. Alternatively, if it is desired to use the attribute values displayed in the character recognition result field 802 and the operator input result field 804 at the time of correction, the selection button 806 or the selection button 808 is pressed to change any attribute value. You can choose. For example, in the above-described example, because “correction form” that is the correct attribute value is already displayed in the result column 810 in step S710, the operator proceeds to the next step S714 when the registration button 812 is pressed. Also, if the “link” has not been corrected as “contact” in the past and there is no equivalent pattern in the learning DB 170, the correction using the learning DB 170 in step S710 is not performed. When the selection button 808 is pressed, the “contact form” that is an input value by the operator is adopted as the correct attribute value, the “contact form” that is the attribute value is displayed in the result column 810, and the registration button 812 is pressed by the operator. The process proceeds to the next step S714.
In step S714, the correction result is registered in the learning DB 170. That is, one line is added to the learning table 500. For example, in the above-mentioned example, “concatenation” is displayed in the character recognition attribute value field 506 before correction, “contact” is input in the input attribute value field 510 before correction, and “contact” is displayed in the attribute value field 514 after correction. Remembered.
In step S716, it is determined whether or not the confirmation and correction work has been completed for all data in the form. If completed, the process proceeds to step S718 (end). If not, the process returns to step S702, and the processes after step S702 are repeated for the next item.

Further, in the flowchart shown in FIG. 7, the process proceeds to step S716 when they coincide with each other in step S706. However, the process may proceed to step S714. That is, even if the character recognition result by the character recognition module 130 and the input result by the operator input module 140 are equivalent, they may be registered in the learning DB 170.
In step S710, the operator ID column 512 of the learning table 500 stored in the learning DB 170 is searched using the input result from the character recognition module 130 as a search keyword, and the attribute value in the searched operator ID column 512 is searched. You may make it correct with the value of the attribute value column 514 after correction corresponding to.

  With reference to FIG. 9, a hardware configuration example of the present embodiment will be described. The configuration illustrated in FIG. 9 is configured by, for example, a personal computer (PC), and illustrates a hardware configuration example including a data reading unit 917 such as a scanner and a data output unit 918 such as a printer.

  A CPU (Central Processing Unit) 901 describes execution sequences of various modules described in the above-described embodiment, that is, the area division module 120, the character recognition module 130, the comparison module 150, the correction module 160, and the like. The control unit executes processing according to the computer program.

  A ROM (Read Only Memory) 902 stores programs used by the CPU 901, operation parameters, and the like. A RAM (Random Access Memory) 903 stores programs used in the execution of the CPU 901, parameters that change as appropriate during the execution, and the like. These are connected to each other by a host bus 904 including a CPU bus.

  The host bus 904 is connected to an external bus 906 such as a PCI (Peripheral Component Interconnect / Interface) bus via a bridge 905.

  A keyboard 908 and a pointing device 909 such as a mouse are input devices operated by an operator. The display 910 includes a liquid crystal display device or a CRT (Cathode Ray Tube), and displays various types of information as text and image information.

  An HDD (Hard Disk Drive) 911 incorporates a hard disk, drives the hard disk, and records or reproduces a program executed by the CPU 901 and information. The hard disk stores input form images, data such as a learning table 500, and the like. Further, various computer programs such as various other data processing programs are stored.

  The drive 912 reads out data or a program recorded in a removable recording medium 913 such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, and the data or program is read out from the interface 907 and the external bus 906. , The bridge 905, and the RAM 903 connected via the host bus 904. The removable recording medium 913 can also be used as a data recording area similar to a hard disk.

  The connection port 914 is a port for connecting the external connection device 915, and has a connection unit such as USB, IEEE1394. The connection port 914 is connected to the CPU 901 and the like via the interface 907, the external bus 906, the bridge 905, the host bus 904, and the like. The communication unit 916 is connected to a network and executes data communication processing with the outside. The data reading unit 917 is a scanner, for example, and executes document reading processing. The data output unit 918 is, for example, a printer, and executes document data output processing.

  Note that the hardware configuration of the information processing system shown in FIG. 9 shows one configuration example, and the information processing system of the present embodiment is not limited to the configuration shown in FIG. 9 and is described in the present embodiment. Any configuration can be used as long as the module can be executed. For example, some modules may be configured by dedicated hardware (for example, ASIC), and some modules may be in an external system and connected via a communication line. A plurality of systems shown in FIG. 5 may be connected to each other via communication lines so as to cooperate with each other. Further, it may be incorporated in a copying machine, a fax machine, a scanner, a printer, a multifunction machine (also called a multi-function copying machine, which has functions of a scanner, a printer, a copying machine, a fax machine, etc.).

In the above embodiment, a form is shown as an image including a character image to be recognized. However, the form may be other than a form, for example, a document image.
A plurality of character recognition modules 130 may be provided. Each character recognition module 130 may have a different character recognition algorithm or character recognition dictionary. In that case, the character recognition result may be determined by majority vote.
A plurality of operator input modules 140 may be provided. An input by a different operator may be performed. Similarly, the input result may be determined by majority vote.
Further, the correction module 160 may perform correction using the certainty factor stored in the certainty factor column 314 in the region division / character recognition result table 300. That is, the correction may be performed only when the certainty factors are similar.

