JP6885449B2 - Information identification program, information identification method, and image reader - Google Patents

Description

The present invention relates to an information identification program, an information identification method, and an image reader.

Conventionally, a program for collecting data on paper forms has been known. The program disclosed in Patent Document 1 includes a step of identifying a fixed character written in advance on a form and an entry character entered on the form by a user from image information of a plurality of forms. The distinction between fixed characters and entered characters is the result of comparing character strings extracted from the same position over multiple forms of the same type, and the learning result of accumulating information to evaluate that they are fixed characters and entered characters. And it is done based on the supplementary data etc. input by the worker. In addition, a step is provided in which a non-entry area based on the fixed character and an entry area based on the entry character are identified according to the identification result between the fixed character and the entry character. In addition, a step of generating a form template in which a form is digitized according to the identification result between a non-entry area and an entry area is provided.

Japanese Unexamined Patent Publication No. 2015-5100

However, some forms include, in addition to the area where characters are always entered by the user, an area where characters may or may not be entered by the user. For a form having such an area to be arbitrarily filled in by the user, there is a possibility that the type of the area cannot be accurately identified even by using the above program.

An object of the present invention is to provide an information identification program, an information identification method, and an image reading device capable of appropriately identifying the type of an area provided on a form and executing processing according to the type of the area.

The information identification program according to the first embodiment is an information identification program executed by a computer constituting the information identification device, and the computer reads an entry form which is a form in which characters including handwritten characters are entered. The first acquisition step of acquiring the read data indicating the image, and the form image data indicating the image of the form according to the entry form based on the read data and in which the handwritten characters are not entered are acquired. The second acquisition step, the first region specifying step for specifying the first region surrounded by the contour corresponding to the first pixel value in the read data, and the first region data showing the image of the first region are read. The third acquisition step to be acquired from the data, the form first area specifying step for specifying the form first area corresponding to the first area in the form image data, and the form first data showing the image of the form first area. The fourth acquisition step of acquiring the data from the form image data, the first determination step of determining whether the first area data and the first form data are different, and the first area data and the first form data. When it is determined that the data is different from the above, the character string data indicating the character string including at least the first character string included in the first area is acquired based on the character recognition process, while the first area data and the form first When it is determined that the data does not differ from the data, the first execution step of not acquiring the character string data is executed.

According to the information identification program according to a first embodiment, the voucher image data corresponding to the type of document that has been determined on the basis of the read data is acquired. Further, the first region surrounded by the contour corresponding to the first pixel value is specified from the read data, and the first region data is acquired. When the first area data and the first form data corresponding to the first area of the form image data are different, the character string data including the first character string data corresponding to the first area data is acquired from the read data. To. On the other hand, if the first area data and the first form data are not different, the character string data is not acquired and an instruction to display the first area data is output. For example, the space that is required to always fill in the form, when applied to a first region according to the first embodiment, the handwritten character or the like is filled in the first region, and a first region data and the form first data If they are different, the character string data including the first character string data is acquired from the read data. On the other hand, if handwritten characters or the like are not entered in the first area and the first area data and the first form data do not differ, the character string data is not acquired from the read data. Therefore, the information identification program according to the first embodiment can appropriately identify the type of the area provided in the form and execute the process according to the type of the area.

The information identification method according to the second embodiment is a method that can be executed by a computer constituting the information identification device, and is read data indicating an image obtained by reading an entry form, which is a form in which characters including handwritten characters are entered. The first acquisition step of acquiring the form image data indicating the image of the form according to the entry form based on the read data and in which the handwritten characters are not entered, and the second acquisition step. A first region identification step for specifying a first region surrounded by a contour corresponding to the first pixel value in the read data, and a third region data indicating an image of the first region are acquired from the read data. From the form image data, the acquisition step, the form first area specifying step for specifying the form first area corresponding to the first area in the form image data, and the form first data showing the image of the form first area are obtained from the form image data. It is determined that the fourth acquisition step to be acquired, the first determination step for determining whether the first area data and the first form data are different, and the first area data and the first form data are different. In this case, the character string data indicating the character string including at least the first character string included in the first area is acquired based on the character recognition process, while the first area data and the form first data are not different from each other. If it is determined, it includes a first execution step that does not acquire the character string data.

Following the information identification method according to the second embodiment, the computer of the information identification apparatus to identify the information, it is possible to achieve the same effect as the first embodiment.

The image reading device according to the third embodiment has a reading unit that reads an image of the form, and a reading unit that indicates an image that is read by the reading unit and is a form in which characters including handwritten characters are written. A first acquisition means for acquiring data, and a second acquisition means for acquiring form image data indicating an image of a form in which handwritten characters are not entered, which is a form corresponding to the entry form based on the read data. The first region specifying means for specifying the first region surrounded by the contour corresponding to the first pixel value in the read data, and the first region data indicating the image of the first region are acquired from the read data. (Iii) The form image data includes the acquisition means, the form first area specifying means for specifying the form first area corresponding to the first area in the form image data, and the form first data indicating the image of the form first area. The fourth acquisition means acquired from, the first determination means for determining whether the first area data and the first form data are different, and the first determination means for determining whether the first area data and the first form data are different. If this is done, the character string data indicating the character string including at least the first character string included in the first area is acquired based on the character recognition process, while the first area data and the form first data are not different from each other. When it is determined that the character string data is not acquired, the first execution means is provided.

By using the image reading apparatus according to the third embodiment, it is possible to achieve the same effect as the first embodiment.

It is an overall structure diagram of the information identification system 4. This is a specific example of the blank form 100. This is a specific example of the entry form 101. It is a figure which shows the data structure of the registration form DB 60 for medical insurance among the registration form DB. It is a figure which shows the data structure of the medical insurance setting table 70 among the setting tables. It is a flowchart which shows the information identification process. It is a flowchart which shows the form type determination process. It is a flowchart which shows the user determination process. It is a flowchart which shows the pre-filled area determination process. It is a flowchart which shows the required entry area determination process. It is a flowchart which shows the arbitrary entry area determination process. It is a flowchart which shows the area determination process. This is a specific example of the error display 81 without the corresponding type. This is a specific example of the entry form reading error display 82. This is a specific example of the reading result display 200. This is a specific example of the reading result display 300.

The embodiments illustrated in the present specification will be described with reference to the drawings. First, the system configuration of the information identification system 4 will be described with reference to FIG. The information identification system 4 includes an information identification device 1, an image reading device (hereinafter referred to as “scanner”) 2, and a management server 3. The information identification device 1 can send and receive data and commands to and from the management server 3 via a network 9 including the Internet, LAN, and other dedicated lines.

