JP2019117520A - Character recognition device, method and program - Google Patents

Abstract

To provide a character recognition device, a method and a program capable of performing highly accurate text data generation for each reading item of a document containing various handwritten characters.SOLUTION: There is provided a character recognition device 200 comprising a control part 230 including a reading range setting part 231 performing setting of a reading range in sample image data by a user operation, a reading item determination part 232 calculating a coordinate position of the reading range in the sample image data to determine a reading item, an attribute information adding part 233 adding predetermined attribute information by a user operation, and a text data generation part 234 generating text data from character information displayed on the reading item determined by the reading item determination part 232 based on the attribute information added by the attribute information adding part 233.SELECTED DRAWING: Figure 2

Description

  The present disclosure relates to a character recognition apparatus, method, and program for reading character information from image data.

  2. Description of the Related Art A technology for generating digital data in which input information is converted into a predetermined character code by using an image scanner or the like to read a document in which characters are written by handwriting and performing an OCR (Optical Character Recognition) process has become widespread.

  For example, according to Patent Document 1, a character identification system for performing character recognition by machine learning is disclosed. The character recognition system includes a character image input reception unit that receives an input of a sample character image, character part extraction that extracts character parts based on the sample character image, and pseudo character model generation that generates a pseudo character model based on the character parts There has been proposed a system configured of a part and an identification dictionary generation that generates a character identification pattern based on a pseudo character model and generates an identification dictionary.

  Also, in recent years, a service called Business Process Outsourcing (BPO) has been provided that outsources part of the company's business process to an external company. For example, in the BPO service, handwritten application forms and account transfer request documents There is a form such as entrusting an external company the process of converting text into text. In this mode, an application form is read by an image scanner or the like and converted into image data, and the converted image data is transmitted to an external company via a network such as the Internet.

Unexamined-Japanese-Patent No. 2015-069256

  However, in such a BPO service, since the image data is ordered from the external company as it is, the external company can not set the reading item. For example, when performing OCR processing on handwritten data, identification may be difficult, such as the number “1” and the small letter “l” (el), but such handwritten characters are entered For example, in the case of telephone numbers, there may be cases where only letters can be entered in the entry fields, such as telephone numbers, and identification accuracy may be improved if the type of letters can be limited for each input item. .

  Furthermore, when the written-in character is a name, etc., an error is usually not permitted. In such a case, it may be more efficient because human errors can be prevented if manually input. Therefore, there is a demand for performing character recognition by OCR by limiting the type of character for each entry column, or enabling manual input by human and character recognition by OCR.

  Therefore, in the present disclosure, a reading item is determined from sample image data of image data of a document including various handwritten characters, and whether or not character recognition by the character type and the number of characters, OCR is performed for each reading item, etc. A character recognition device, method and program capable of generating text data with high accuracy for each read item by adding attribute information of

  The character recognition device according to one aspect of the present disclosure is a character recognition device that reads character information from image data obtained by scanning a document as an image, and is operated by the user from sample image data obtained by scanning a sample of the document as an image. The read range setting unit for setting the read range of the document, the read item determination unit for determining the read item of the document from the coordinate position of the read range in the sample image data, and predetermined attributes by the user operation for the read item. An attribute information addition unit for adding information, and a text data generation unit for generating, from the image data, text information displayed on the read item of the image data as text data based on the read item and the attribute information.

  A character recognition method according to an aspect of the present disclosure is a character recognition method for reading character information from image data obtained by scanning a document as an image, and is read from sample image data obtained by scanning a sample of the document as an image by user operation. A reading range setting step of setting a reading range of the document performed by the range setting unit; a reading item determination step of determining a reading item of the document performed by the reading item determination unit from the coordinate position of the reading range in the sample image data; An attribute information addition step of adding predetermined attribute information performed by the attribute information addition unit to the read item by the operation of the user, and an image data of the image data performed by the text data generation unit based on the read item and the attribute information A text data generation step of generating the text information displayed in the read item as text data; .