The described program can be stored in a recording medium, and the program can be provided by communication means. In that case, for example, the above-described program can also be regarded as an invention of a “computer-readable recording medium recording the program”.
The “computer-readable recording medium on which a program is recorded” refers to a computer-readable recording medium on which a program is recorded, which is used for program installation, execution, program distribution, and the like.
The recording medium is, for example, a digital versatile disc (DVD), which is a standard established by the DVD Forum, such as “DVD-R, DVD-RW, DVD-RAM,” and DVD + RW. Standards such as “DVD + R, DVD + RW, etc.”, compact discs (CDs), read-only memory (CD-ROM), CD recordable (CD-R), CD rewritable (CD-RW), etc. MO), flexible disk (FD), magnetic tape, hard disk, read only memory (ROM), electrically erasable and rewritable read only memory (EEPROM), flash memory, random access memory (RAM), etc. It is.
The program or a part of the program can be recorded on the recording medium and stored or distributed. Also, by communication, for example, a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a wired network used for the Internet, an intranet, an extranet, etc., or wireless communication It can be transmitted using a transmission medium such as a network or a combination of these, and can also be carried on a carrier wave.
Furthermore, the above program may be a part of another program, or may be recorded on a recording medium together with a separate program. Moreover, it may be divided and recorded on a plurality of recording media.

It is a conceptual module block diagram about the structural example of this Embodiment. It is a block diagram of the whole system example in the case of implement | achieving this Embodiment. It is an example of the data structure of the area | region division process result and character recognition process result which this Embodiment handles. It is an example of the data structure of the operator input which this Embodiment handles. It is an example of the data structure of the learning data which this Embodiment handles. It is an example of a flowchart which shows processes, such as character recognition by this Embodiment. It is an example of a flowchart which shows the correction process by this Embodiment. It is an example of a screen display in the correction process by this Embodiment. It is a block diagram which shows the hardware structural example of the computer which implement | achieves this Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 110 ... Image input module 120 ... Area division module 130 ... Character recognition module 140 ... Operator input module 150 ... Comparison module 160 ... Correction module 170 ... Learning DB
210: Scan server 220 ... Client PC
230 ... Communication line 300 ... Area division / character recognition result table 400 ... Operator input table 500 ... Learning table

Claims (9)

A comparison means for comparing a first input value obtained by an operator's operation on an equivalent item with a second input value obtained by a character recognition process;
Storage means for associating and storing past first input values, past second input values for equivalent items corresponding to the first input values, and corrected values for equivalent items;
When the first input value and the second input value are different from each other by the comparison by the comparison means, a correction means for correcting based on the contents stored in the storage means is provided. Information processing system. The storage means stores the first input value or the second input value when the first input value and the second input value are equal by the comparison by the comparison means,
When the first input value is different from the second input value, the correcting means searches for the second input value stored in the storage means based on the second input value, and is searched. The information processing system according to claim 1, wherein the information is corrected by a corrected value corresponding to the second input value. The information processing system according to claim 1, wherein the correction unit corrects based on past corrections in an equivalent document stored in the storage unit. The character recognition process also outputs a character recognition result and a certainty factor by the character recognition process,
The storage means further stores a certainty factor by the character recognition process,
The information processing system according to claim 1, wherein the correction unit corrects according to a certainty factor stored in the storage unit. An image input means for inputting an image;
Character recognition processing means for recognizing characters of the image input by the image input means;
A character input means for inputting characters in the image by an operation of an operator;
Comparing means for comparing the first input value by the character input means for the equivalent item with the second input value by the character recognition processing means;
Storage means for associating and storing past first input values, past second input values for equivalent items corresponding to the first input values, and corrected values for equivalent items;
When the first input value and the second input value are different by the comparison by the comparison means, a first correction means for correcting based on the storage means;
An information processing system comprising: second correction means for correcting a result of correction by the first correction means by an operation of an operator. Computer
A comparison means for comparing a first input value obtained by an operator's operation on an equivalent item with a second input value obtained by a character recognition process;
Storage means for associating and storing past first input values, past second input values for equivalent items corresponding to the first input values, and corrected values for equivalent items;
When the first input value and the second input value are different by the comparison by the comparison unit, the comparison unit functions as a correction unit that corrects the content based on the contents stored in the storage unit. Information processing program. The storage means stores the first input value or the second input value when the first input value and the second input value are equal by the comparison by the comparison means,
When the first input value is different from the second input value, the correcting means searches for the second input value stored in the storage means based on the second input value, and is searched. The information processing program according to claim 6, wherein the information is corrected by a corrected value corresponding to the second input value. The information processing program according to claim 6, wherein the correction unit corrects based on past corrections in an equivalent document stored in the storage unit. The character recognition process also outputs a character recognition result and a certainty factor by the character recognition process,
The storage means further stores a certainty factor by the character recognition process,
The information processing program according to claim 6, wherein the correction unit corrects the degree of certainty stored in the storage unit.

Images