The configuration of the information identification device 1 will be described. Image data corresponding to the contents described in the form in which characters including handwritten characters are entered is input to the information identification device 1. The information identification device 1 identifies the types of various entry fields provided in the form from the input image data. The information identification device 1 executes processing according to the identification result. The form is, for example, a paper medium, and includes various insurance contract applications, mobile phone contract applications, various applications in administrative procedures, and the like. In this embodiment, the form is an application for various insurance contracts. The information identification device 1 may be a device dedicated to information identification, or may be a general-purpose device such as a so-called personal computer (PC) in which a dedicated application is installed. In this embodiment, the information identification device 1 by a general-purpose PC is illustrated. As shown in FIG. 1, the information identification device 1 includes a CPU 10 which is a controller that controls the information identification device 1. The CPU 10 is electrically connected to a ROM 11, a RAM 12, a hard disk drive (hereinafter referred to as “HDD”) 13, and an input / output interface (hereinafter referred to as “input / output I / F”) 14 via a data bus. .. The ROM 11 stores a program such as a BIOS for operating the information identification device 1, an initial value, and the like. The RAM 12 temporarily stores various data such as the calculation result obtained by the calculation process by the CPU 10.

HDD 13 is a non-volatile storage device. In the present embodiment, the HDD 13 stores a program that causes the CPU 10 to execute the information identification process (see FIG. 6) described later. The HDD 13 is an example of a non-temporary storage medium. The non-temporary storage medium may be any storage medium capable of storing information regardless of the period for storing the information. The non-temporary storage medium may not include a temporary storage medium (eg, a signal to be transmitted). The storage device in the present embodiment is the HDD 13, but the storage device is composed of other non-temporary storage media such as flash memory and RAM that can hold data regardless of the length of time for storing information. May be done.

The input / output I / F 14 mediates the transfer of data. The input / output I / F 14 includes an external communication interface (hereinafter referred to as "external communication I / F") 15, a display unit 16 for displaying an image, an operation unit 17 such as a keyboard and a mouse, and a general-purpose image reading. The scanner 2 which is a device is electrically connected. The external communication I / F 15 connects the information identification device 1 to the network 9. The external communication I / F15 is an interface circuit conforming to a standard such as Ethernet (registered trademark). An external display unit 19 that displays an image outside the information identification device 1 can be connected to the external communication I / F15. The external display unit 19 may be a monitor that can be connected via the network 9, or may be a display of a smartphone or tablet PC that is connected to the information identification device 1 via the network 9 and the external communication I / F15. May be good.

The information recognition process (see Fig. 6) is based on the read data read from various insurance application forms, and is the data of the character string entered in the application form according to the type of entry area provided in the application form. It is a part of the processing of the software that manages. The input / output I / F 14 is an interface element conforming to a predetermined standard (for example, Universal Serial Bus (USB)) for connecting the information identification device 1 and an external device such as a scanner 2. The CPU 10 transmits / receives data to / from an external device such as a scanner 2 via the input / output I / F 14.

The configuration of the scanner 2 will be described. The scanner 2 is a sheet feed scanner or a flatbed scanner. The scanner 2 performs a reading operation for capturing an image such as an entry form 101, which will be described later, forms image data, and transmits the image data to the information identification device 1. The scanner 2 may be a read-only scanner for the entry form 101, or may be a general-purpose scanner.

The scanner 2 includes a CPU 20 which is a controller that controls the scanner 2. The CPU 20 is electrically connected to a ROM 21, a RAM 22, an input / output interface (hereinafter referred to as “input / output I / F”) 24, a reading unit 26, a display unit 27, and an operation unit 28 via a data bus. The ROM 21 stores a program such as an OS for operating the scanner 2, initial values, and the like. The RAM 22 temporarily stores various data such as calculation results obtained by calculation processing by the CPU 20. In particular, the RAM 22 secures at least a storage area for storing the image data of the image of the entry form 101. The input / output I / F 24 mediates the transfer of data. An external communication I / F 25 is electrically connected to the input / output I / F 24. The external communication I / F 25 connects the scanner 2 to the network 9. The scanner 2 may be connected to the information identification device 1 via the network 9 and the external communication I / F 25. An external display unit 29 that displays an image outside the scanner 2 can be connected to the external communication I / F 25. The external display unit 29 is, for example, a display of a smartphone or tablet PC connected to the scanner 2 via the network 9 and the external communication I / F25.

The reading unit 26 includes a well-known contact type image sensor. The reading unit 26 reads the image of the entry form 101. The reading unit 26 converts the read analog image into digital image data. The image data is transferred to the information identification device 1 via the input / output I / F24 or the external communication I / F25 in response to the transfer command of the information identification device 1. The display unit 27 is a display screen for displaying an image provided on the scanner 2. The operation unit 28 is an operation button or the like for operating various operations of the scanner 2. The scanner 2 performs an operation of reading the image of the entry form 101 via the operation unit 28 or in response to the input of the reading instruction of the image of the entry form 101 from the information identification device 1.

The configuration of the management server 3 will be described. The management server 3 may be a general-purpose server as hardware. The management server 3 includes a CPU 30 that controls the management server 3. The CPU 30 is electrically connected to the ROM 31, RAM 32, HDD 33, and input / output I / F 34 via a data bus. The ROM 11 stores a program such as a BIOS for operating the management server 3, an initial value, and the like. The RAM 32 temporarily stores various data such as calculation results obtained by calculation processing by the CPU 30. In the present embodiment, the HDD 33 stores a registration form DB and a form type setting table as a database (hereinafter, referred to as “DB”) for managing various information. The registration form DB is, for example, the medical insurance registration form DB 60 (see FIG. 4), which will be described later. The form type setting table is, for example, the medical insurance setting table 70 (see FIG. 5), which will be described later. An external communication I / F35 is connected to the input / output I / F34. The external communication I / F35 connects the management server 3 to the network 9. The external communication I / F35 is an interface circuit that conforms to a standard such as Ethernet (registered trademark).

The unfilled form 100 and the filled-in form 101 will be described with reference to FIGS. 2 and 3. The blank form 100 is various insurance contract application forms distributed by insurance companies and the like to insurance contract applicants (hereinafter referred to as users). Handwritten characters and the like have not yet been entered in the blank form 100 by the user. The entry form 101 is a form in a state after the user has filled in the necessary items on the unfilled form 100. When the unfilled form 100 and the filled-in form 101 are not particularly distinguished, or when they are collectively referred to, they are simply referred to as "forms".

The form mainly includes a form type display column 110, a date entry column 120, a user ID column 130, a pre-entry column 140, a first optional entry column 150, and a second optional entry column 160. The form may be provided with various entry fields other than these, but the description thereof will be omitted in the present embodiment. The form type display field 110 is a field in which a character string is previously described by printing or the like. The character string described in advance indicates which type of insurance contract the form is for applying for, among various insurance contract application forms. For example, the types of insurance contracts include medical insurance, automobile insurance, fire insurance, earthquake insurance, and the like. The type of form for each type of insurance contract is referred to as the form type below. That is, the form type display column 110 is a column indicating which form type the form is. In the form type display field 110 of the blank form 100 shown in FIG. 3, the characters "medical insurance contract application form" indicating that the application form is a medical insurance contract application form are printed as the form type. The form type display field 110 is provided in the upper left portion of the form. In the present embodiment, the position where the form type display field 110 is provided in the form is the same position regardless of the type of the form.