  A character recognition program according to one aspect of the present disclosure is a character recognition program that reads character information from image data obtained by scanning a document as an image, and is operated by the user from sample image data obtained by scanning a sample of the document as an image. The reading range setting step of setting the reading range of the document, the reading item determining step of determining the reading item of the document from the coordinate position of the reading range in the sample image data, and the reading item An attribute information adding step for adding attribute information of the above, and a text data generation step for generating text information displayed in the read item of the image data as text data based on the read item and the attribute information from the image data Run on a computer.

  According to the present disclosure, it is possible to perform highly accurate text data generation for each read item for image data of a document including various handwritten characters.

1 is a block diagram illustrating a character recognition system according to an embodiment of the present disclosure. It is a functional block diagram which shows an example of a structure of the character recognition apparatus shown in FIG. It is a flowchart which shows the character recognition method by the character recognition system shown in FIG. FIG. 6 is a schematic diagram showing a bank transfer request document that is an example of a document to be scanned. It is a schematic diagram which shows the account transfer request document (sample) which is an example of the sample of the document scanned. It is a schematic diagram which shows the example of reading range setting. It is a schematic diagram which shows the example of the attribute information added to the reading range of a reading place of FIG. 6, and a financial institution code reading range. It is a figure which shows the example of a financial institution table. It is a figure which shows the example of the produced | generated text data. It is a functional block diagram showing an example of composition of a character recognition device concerning one embodiment of this indication. It is a flowchart which shows the character recognition method by the character recognition system provided with the character recognition apparatus shown in FIG. It is a schematic diagram which shows the example of the attribute information added to the name reading range of FIG. 6, and a financial institution code reading range.

  Embodiments of the present disclosure will be described with reference to the drawings. Note that the embodiments described below do not unduly limit the content of the present disclosure described in the claims. In addition, not all of the components shown in the embodiment are necessarily essential components of the present disclosure.

(Embodiment 1)
<Configuration>
FIG. 1 is a block diagram of a character recognition system 1 according to a first embodiment of the present disclosure. The character recognition system 1 is a system that reads character information from image data obtained by scanning a document such as a handwritten application form or a bank transfer request as an image, for example, and receives an application form or a bank transfer request from a customer Used by a company (a client company) to outsource to an external company (a client company) that provides BPO services.

  The character recognition system 1 includes an image data conversion device 100, a character recognition device 200, and a network NW. The image data conversion apparatus 100 and the character recognition apparatus 200 are connected via the network NW. The network NW is configured by the Internet, a LAN (Local Area Network), a WAN (Wide Area Network), and the like. The network NW may be wired communication or wireless communication, and includes a communication method called 4G such as LTE (Long Term Evolution) and a communication method by 5G.

  The image data conversion apparatus 100 is installed at the ordering company, and documents such as application forms and fund transfer request documents composed of a plurality of items are scanned and converted into image data, and the converted image data is stored. Do. The image data conversion device 100 includes a scanner device 110, a server device 120, and a user terminal 130, and is communicably connected to one another by, for example, USB (registered trademark) or LAN.

  The scanner device 110 is a device that scans documents such as application forms and account transfer request documents into image data, and scans documents that become samples of these documents into sample image data. Although the scanner device is used in the first embodiment, any device that can convert a paper medium into electronic data may be used. For example, a camera may be used.

  The server apparatus 120 is a server that manages the business of the ordering company. Further, the server device 120 includes an image data DB 121 storing image data converted by the scanner device 110 and sample image data, and a read item DB 122 storing a read item determined by the character recognition device 200 from the sample image data. There is. The image data DB 121 and the read item DB 122 are databases that can be accessed from the character recognition device 200 and can be referenced and updated.

  The user terminal 130 is a terminal installed in the company on the ordering side, and is a terminal for setting a reading range from the sample image data by the operation of the user. The user terminal 130 includes a display unit for displaying sample image data, and can set a reading range by operating the operation unit. The display unit includes a display or the like, and the operation unit is a keyboard or a mouse. Etc.

  FIG. 2 is a functional block diagram showing an example of the configuration of the character recognition device shown in FIG. The character recognition device 200 is installed in the company on the order receiving side, determines the read item on the sample image data based on the read range set by operating the user terminal 130, and sets the read item of the image data as the read item. Generates the displayed text information as text data. The character recognition device 200 includes a communication unit 210, a storage unit 220, and a control unit 230.