The date entry field 120 is provided below the form type display field 110 in the form. The date entry field 120 is a field in which the user enters the application date when applying for an insurance contract. As shown in FIG. 2, in the blank form 100, the date is not entered in the date entry field 120 and is blank. In the present embodiment, if the application date is not entered in the date entry column 120, the insurance contract will not be concluded even if the application date is entered in the other columns of the form. That is, the date entry field 120 is a field that the user must fill out in the form. Therefore, in the following, the date entry field 120 is also referred to as a "required entry field". As shown in FIG. 3, the date entry field 120 of the entry form 101 is handwritten by the user as "2015", "August", "2", and "2". By bordering the outer peripheral portion of the date entry field 120 with a thick red straight line, a red rectangle showing the outline of the date entry field 120 is formed on the form. This rectangle is called the first rectangle 121.

The user ID field 130 is provided on the right side of the date entry field 120 in the form. The user ID column 130 is a column in which the user ID is printed by a printer or the like used by a form manager of an insurance company or the like. The user ID is a contractor code or the like indicating a user who intends to make an insurance contract, and in the present embodiment, the user ID is a character string assigned to each user in order to identify the user of the form. In an example of this embodiment, the user ID is "AABBCC1234" shown in FIGS. 3 and 4.

The pre-entry column 140 is provided below the date entry column 120 and the user ID column 130 in the form. The pre-filled column 140 is a column for the administrator to fill in the information that the administrator knows among the user-specific information before the form is distributed to the user. In the unfilled form 100 at the time of distribution to the user, the user's own personal information is already entered in the pre-filled field 140. In the example of this embodiment, the user "Taro Yamada" is associated with the user ID of "AABBCC1234" and is registered in the medical insurance registration form DB in advance. The administrator creates an unfilled form 100 in which the user information registered in the medical insurance registration form DB 60 is pre-printed in the pre-filled field 140 by a printer or the like used by the administrator. The administrator distributes the created blank form 100 to the users. Specifically, as shown in FIG. 2, among the pre-filled fields 140, "Yamada Taro", "Yamada Taro", which indicate the user's name which is the user's information registered in the medical insurance registration form DB60, An unfilled form 100 on which the user's gender, date of birth, address and telephone number are pre-printed is distributed to the user. By edging the outer peripheral portion of the pre-filled field 140 with a thick green straight line, a green rectangle showing the outline of the pre-filled field 140 is formed on the form. This rectangle is called the second rectangle 141.

In the pre-filled field 140, there is an area for stamping the application stamp required for contract application, a check box area for entering a check when the address and telephone number have changed, and the changed address. And there is an area for entering a telephone number. The user stamps these areas of the distributed blank form 100 and fills in necessary items. In the entry form 101 shown in FIG. 3, an application stamp is stamped in the pre-filled field 140, and the check box area is checked on the assumption that the address pre-printed in the pre-filled field 140 has been changed, and the changed address. Indicates a state in which is handwritten by the user.

The first optional entry field 150 is provided below the pre-entry field 140 in the form. The first optional entry field 150 is provided for entering information such as a name of the user's spouse. For example, when the user has a spouse and the spouse also makes an insurance contract using the same form, the user manually fills in the spouse information required for the contract in the first optional entry field 150. On the other hand, if the user does not have a spouse, or if the user has a spouse but does not apply for an insurance contract for the spouse, the user does not need to fill in the first optional entry field 150. As described above, in the first optional entry field 150, handwritten characters may or may not be entered according to the needs of the user. Therefore, as shown in FIG. 2, in the blank form 100, the first optional entry field 150 is left blank. In addition, FIG. 3 shows an entry form 101 in which the name, gender, and date of birth of the user's spouse are handwritten and the application seal is affixed. By edging the outer peripheral portion of the first optional entry field 150 with a thick blue straight line, a blue rectangle showing the outline of the first optional entry field 150 is formed on the form. This rectangle is called the third rectangle 151.

The second optional entry field 160 is provided below the first optional entry field 150 in the form. The second optional entry field 160 is provided for entering information such as a name of the user's child. For example, when the user has a child and the same form is used for the insurance contract, the user manually fills in the child information required for the contract in the second optional entry field 160. On the other hand, if the user does not have a child, or if the user has a child but does not apply for an insurance contract for the child, the user does not need to fill in the second optional entry field 160. As described above, the second optional entry field 160 may or may not be filled with handwritten characters as required by the user, as in the first optional entry field 150. Therefore, as shown in FIG. 2, in the blank form 100, the second optional entry field 160 is left blank. FIG. 3 shows an entry form 101 in which information about the child is not entered and the second optional entry field 160 is blank because the insurance contract is not made for the child of the user. The outer peripheral portion of the second optional entry field 160 is bordered by a thick light blue (pale blue) straight line, so that a light blue rectangle showing the outline of the second optional entry field 160 is formed on the form. This rectangle is called the fourth rectangle 161. In the following, when the first optional entry column 150 and the second optional entry column 160 are collectively referred to, they are also simply referred to as "optional entry column".

In this way, various forms may or may not be filled in by the user, such as the required entry field of the present embodiment and the optional entry field. There may be a mixture of entry fields. Further, the form may include an entry field in which the form to be distributed to the user has a pre-filled item, such as the pre-filled field 140 of the present embodiment. In a form that includes a plurality of types of entry fields with different entry requirements, the necessity of entry mode differs in the form by surrounding the outline of the entry column with a rectangle using a different color for each entry column type. It is easy to visually convey to the user that multiple entry fields are included.

The registration form DB stored in the HDD 33 will be described. In the registration form DB, information such as name, gender, address, telephone number, etc., which is the personal information of a plurality of contract applicants, is associated with the user ID given to each of the contract applicants, and the form type. It is registered every time.

FIG. 4 shows a DB that manages information about a contract applicant for a medical insurance form type and information on a form that is a contract application form among various types of insurance contracts registered in the registration form DB. is there. Note that FIG. 4 illustrates the data of the smallest unit registered in association with the user ID "AABBCC1234" in the entire medical insurance registration form DB60.

The medical insurance registration form DB60 manages information by classifying registration items into each of the "form management information" column, the "personal information" column, the "spouse information" column, and the "child information" column. The form management information column includes registration items of "user ID", "registration date", "application date", and "registration form image data". In the "registration date", the date on which the user ID and the information associated with the user ID are registered in the medical insurance registration form DB 60 by the administrator is registered. In the "application date", when the character string indicating the date entered in the date entry field 120 of the entry form 101 is acquired by executing the information identification process described later, the date based on the acquired character string is ". It is registered as "application date". In the "registered form image data", image data showing an image of the entire unfilled form 100 corresponding to the user ID is registered. In the present embodiment, the image data is in JPEG format, but the image data format may be in another format such as GIF format.

In each registration item included in the personal information column, information indicating the contract applicant's name, gender, address, and telephone number is pre-registered in association with the user ID. The registered content in the personal information column is the content pre-printed in the pre-entry column 140 of the unfilled form 100.

Each registration item included in the spouse information column is acquired when a character string or the like indicating the content entered in the first optional entry column 150 of the entry form 101 is acquired by executing the information identification process described later. Each content based on the character string etc. is registered. In the child information column, when a character string or the like indicating the content entered in the second optional entry field 160 of the entry form 101 is acquired by executing the information identification process described later, each based on the acquired character string or the like. The content is registered. Therefore, the spouse information column and the child information column do not have registration contents on the registration date of the medical insurance registration form DB60 and at the time when the blank form 100 is created based on the medical insurance registration form DB60. So it is left blank. In FIG. 4, "-" indicates that the registration content corresponding to the registration item does not exist.

The medical insurance setting table 70 stored in the HDD 33 will be described with reference to FIG. The medical insurance setting table 70 stores setting information for the medical insurance contract application form among the forms corresponding to a plurality of types of insurance contract application forms handled by the administrator. Although not shown, the HDD 33 stores a plurality of setting tables for storing form setting information corresponding to various form types other than the medical insurance setting table 70. The setting information is information indicating the arrangement of the form type display column 110, the date entry column 120, the pre-entry column 140, the first optional entry column 150, and the second optional entry column 160 in each form.

The medical insurance setting table 70 has each of "type area" column, "ID area" column, "pre-entry area" column, "required entry area" column, "first optional area" column, and "second optional area". The setting information is classified and stored in the column. The type area column stores each information of "coordinate position", "area range", and "type area data" for the form type display field 110 in the form. The coordinate position (x0, y0) is the position where the upper left part of the form type display field 110 is arranged in the form in the read data obtained by reading the form, and the position of the upper left corner of the form is the origin of the xy coordinates. The coordinates are displayed as (0,0). The area range is information used for extracting the image data corresponding to the form type display field 110 from the read data in the form type determination process (see FIG. 7) described later. The area range represents the width and height of the extraction area with (x0, y0) as the origin in terms of the number of dots (number of pixels). The area range "25 x 80" based on the coordinate position (x0, y0) is set to be slightly wider than the position and area occupied by the form type display field 110 in the form. The registration type area data stores image data indicating the form type display field 110 in the form in the state before the content corresponding to the pre-filled field 140 is printed on the blank form 100 as a template for pattern matching. The data format of the registration type area data may be any format capable of pattern matching, including the JPEG format. In FIG. 5, one image data with a file name of "iryohoken.jpg" is shown as the registration type area data. “Iryohoken.jpg” is image data associated with a form type indicating a medical insurance contract application form. The registration type area data also includes image data indicating the form type display field 110 corresponding to various form types. In addition, each of the setting tables for other form types is provided with the same columns as the above-mentioned "type area" column.

The "ID area" column stores each information of "coordinate position" and "area range" for the user ID column 130 in the form. The coordinate positions (x1, y1) indicate the positions where the upper leftmost portion of the user ID field 130 is arranged in the form in the read data with respect to the above-mentioned origin (0,0). The area range is information used for determining the user ID from the read data in the user ID determination process (see FIG. 8) described later. The area range represents the height and width of the area with (x1, y1) as the origin in terms of the number of dots.

The pre-filled area column stores information of "coordinate position", "area range", "first pixel value", and "first threshold value" for the pre-filled field 140 in the form. The coordinate position (x2, y2) indicates the position where the upper leftmost portion of the pre-entry field 140 is arranged in the form in the read data with respect to the above-mentioned origin (0,0). The area range is information used to identify the second rectangle 141 corresponding to the pre-filled field 140 from the read data in the area determination process (see FIG. 12) described later. The area range represents the height and width of the area with (x2, y2) as the origin in terms of the number of dots. The area range "100 x 200" based on the coordinate position (x2, y2) is set to be slightly wider than the position and area occupied by the second rectangle 141 in the form. The first pixel value is a value indicating the green color of the second rectangle 141 provided on the form, which is shown in RGB format with 256 gradations for each color. The first threshold value defines a range to be regarded as the first pixel value when the second rectangle 141 is specified in the area determination process. In the present embodiment, the first threshold value "± 5" means that the color of the pixel value included in the range of five steps before and after the pixel value defined in the first pixel value is the same as the color of the first pixel value. It defines what is considered a color.

Similar to the pre-entry area column, the required entry area column stores each information of "coordinate position", "area range", "second pixel value" and "second threshold value" for the date entry field 120 in the form. ing. The area range "30 x 130" based on the coordinate positions (x3, y3) is set to be slightly wider than the position and area occupied by the first rectangle 121 in the form. The second pixel value is a value indicating the red color of the first rectangle 121 provided on the form, which is shown in RGB format with 256 gradations for each color.

Similarly, the first optional area column also stores information on the "coordinate position", "area range", "third pixel value", and "third threshold value" for the first optional entry field 150 in the form. Similarly, the second optional area column also stores information on the "coordinate position", "area range", "fourth pixel value", and "fourth threshold value" for the second optional entry field 160 in the form. The area range "50 x 200" based on the coordinate position (x4, y4) is set to be slightly wider than the position and area occupied by the third rectangle 151 in the form. The area range "150 x 200" based on the coordinate position (x5, y5) is set to be slightly wider than the position and area occupied by the fourth rectangle 161 in the form. The third pixel value is a pixel value indicating a color different from the first pixel value and the second pixel value, and the value indicating the blue color of the third rectangle 151 provided on the form is RGB of 256 gradations for each color. It is shown in the format. The fourth pixel value is a pixel value indicating a color different from the first pixel value, the second pixel value, and the third pixel value, and the value indicating the light blue of the fourth rectangle 161 provided on the form is the 256th floor of each color. It is shown in the RGB format of the key. The second threshold value, the third threshold value, and the fourth threshold value also define that the colors of the pixel values included in the range of five levels before and after the corresponding pixel values are regarded as the same color as in the first threshold value. For the first pixel value, the second pixel value, the third pixel value, and the fourth pixel value, a color space other than the RGB format (for example, YCbCr format using brightness and color difference, hue, saturation, and lightness was used. It may be defined in HSV format, etc.).

The information identification process in the present embodiment will be described with reference to FIGS. 6 to 16. The information identification process is a process of identifying each area provided on the form based on the read data read by the scanner 2 for the image of the entry form 101, and acquiring the read result data according to the identification result.