  The communication unit 210 is a communication interface for communicating with the image data conversion apparatus 100, and communication is performed according to a communication protocol such as TCP / IP (Transmission Control Protocol / Internet Protocol).

  The storage unit 220 stores programs for executing various control processing and each function in the control unit 230, input data, and the like, and includes a RAM (Random Access Memory), a ROM (Read Only Memory), and the like. Ru. The storage unit 220 also temporarily stores the read item determined from the sample image data and the generated text data, and also stores the result of reading the handwritten information in the past as learning information.

  The control unit 230 controls the overall operation of the character recognition apparatus 200 by executing a program stored in the storage unit 220, and is controlled by a central processing unit (CPU), a graphics processing unit (GPU) or the like. Configured As functions of the control unit 230, a read range setting unit 231, a read item determination unit 232, an attribute information addition unit 233, and a text data generation unit 234 are provided. The read range setting unit 231, the read item determination unit 232, the attribute information addition unit 233, and the text data generation unit 234 are activated and executed by a program stored in the storage unit 220.

  The reading range setting unit 231 accesses the server apparatus 120 via the communication unit 210 by operating the user terminal 130, reads sample image data stored in the image data DB 121, and displays the image data on the display unit of the user terminal 130. indicate.

  Further, the reading range setting unit 231 sets the reading range in the sample image data by operating the operation unit of the user terminal 130. The setting of the reading range is performed by so-called drag and drop range setting when the operation unit (for example, a mouse) is operated on the display unit on which the sample image data is displayed.

  The read item determination unit 232 calculates the coordinate position of the read range in the sample image data, for example, the coordinate positions of the four corners of the range by dragging and dropping from the read range performed by the read range setting unit 231, and reads the inside of the read item Decide. The read item is accessed by the server device 120 and stored in the read item DB 122.

  The attribute information addition unit 233 selects, for example, an item name (name, telephone number, etc.), a generation condition (select from a predetermined table, or the like) for the read item determined by the read item determination unit 232 by user operation. Attribute information such as character type conditions (Kanji + Hiragana + Katakana, only numbers, only alphabets, etc.) is added. The attribute information is accessed by the server device 120 and stored in the read item DB 122 for each read item.

  The text data generation unit 234 has an OCR function, reads out the image data stored in the image data DB 121, and based on the attribute information added by the attribute information addition unit 233, the read item determined by the read item determination unit 232 The character information displayed in is read by OCR to generate text data.

<Flow of processing>
Hereinafter, an example of the character recognition method performed by the character recognition system 1 will be described with reference to FIG. FIG. 3 is a flowchart showing a character recognition method by the character recognition system 1 shown in FIG.

  As the process of step S101, the image data conversion apparatus 100 uses the scanner device 110 to scan a sample document serving as a sample of a document such as an application form or a bank transfer request document, convert it into sample image data, and convert it. The image data is stored in the image data DB 121.

  FIG. 4 is a schematic diagram showing an account transfer request document, which is an example of a document, which is the source of the sample document scanned in step S101. The account transfer request form is an application form for requesting a financial transfer from a financial institution for the purpose of debiting public utility charges, etc., and it includes a reading field A1, a name entry field A2, a financial institution name entry field A3, A branch name entry column A4, a financial institution code entry column A5, a branch code entry column A6, a deposit item entry column A7, and an account number entry column A8 are provided.

  As shown in FIG. 4, the client of the ordering company enters his / her name etc. in these fields and requests the ordering company to transfer money. The ordering company receives the completed account transfer request form, performs scanning using the scanner device 110, and stores the image data in the image data DB 121.

  FIG. 5 is a schematic view showing a bank transfer request document (sample) which is an example of a sample document scanned in step S101. This account transfer request document (sample) is a sample similar to the account transfer request document shown in FIG. 4 and, like the account transfer request document, a kurigana entry field A11, a name entry field A12, a financial institution name entry field A13, A branch name entry field A14, a financial institution code entry field A15, a branch code entry field A16, a deposit item entry field A17, and an account number entry field A18 are provided. This sample is used to determine what to read when reading the letters of the account transfer request by OCR.

  When consigning the process of converting the bank transfer request into a text to the order receiving company, the ordering company scans this bank transfer request (sample) as a sample using the scanner device 110 and generates sample image data. It is stored in the image data DB 121.

  In the character recognition device 200, the server device 120 is accessed by the reading range setting unit 231 via the communication unit 210 by operating the operation unit of the user terminal 130 as the process of step S102, and is stored in the image data DB 121. The sample image data being read out is read out and displayed on the display unit of the user terminal 130. The operation unit of the user terminal 130 is operated by the operation of the user, and the setting of the reading range in the sample image data is performed.

  FIG. 6 is a schematic view showing an example of reading range setting in step S102. As shown in FIG. 6, the display unit displays sample image data of the bank transfer request document (sample) shown in FIG. In the same way as for the fund transfer request (sample), entry fields A21, names A22, financial institutions name A23, branch names A24, financial institution codes A25, branch codes A26, deposit items A26 A field A27 and an account number entry field A28 are displayed.

  In this state, drag and drop is performed by operating the operation unit of the user terminal 130, and range setting is performed as indicated by a broken line in FIG. For example, the reading range S1 for reading, name reading range S2, financial institution name reading range S3, branch name reading range S4, financial institution code reading range S5, branch code reading range S6, and account number reading range S7 are set as the reading range. .

  As the process of step S103, in the character recognition device 200, the coordinate position of the four corners of the reading range in the sample image data is calculated by the reading item determination unit 232, and the inside is determined as the reading item. For example, with the coordinates of the upper left point P1 of the sample image data shown in FIG. 6 as (0, 0), relative coordinates from the point P1 are calculated in dot units for the four corners of the reading range. The read range determined by this is accessed by the server device 120 via the communication unit 210 and stored in the read item DB 122.

  As the process of step S104, in the character recognition device 200, attribute information such as an item name, a generation condition, and a character type condition by the user operation on the read item determined by the attribute information addition unit 233 in step S103. Is added. The added attribute information is accessed by the server device 120 via the communication unit 210 and stored in the read item DB 122.

  FIG. 7 is a schematic view showing an example of attribute information added to the kana reading range S1 and the financial institution code reading range S5 in FIG. 6 in step S104. For example, with respect to the kana reading range S1 shown in FIG. 6, as in kichigana attribute information T1 shown in FIG. 7, for the item name column T11, the generation condition column T12 and the character type column T13, respectively, "," "Katakana" is input. Also, for the financial institution code reading range S5, as in the financial institution code attribute information T2, “Financial institution code” and “table selection” for the item name column T21, generation condition column T22, and character type column T23, respectively. "4 digits" is entered.

  In the item name fields T11 and T21, item names are input as they are. The generation condition fields T12 and T22 can select, for example, either “OCR read” or “table selection”, and “OCR read” indicates that OCR is read by an OCR and is converted into text, “table selection “Indicates that the financial institution table shown in FIG. 8 is selected. The character type column T13, T23 can select, for example, "Katakana", "Numeral 4 digit", etc. Since "Katakana" is only filled with Katakana in the reading range S1, the type of the character to be read is also Katakana only. Because "number four digits" is only entered in the financial institution code reading range S5 and the number of digits is also determined to be four digits, the type of characters to be read may be only four digits. It shows. In addition, for example, in the case where a grid is provided for each digit as in financial institution code reading range S5, branch code reading range S6, and account number reading range S7 (numeral with numbers), It may be distinguished as in the case where no mass is provided (number without numbers).

  FIG. 8 is a diagram showing an example of a financial institution table. The financial institution table is stored, for example, in the server device 120. Only the financial institution code and the financial institution name stored in the financial institution table are input to the financial institution code reading range S5 and the financial institution name reading range S3 in FIG. Therefore, the reading of the financial institution code reading range S5 may be selected from the financial institution code of this financial institution table.

  The condition for reading financial institution name reading range S3 may be either “OCR read” or “table selection”. For example, “OCR read” is usually made and read by OCR and text conversion is performed, and handwritten characters are read If it can not be done, the financial institution name of the financial institution table corresponding to the financial institution code read from the financial institution code reading range S5 may be acquired.

  Further, the same applies to the branch name reading range S4 and the branch code reading range S6, and the branch code table is stored in the server apparatus 120 and processed as in the financial institution name reading range S3 and the financial institution code reading range S5. good.