Prior to the start of the information identification process, the administrator prepares a registration form DB including a medical insurance registration form DB 60 for each type of insurance contract. In addition, the administrator registers the user ID and personal information in association with the registration form DB corresponding to the type of insurance contract that the user desires to make a contract. Then, the administrator creates an unfilled form 100 in which the user ID is printed in the user ID field 130 and the personal information is printed in the pre-filled field 140 based on the registered contents registered in the registration form DB. The blank form 100 is not limited to being created by being printed on a paper medium or the like by a printer or the like. For example, the unfilled form 100 may be generated as an image file by outputting the registered contents registered in the registration form DB to a file. The administrator obtains image data showing the image of the created blank form 100 by using a scanner 2 or the like, associates the obtained image data with the form type and the user ID, and registers the form image in the registration form DB. Register as data. In the following, it is assumed that the administrator distributes the blank form 100 to users who desire a medical insurance contract. It is assumed that the user submits an entry form 101 in which necessary items are filled in the distributed blank form 100 to the administrator. In the information identification process, it is assumed that the process related to this entry form 101 is performed.

As shown in FIG. 6, when the information identification process is started, the CPU 10 transmits an instruction to read the image of the entry form 101 to the scanner 2 via the input / output I / F 14 (S11). Upon receiving the instruction, the scanner 2 reads the entry form 101 by the reading unit 26, generates read data which is the data of the read image, and transmits the read data to the information identification device 1. The CPU 10 acquires the read data transmitted from the scanner 2 via the input / output I / F 14 (S12). When the blank form 100 is generated as an image file, the CPU 10 does not have to perform the process of S11. In this case, the CPU 10 may acquire an image file based on the blank form 100 in the process of S12. After that, the CPU 10 executes the form type determination process (S13), the user determination process (S14), the pre-entry area determination process (S15), the required entry area determination process (S16), and the optional entry area determination process (S18) in this order.

The form type determination process (see S13 and FIG. 6) will be described with reference to FIG. 7. The form type determination process is a process for determining the form type of the entry form 101 based on the read data. When the form type determination process is started, the CPU 10 determines whether or not the read data is color image data (S31). For this determination, a well-known determination method using the magnitude and distribution of saturation indicated by the scanned image data may be used. In the present embodiment, the regular form distributed to the user by the administrator includes a date entry field 120 surrounded by a red rectangle, a pre-fill field 140 surrounded by a green rectangle, and a first entry field surrounded by a blue rectangle. It includes an optional entry field 150 and a second optional entry field 160 surrounded by a light blue rectangle. When the scanned image data is not color image data, the user may use a form other than the form officially distributed by the administrator, and therefore, the process of S31 is provided to eliminate this.

When the read data is color image data (S31: YES), the CPU 10 extracts the form type area data from the read data (S32). The form type area data is image data of a portion of the read data corresponding to the form type display field 110 of the entry form 101. In the present embodiment, in the form handled by the administrator, the position and area range where the form type display field 110 is provided in the form are the same throughout the form type. The CPU 10 refers to any setting table including the medical insurance setting table 70 stored in the HDD 33. The CPU 10 extracts from the read data the image data of the portion corresponding to the coordinate position and the coordinate range defined in the type area column of the referenced setting table as the form type area data.

The CPU 10 compares the extracted form type area data with the registration type area data stored in the setting table type area column described above (S33). In this process, the CPU 10 acquires the registration type area data by referring to the setting table stored in the HDD 33, and performs template matching of the extracted form type area data with the acquired registration type area data.

The CPU 10 determines whether or not the registration type area data matches the form type area data (S34). When there is a registration type area data that matches the form type area data (S34: YES), the CPU 10 determines the form type associated with the matching registration type area data as the form type of the entry form 101. (S35). The CPU 10 stores the determination result of the form type in the RAM 12. In the example of the present embodiment, it is assumed that the fact that the form type is determined to be "for medical insurance" is stored in the RAM 12. The CPU 10 returns the process to the information identification process (see FIG. 6).

On the other hand, when the read data is not color image data (S31: NO), or when there is no registered type area data that matches the form type area data (S34: NO), the CPU 10 displays an error display 81 without the corresponding type (FIG. An instruction to output (see 13) is transmitted to the display unit 16 via the input / output I / F 14 (S36). The error display 81 without the corresponding type indicates that the entry form 101 is not a regular form and that the form type shown in the form type display field 110 provided in the entry form 101 does not correspond to the information identification process. It is an image displayed to show in. The image data indicating the error display 81 without the corresponding type is stored in the HDD 33. The display unit 16 displays the corresponding typeless error display 81 in response to the instruction by the process of S36. The CPU 10 shifts the process to the determination of S29 (see FIG. 6).

The user determination process (see S14 and FIG. 6) will be described with reference to FIG. The user determination process is a process for determining the user of the entry form 101 based on the read data. When the user determination process is started, the CPU 10 extracts the user ID area data from the read data (S41). The user ID area data is image data of a portion of the read data corresponding to the user ID field 130 of the entry form 101. In this process, the CPU 10 refers to the determination result of the form type stored in the RAM 12 and acquires the information indicating that the form type of the read data is "for medical insurance". The CPU 10 refers to the medical insurance setting table corresponding to the acquired form type among the setting tables stored in the HDD 33. From the read data, the CPU 10 extracts the image data of the portion corresponding to the coordinate position and the coordinate range defined in the ID area column of the medical insurance setting table 70 as the user ID area data. The CPU 10 stores the extracted user ID area data in the RAM 12.

The CPU 10 executes character recognition processing on the extracted user ID area data (S42). In the present embodiment, as the character recognition process, a known OCR process is adopted in which a feature corresponding to the outline of a character is extracted from an image and compared with a standard pattern of the character. The CPU 10 extracts characters from the user ID area data stored in the RAM 12 by executing the OCR process. In the OCR process, a well-known library including a feature extraction step such as a well-known edge detection technique and a matching step for comparing the extracted feature with a character template can be used. The CPU 10 analyzes the extracted characters and identifies the characters printed in the user ID field 130. The CPU 10 generates ID character string data, which is character string data indicating a character string printed in the specified user ID field 130, and stores it in the RAM 12.

The CPU 10 refers to the registration form DB of the form type determined in the process of S35 (see FIG. 7) among the registration form DB stored in the HDD 33 (S43). In the example of this embodiment, since the form type of "for medical insurance" is stored in the RAM 12, it is assumed that the CPU 10 refers to the medical insurance registration form DB 60 accordingly. The CPU 10 compares the registered user ID, which is a plurality of user IDs registered in the referenced registration form DB, with the ID character string data stored in the RAM 12 (S45).

The CPU 10 determines whether the registered user ID has a character string that matches the character string indicated by the ID character string data (S46). If the registered user ID does not have a character string that matches the character string indicated by the ID character string data (S46: NO), the CPU 10 indicates that there is no registered user ID corresponding to the character string indicated by the ID character string data. None An instruction to output an error display (not shown) is transmitted to the display unit 16 via the input / output I / F 14 (S47). The error display without the corresponding ID is an image displaying a message such as "the corresponding registered user ID does not exist". The display unit 16 displays an error display without the corresponding ID in response to the instruction in S47.

The CPU 10 transmits an instruction to directly output an input display (not shown) to the display unit 16 via the input / output I / F 14 (S48). The direct input display is a screen for urging the operator to input the user ID described in the user ID field 130 of the entry form 101 by using the operation unit 17, for example. This process is performed to acquire the ID character string data when the ID character string data is not correctly generated from the read data, such as when the print density of the user ID printed on the entry form 101 is low. The display unit 16 directly displays the input display in response to the instruction in S48.

The CPU 10 acquires the ID character string data input via the operation unit 17 (S49). After that, the CPU 10 repeats the processes of S43 to S47 until it is determined that the registered user ID has a character string that matches the character string indicated by the ID character string data (S46: YES). When the processes of S43 to S47 are repeated a predetermined number of times or more, the CPU 10 may shift the processes to the determination of S29 (see FIG. 6).

When the registered user ID has a character string that matches the character string indicated by the ID character string data (S46: YES), the CPU 10 corresponds to the read data with the registered user ID that matches the character string indicated by the ID character string data. It is acquired as a user ID to be used (S50). The CPU 10 stores the acquired data indicating the user ID in the RAM 12. The CPU 10 returns the process to the information identification process (see FIG. 6). In the example of this embodiment, it is assumed that the user ID indicating the character string "AABBCC1234" printed in the user ID field 130 of the entry form 101 is stored in the RAM 12.

The pre-filled determination process (see S15 and FIG. 6) will be described with reference to FIG. In the pre-filling determination process, it is determined whether the read data includes the first area corresponding to the pre-filling column 140. When the read data includes the first region, the first region data showing the image of the first region is acquired, and processing is performed according to the content of the acquired first region data. As shown in FIG. 9, when the pre-entry area determination process is started, the CPU 10 refers to the form type setting table determined in the process of S35 (see FIG. 7) among the setting tables stored in the HDD 33. (S51). In the example of the present embodiment, since it is stored in the RAM 12 that the form type is "for medical insurance" by the processing of S35, it is assumed that the CPU 10 refers to the medical insurance setting table 70 accordingly. The CPU 10 acquires the first pixel value and the first threshold value from the pre-filled area column of the setting table (S52). The first pixel value and the first threshold value acquired here are also referred to as a set pixel value and a set threshold value in the pre-filled area determination process.

The CPU 10 executes the area determination process (S53). The area determination process (see S53 and FIG. 9) will be described with reference to FIG. The area determination process is a process of determining whether the read data includes an area surrounded by a color corresponding to the set pixel value, based on the set pixel value and the set threshold value in the pre-filled area determination process. The area determination process is also executed in the required entry area determination process (see FIG. 10) and the optional entry area determination process (see FIG. 11), which will be described later. In this case, is the area surrounded by the color corresponding to the set pixel value included based on the set pixel value and the set threshold value in the required entry area determination process or the set pixel value and the set threshold value in the optional entry area determination process? Is determined.

As shown in FIG. 12, when the area determination process is started, the CPU 10 sets the search area range (S111). As described above, the search area range is the range of the area for searching the coordinate positions of the pixels having the pixel values corresponding to the set pixel values in the read data. Specifically, the CPU 10 refers to the form type setting table determined in the process of S35 (see FIG. 7) among the setting tables stored in the HDD 33, and refers to the coordinate position and the area range corresponding to the set pixel values. To get. In the example of the present embodiment, the medical insurance setting table 70 is referred to, and the coordinate positions (x2, y2) associated with the first pixel value, which is the set pixel value, in the pre-filled area column of the medical insurance setting table 70. ), And the area range "100 x 200" with the coordinate position as the origin is acquired. The CPU 10 sets the area range based on the acquired coordinate position in the search area range.

The CPU 10 determines whether or not each of the pixels indicating the coordinate position in the search area range of the read data has a pixel value corresponding to the set pixel value. The pixel value corresponding to the set pixel value is a pixel value within the range of the set threshold value with respect to the set pixel value. Here, the pixel value within each range of 0 ≦ R ≦ 5, 250 ≦ G ≦ 255, 0 ≦ B ≦ 5 corresponds to the set pixel value. In the present embodiment, it is repeatedly determined whether or not each of the pixels indicating the coordinate position in the search area range of the read data has a pixel value within the range of the first threshold value with respect to the first pixel value. Then, the CPU 10 acquires the coordinate position of each pixel for the pixel determined to have the pixel value corresponding to the set pixel value (S112). The CPU 10 stores the acquired coordinate position in the RAM 12.

The CPU 10 calculates an approximate straight line by approximating a plurality of coordinate positions acquired in the process of S112 using a well-known straight line approximation method such as the least squares method (S113). In the example of the present embodiment, in the search area range of the read data, an approximate straight line corresponding to each side of the green second rectangle 141 that borders the pre-entry field 140 provided in the entry form 101 is calculated. In this process, for example, the CPU 10 allocates the pixels acquired in S112 to the clusters by using the K-means method. Since it is assumed that the approximate straight line corresponding to the four sides of the rectangle is calculated, the number of clusters is set to "4". By approximating each cluster using the linear approximation method, an approximate straight line corresponding to each side of the second rectangle 141 is calculated. The number of clusters to be set may be appropriately adjusted according to the number of pre-filled areas.

The CPU 10 acquires the registered form image data according to the form type and the user ID from the registered form DB stored in the HDD 33 (S115). Specifically, the CPU 10 refers to the registration form DB of the form type determined in the process of S35 (see FIG. 7) among the registration form DB stored in the HDD 33, and from the registration form DB of the referenced type, The registration form image data corresponding to the user ID acquired in the process of S50 (see FIG. 8) is acquired. In the example of this embodiment, the form type of "for medical insurance" is stored in the RAM 12 by the processing of S35, and the user ID indicating the character string of "AABBCC1234" is stored in the RAM 12 by the processing of S50. .. Therefore, it is assumed that the medical insurance registration form DB 60 shown in FIG. 4 is referred to in the registration form DB, and the registration form image data “AABBCC1234_0.jpg” associated with the user ID “AABBCC1234” is acquired.