  As the process of step S105, the image data conversion apparatus 100 uses the scanner device 110 to scan a document such as the application form and the account transfer request form shown in FIG. 4 and convert it into image data and convert it into a sample image. Data is stored in the image data DB 121. In this state, the ordering company indicates the image data and the sample image data in the image data DB 121 and consigns it to the accepting company.

  As the process of step S106, in the character recognition device 200, the text data generation unit 234 reads the kana reading range S1-account based on the kana attribute information T1, the financial institution code attribute information T2, ... added in step S105. The character information displayed in the number reading range S7 is read by OCR to generate text data.

  FIG. 9 is a diagram showing an example of text data generated in step S106. For example, as shown in FIG. 4, it is described in a hurigana entry field A1, a name entry field A2, a financial institution name entry field A3, a branch name entry field A4, a financial institution code entry field A5, a branch code entry field A6 and an account number entry field A8. The “TOKYO Taro”, “patent Taro”, “0001”, “Mizuho”, “021”, “Marunouchi”, and “1111111” that have been read are respectively read and generated as text data.

  As the process of step S107, in the character recognition device 200, the text data generated in step S106 is transmitted to the server apparatus 120 via the communication unit 210 as delivery to the ordering company.

  As described above, in the character recognition system according to the present embodiment, when the reading range is set on the sample image data by dragging and dropping, the coordinate position of the reading range in the sample image data is calculated and the reading item is determined. Text data is generated according to this read item. As a result, it becomes possible for the order-receiving company to determine the read item to be subjected to the character recognition by the OCR.

  In addition, since attribute information such as the character type, the number of characters, and whether to read by OCR or to select from a table can be added to each read item, unnecessary character recognition can be eliminated. It becomes possible to perform highly accurate text data generation for each.

Second Embodiment
<Configuration>
FIG. 10 is a functional block diagram showing an example of the configuration of the character recognition device according to the second embodiment of the present disclosure. In the character recognition device 200, a text data input unit 235 and a text data determination unit 236 are added as functions of the control unit 230 to the same configuration as that of the first embodiment.

  The text data input unit 235 reads out the image data stored in the image data DB 121 and displays it on the display unit, and inputs text data by the operation of the operation unit of the user terminal 130. This is to ensure the accuracy of the input information, since it is possible to prevent an error by human input if errors are not permitted in the process of converting the account transfer request into a text. Moreover, in order to perform double check, it is also possible to perform reading by OCR and also to manually input text data by the user. In such a case, it is a function for enabling manual input of text data by the user.

  As described above, the text data determination unit 236 performs double-checking when reading by OCR and performing manual input of text data by the user. This is to prevent errors by double checking if errors are not permitted in the process of converting the account transfer request and the like into text, and further secure the accuracy of the input information. Then, when the read content by the OCR and the content of the manual input of the text data by the user do not match, the user is warned that the input field is displayed in red, for example. The other configuration is the same as that of the first embodiment.

<Flow of processing>
Hereinafter, processing different from that of the first embodiment will be described as an example of the character recognition method performed by the character recognition system 1 with reference to FIG. FIG. 11 is a flow chart showing a character recognition method by the character recognition system 1 provided with the character recognition device 200 shown in FIG. The steps up to step S103 shown in FIG. 3 are the same as in the first embodiment, so the following step S104 will be described.

  The process of step S104 is similar to that of the first embodiment in that attribute information is added to the read item and stored in the read item DB 122, but the added attribute information is different from that of the first embodiment.

  FIG. 12 is a schematic view showing an example of attribute information added to the name reading range S2 and the financial institution code reading range S5 of FIG. 6 in step S104. For example, in the name reading range S1 shown in FIG. 6, like the kana attribute information T3 shown in FIG. 12, in addition to the item name column T31, the generation condition column T32 and the character type column T33 similar to the first embodiment In the condition column T34, "name", "-", "kanji + hiragana + katakana", and "OCR + entry" are input. Also, for the financial institution code reading range S5, as in the financial institution code attribute information T4, "Financial institution code" for the item name column T41, the generation condition column T42, the character type column T43, and the generation presence / absence condition column T44. , "Table selection", "4 digits", "OCR" are input.