The CPU 10 has a shape formed by an approximate straight line calculated in the process of S113 and a rectangular shape composed of pixels having a pixel value corresponding to a set pixel value in the registered form image data acquired in the process of S115. Compare (S116). When the area determination process is executed in the pre-filled area determination process (see FIG. 9), the set pixel value is the first pixel value defined in the pre-filled area of the setting table. In the present embodiment, the first pixel value defined in the pre-entry area of the medical insurance setting table 70 is a pixel value indicating green corresponding to the color indicated by the second rectangle 141 provided on the entry form 101. is there. The CPU 10 specifies the shape of the figure indicated by the set of pixels of the first pixel value, which is the set pixel value, included in the registered form image data. In specifying the shape of the figure indicated by the set of pixels of the first pixel value, the CPU 10 sets a predetermined search area in the registered form image data as in the process of S111. Similar to the process of S112, the CPU 10 acquires the coordinate position of each pixel for the pixel determined to have the pixel value corresponding to the set pixel value in the search area. Then, the CPU 10 calculates an approximate straight line based on the acquired plurality of coordinate positions, as in the process of S113. In this case, the shape of the second rectangle 141 included in the registered form image data is specified. Then, the CPU 10 compares the shape of the specified second rectangle 141 with the shape formed by the approximate straight line based on the read data.

The CPU 10 determines whether or not the shape of the specified second rectangle 141 and the shape formed by the approximate straight line based on the read data are within a predetermined same range (S117). When the shape of the specified second rectangle 141 and the shape formed by the approximate straight line based on the read data are not within the same predetermined range (S117: NO), the CPU 10 performs the processing in the pre-entry area determination processing (see FIG. 9). Return to. On the other hand, when the shape of the specified second rectangle 141 and the shape formed by the approximate straight line based on the read data are within a predetermined same range (S117: YES), the CPU 10 determines the shape formed by the approximate straight line based on the read data. The area surrounded by is specified as the area corresponding to the set pixel value (S118). In this case, the area specified with the set pixel value as the first pixel value is hereinafter referred to as the first area. The CPU 10 stores information such as coordinate values indicating a position and a size in the read data of the first region in the RAM 12. The CPU 10 returns the process to the pre-filled area determination process (see FIG. 9). If the area determination process is executed in the required entry area determination process or the optional entry area determination process, the CPU 10 returns the process to the required entry area determination process or the optional entry area determination process.

Returning to the description of FIG. The CPU 10 determines whether or not the first region corresponding to the first pixel value is specified in the area determination process (S54). In this process, the CPU 10 determines that the first area is specified in the area determination process when the RAM 12 stores information such as coordinate values indicating the position and size in the read data of the first area. .. When the first area is not specified in the area determination process (S54: NO), the CPU 10 returns the process to the information identification process (see FIG. 6). It should be noted that this process takes into consideration that the pre-entry field 140 may not be provided in the form and the first area may not be specified.

When the first area is specified in the area determination process (S54: YES), the CPU 10 stores the first area data indicating the image of the first area as information such as the coordinate values of the first area stored in the RAM 12. Is acquired from the read data based on (S55). The first area data corresponds to the image of the pre-entry field 140 in the entry form 101.

The CPU 10 specifies the position and size of the area surrounded by the figure indicated by the set of pixels of the first pixel value, which is the set pixel value, in the registered form image data (S56). In this case, the CPU 10 stores information such as coordinate values indicating the position and size in the registered form image data of the area surrounded by the second rectangle 141 corresponding to the first pixel value in the RAM 12. The CPU 10 acquires the registered first area data indicating the image of the area surrounded by the second rectangle 141 from the registered form image data based on the information such as the coordinate values stored in the RAM 12 (S58). The registration first area data corresponds to the image of the pre-filled field 140 in the blank form 100.

The CPU 10 compares the first area data with the registered first area data (S59). In this comparison, the CPU 10 analyzes, for example, color elements (gradation, hue) for each pixel constituting the first region data. Then, the CPU 10 acquires a histogram whose horizontal axis is the color element of the first region data, which is the result of analysis for each pixel. Further, the CPU 10 also acquires a histogram with the color element of the registered first area data as the horizontal axis for the registered first area data. The CPU 10 compares each of the acquired histograms. The histogram of the first area data corresponds to the contents printed and printed in the pre-filled field 140 of the blank form 100. In the histogram of the registration first area data, in addition to the contents printed and printed in the pre-entry field 140 of the entry form 101, the contents of the handwritten characters etc. entered when the handwritten characters etc. are entered by the user. Corresponds to the content in which is added.

The CPU 10 determines whether there is a difference between the first area data and the registered first area data (S61). This judgment is made based on, for example, whether or not there is a difference between the histograms as a result of comparing the histogram of the first region data with the histogram of the registered first region data. That is, by the process of S61, it is determined whether or not the handwritten characters or the like have been entered by the user in the pre-entry field 140 of the entry form 101 with respect to the pre-entry field 140 of the unfilled form 100. In the example of the present embodiment, the pre-filled field 140 of the unfilled form 100 shown in FIG. 2 corresponds to the content registered in advance by the administrator in association with the user ID "AABBCC1234" in the medical insurance registration form DB60. The printed content is shown. In the pre-filled field 140 of the entry form 101 shown in FIG. 3, in addition to the content pre-printed in the pre-filled field 140, the address change content is handwritten and the user's application stamp is stamped. In this case, a difference occurs between the histogram of the first region data and the histogram of the registered first region data.

When there is a difference between the first area data and the registered first area data (S61: YES), the CPU 10 sets the first area data as the execution target of the character recognition process in the process of S21 (see FIG. 6) described later. It is decided (S62). The CPU 10 stores in the RAM 12 that the first area data is to be executed for the character recognition process. The CPU 10 returns the process to the information identification process (see FIG. 6).

On the other hand, when there is no difference between the first region data and the registered first region data (S61: NO), the CPU 10 determines the first region as the brightness change display target (S63). The CPU 10 acquires the average value of the brightness values indicated by the entire first region data from the first region data. The CPU 10 stores the average value of the acquired luminance values in the RAM 12 in association with the fact that the first region is determined as the luminance change display target. The CPU 10 returns the process to the information identification process (see FIG. 6).

The required entry determination process (see S16 and FIG. 6) will be described with reference to FIG. In the mandatory entry determination process, it is determined whether the read data includes the second area corresponding to the date entry field 120. When the read data includes a second region, the second region data indicating the image of the second region is acquired, and processing is performed according to the content of the acquired second region data. As shown in FIG. 10, when the required entry area determination process is started, the CPU 10 refers to the form type setting table determined in the process of S35 (see FIG. 7) among the setting tables stored in the HDD 33. (S71). The CPU 10 acquires the second pixel value and the second threshold value from the required entry area column of the setting table (S72). In the example of this embodiment, it is assumed that the second pixel value and the second threshold value indicating red are acquired according to the reference of the medical insurance setting table 70. The second pixel value and the second threshold value acquired here are also referred to as a set pixel value and a set threshold value in the required entry area determination process.