  For example, one of “entry”, “OCR + entry”, and “OCR” can be selected in the generation presence / absence condition fields T34 and T44. The “entry” is a text data hand by the user without reading by OCR. "OCR + entry" indicates reading by OCR, and then, the user manually inputs text data, double-checks both input contents, and "OCR" reads by OCR Indicates to do.

  As in the process of step S105, the document is scanned as in the first embodiment.

  As the process of step S106, in the character recognition device 200, the text data generation unit 234 reads the kana reading range S1-account based on the kana attribute information T1, the financial institution code attribute information T2, ... added in step S104. The character information displayed in the number reading range S7 is read by OCR to generate text data. At this time, when the condition of generation of the attribute information is "entry", generation of text data is not performed.

  As the process of step S108, in the character recognition device 200, the text data input unit 235 reads out the image data stored in the image data DB 121 and displays it on the display unit. An input column of an item in which “entry” and “OCR + entry” are added to the attribute information generation presence / absence condition column is displayed on the display unit of the user terminal 130, and text data is input by the operation of the operation unit.

  As the process of step S109, in the character recognition device 200, the text data determination unit 236 adds the OCR read content by the user and the manual input of the text data by the user for the item in which "OCR + entry" is added to the generation information presence / absence condition column. Double check is performed with the contents of. And when both the input contents correspond, alerting is not performed, when the both input contents do not match, the input column of the display part of the user terminal 130 is displayed in red, and alerting is carried out. The flow of processing after this is the same as that of the first embodiment.

  According to the present embodiment, in addition to the effects of the first embodiment, the user manually inputs text data by the user for each reading item, or performs both reading by OCR and manual input of text data by the user. Since it has become possible to add attribute information on whether to perform a check or to perform reading by OCR and to enable manual input of text data, it is possible to prevent an error by manual input by each person, it is possible to read each item It is possible to generate text data with high accuracy.

  As another embodiment, it is also possible to connect the scanner device directly to the character recognition device and configure the character recognition device in a stand-alone manner without going through a network. This is effective when it is desired to configure the character recognition device without connecting it to the network in order to prevent information leakage to the outside.

  Although the embodiments according to the disclosure have been described above, they can be implemented in other various forms, and can be implemented with various omissions, substitutions, and modifications. Those embodiments and modifications, and omissions, substitutions and changes are included in the technical scope of the claims and the equivalents thereof.

1 character recognition system, 100 image data conversion device, 110 scanner device, 120 server device, 121 image data DB, 122 reading item DB, 200 character recognition device, 210 communication unit, 220 storage unit, 230 control unit, 231 reading range setting Part 232 Reading item determination part 233 Attribute information addition part 234 Text data generation part 235 Text data input part 236 Text data judgment part NW network

The character recognition device according to one aspect of the present disclosure is a character recognition device that reads character information from image data obtained by scanning a document as an image, and is operated by the user from sample image data obtained by scanning a sample of the document as an image. The reading range setting unit that sets the reading range of the document, the reading item determination unit that determines the reading item of the document from the coordinate position of the reading range in the sample image data, and the reading item To indicate whether the text data is generated from the character information, the user is allowed to input without generating the text data from the character information, or the text data is generated from the character information to be also input by the user and attribute information adding unit for adding a predetermined attribute information including, from the image data, on the basis of the read items and attribute information, the image data If the text data generator for generating character information of the read items are displayed as text data, generating whether conditions read item is performed also input by the user by generating text data from the text information, generates the character information It is determined whether or not the input text data matches the input data from the user, and a text data determination unit that issues a warning when they do not match .

A character recognition method according to an aspect of the present disclosure is a character recognition method for reading character information from image data obtained by scanning a document as an image, and is read from sample image data obtained by scanning a sample of the document as an image by user operation. A reading range setting step of setting a reading range of the document performed by the range setting unit; a reading item determination step of determining a reading item of the document performed by the reading item determination unit from the coordinate position of the reading range in the sample image data; For the read item, the attribute information addition unit performs by the operation of the user, generates text data from character information for each read item, or allows the user to input without generating text data from character information, or character information It is added a predetermined attribute information including the generated existence condition indicating also either perform input by the user by generating text data from And attribute information adding step, the image data, and text data generating step of generating text data generating unit is performed based on the read item and attribute information, a character information displayed in the read items of image data as text data, read item Whether the text data generated by the text data determination unit matches the text data generated by the text information with the input content by the user or not, if the text generation determination condition generates text data from the text information and the user also inputs And a text data determination step of giving a notice if it does not match .

A character recognition program according to one aspect of the present disclosure is a character recognition program that reads character information from image data obtained by scanning a document as an image, and is operated by the user from sample image data obtained by scanning a sample of the document as an image. Accordingly, the reading range setting step sets the reading range of the document, from the coordinate position of the read range in the sample image data, a reading item determining step of determining the read item of the document, to read item by operation of the user, reading For each item, it indicates whether to generate text data from character information, let the user input without generating text data from character information, or generate text data from character information and also input by the user an attribute information addition step of adding a predetermined attribute information including the existence condition, the image data, read Based on the item and attribute information, and the text data generation step of generating character information displayed in the read items of image data as text data, generating existence conditions of reading items, the user generates the text data from the character information When the input is also performed, it is determined whether or not the text data generated from the character information and the input content by the user match, and the computer is made to execute a text data determination step to call attention when they do not match .

Claims (9)

A character recognition apparatus for reading character information from image data obtained by scanning a document as an image,
A reading range setting unit configured to set a reading range of the document by a user operation from sample image data obtained by scanning the sample of the document as an image;
A reading item determination unit that determines a reading item of the document from the coordinate position of the reading range in the sample image data;
An attribute information adding unit that adds predetermined attribute information to the read item by a user operation;
And a text data generation unit that generates, from the image data, text information displayed in the read item of the image data as text data, based on the read item and the attribute information.
Character recognition device. The attribute information generates the text data from the character information for each reading item, or allows the user to input the text data without generating the text data from the character information, or the text data from the character information Including a generation presence / absence condition indicating whether to generate the input by the user and
The text data generation unit generates the text data from the character information or the generation condition of the reading item generates the text data from the character information and also inputs the user by the user. The character recognition device according to claim 1, which generates text data.   When the generation presence / absence condition of the reading item generates the text data from the character information and also inputs by the user, the text data generated from the character information matches the input content by the user The character recognition device according to claim 2, further comprising a text data determination unit that determines whether or not there is a match and warns when there is no match. The attribute information includes a generation condition indicating whether the character information is selected from registration information registered in a predetermined table, or whether the character information is read by OCR.
The character recognition device according to any one of claims 1 to 3, wherein the text data generation unit generates the text data based on the generation condition.   The one read item and the other read item are associated, and the generation condition of the one read item reads the character information by OCR, and the generation condition of the other read item is read from the registration information In the case of selection, when the character information of the one read item can not be read, the text data generation unit determines the character information of the one read item from the registration information selected in the other read item. The character recognition device according to claim 4, which determines The attribute information includes a character type condition specifying a character type of the character information,
The character recognition device according to any one of claims 1 to 5, wherein the text data generation unit generates the text data based on the character type condition. The attribute information includes a condition for specifying the number of characters of the character information, and
The character recognition device according to any one of claims 1 to 6, wherein the text data generation unit generates the text data based on the character number condition. A character recognition method for reading character information from image data obtained by scanning a document as an image.
A reading range setting step for setting a reading range of the document, performed by a reading range setting unit by a user operation from sample image data obtained by scanning a sample of the document as an image;
A reading item determination step of determining a reading item of the document, performed by a reading item determination unit from the coordinate position of the reading range in the sample image data;
An attribute information addition step of adding predetermined attribute information, which is performed by an attribute information addition unit by a user operation on the read item;
And a text data generation step of generating text information displayed in the read item of the image data as text data, performed by the text data generation unit based on the read item and the attribute information from the image data.
Character recognition method. A character recognition program for reading character information from image data obtained by scanning a document as an image,
A reading range setting step of setting a reading range of the document by a user operation from sample image data obtained by scanning the sample of the document as an image;
A reading item determination step of determining a reading item of the document from the coordinate position of the reading range in the sample image data;
An attribute information addition step of adding predetermined attribute information to the read item by a user operation;
Causing a computer to execute a text data generation step of generating, as text data, character information displayed in the read item of the image data based on the read item and the attribute information from the image data.
Character recognition program.

Images