The CPU 10 executes the above-mentioned area determination process (see FIG. 12) based on the set pixel value and the set threshold value in the required entry area determination process (S73). In the area determination process, a search area range for searching the coordinate position of the pixel having the pixel value corresponding to the second pixel value in the read data is set based on the setting table (S111). Then, the coordinate positions of the pixels having the pixel values corresponding to the second pixel values in the search area range are acquired (S112). An approximate straight line is calculated from the acquired coordinate positions (S113). From the registration form DB, the registration form image data corresponding to the form type and the user ID is acquired (S115). The shape formed by the approximate straight line is compared with the rectangular shape formed by the pixels having the pixel values corresponding to the second pixel values in the registered form image data (S116). The rectangle formed by the pixels having the pixel value corresponding to the second pixel value in the registered form image data corresponds to the shape of the first rectangle 121 provided on the form. It is determined whether or not the shape of the second rectangle 141 in the registered form image data and the shape formed by the approximate straight line based on the read data are within a predetermined same range (S117). When the shape of the second rectangle 141 in the registered form image data and the shape formed by the approximate straight line based on the read data are within the same predetermined range (S117: YES), the shape is surrounded by the shape formed by the approximate straight line. The region is specified as the second region corresponding to the second pixel value (S118).

The CPU 10 determines whether or not the second area is specified in the area determination process (S74). The CPU 10 determines that the second area is specified when the RAM 12 stores information such as coordinate values indicating the position and size in the read data of the second area.

When the area corresponding to the second pixel value is not specified in the area determination process (S74: NO), the CPU 10 gives an input / output I / F 14 an instruction to output an entry form reading error display 82 (see FIG. 14). It is transmitted to the display unit 16 via (S84). Since the date entry field 120 is a required entry field in the form, it is assumed that the date entry field 120 always exists in the form. However, there are cases where the second area corresponding to the date entry field 120 is not specified in the area determination process because the read data is not properly acquired. The entry form reading error display 82 is an image displayed to encourage the operator to read the entry form 101 again by the reading unit 26 and acquire new reading data. In the example of the present embodiment, the entry form reading error display 82 prompts the operator to read the entry form 101 by displaying the message "Please scan the form again". The image data showing the entry form reading error display 82 is stored in the HDD 33. The display unit 16 displays the corresponding typeless error display 81 in response to the instruction by the process of S74.

The CPU 10 determines whether or not an instruction for causing the reading unit 26 to read the entry form 101 has been input via the operation unit 17 (S85). When an instruction for causing the reading unit 26 to read the entry form 101 is input (S85: YES), the CPU 10 returns the process to S11 (see FIG. 6). When the instruction for reading the entry form 101 is not input by the reading unit 26 (S85: NO), the CPU 10 shifts the process to the determination in S29 (see FIG. 6).

When the area corresponding to the second pixel value is specified in the area determination process (S74: YES), the CPU 10 stores the second area data indicating the image of the second area in the RAM 12 of the second area. Obtained from the read data based on information such as coordinate values (S75). The second area data corresponds to the image of the date entry field 120 in the entry form 101.

The CPU 10 specifies the position and size of a region surrounded by a figure indicated by a set of pixels having a second pixel value, which is a set pixel value, in the registered form image data (S76). In this case, the CPU 10 stores in the RAM 12 information such as coordinate values indicating the position and size of the registered form image data in the area surrounded by the first rectangle 121 in the registered form image data. The CPU 10 acquires the registered second area data indicating the image of the area surrounded by the first rectangle 121 from the registered form image data based on the information such as the coordinate values stored in the RAM 12 (S78). The registration first area data corresponds to the image of the date entry field 120 in the blank form 100.

The CPU 10 compares the second area data with the registered second area data (S79). In this process, the CPU 10 acquires, for example, histograms relating to the color elements of the second region data and the registered second region data, respectively. The CPU 10 compares each histogram.

The CPU 10 determines whether there is a difference between the first area data and the registered first area data (S81). This determination is made, for example, by whether there is a difference between the histograms. That is, by the process of S81, it is determined whether or not the user has entered the handwritten characters or the like in the date entry field 120 of the entry form 101 with respect to the date entry field 120 of the unfilled form 100. In the example of this embodiment, the date entry field 120 of the blank form 100 shown in FIG. 2 is blank, but the date entry field 120 of the entry form 101 shown in FIG. 3 is filled with "2015", "8", and "2". The date based on the numbers is written by hand. In this case, a difference occurs between the histogram of the second region data and the histogram of the registered second region data.

When there is a difference between the second area data and the registered second area data (S81: YES), the CPU 10 sets the second area data as the execution target of the character recognition process in the process of S21 (see FIG. 6) described later. It is decided (S82). The CPU 10 stores in the RAM 12 that the second area data is to be executed for the character recognition process. The CPU 10 returns the process to the information identification process (see FIG. 6).

On the other hand, when there is no difference between the second area data and the registered second area data (S81: NO), the CPU 10 stores in the RAM 12 an instruction to output an error display 83 (see FIG. 16) in which the required entry field is not entered. (S83). In the required entry field blank error display 83, there is no difference between the second area data and the registered second area data in the reading result display described later, that is, the necessary entry is not made in the date entry field 120. Is an image displayed to show the operator in association with the second area. The CPU 10 returns the process to the information identification process (see FIG. 6).

The optional entry determination process (see S18 and FIG. 6) will be described with reference to FIG. In the optional entry determination process, it is determined whether the read data includes an area corresponding to the optional entry column. If the read data includes an area corresponding to the optional entry field, the corresponding area data showing the image of the area corresponding to the optional entry field is acquired, and processing is performed according to the content of the acquired corresponding area data. Will be. As shown in FIG. 11, when the arbitrary entry area determination process is started, the CPU 10 refers to the form type setting table determined in the process of S35 (see FIG. 7) among the setting tables stored in the HDD 33. (S91). The CPU 10 sets a third as a set pixel value and a setting threshold value from the first optional area column set in the upper column of the first optional area column and the second optional area column corresponding to the optional entry column of the setting table. The pixel value and the third threshold value are acquired (S92). In the example of this embodiment, it is assumed that the third pixel value and the third threshold value indicating blue are acquired according to the reference of the medical insurance setting table 70.

The CPU 10 executes the above-mentioned area determination process (see FIG. 12) based on the set pixel value and the set threshold value (S93). In the area determination process, a search area range for searching the coordinate position of the pixel having the pixel value corresponding to the third pixel value in the read data is set based on the setting table (S111). Then, the coordinate positions of the pixels having the pixel values corresponding to the third pixel values in the search area range are acquired (S112). An approximate straight line is calculated from the acquired coordinate positions (S113). From the registration form DB, the registration form image data corresponding to the form type and the user ID is acquired (S115). The shape formed by the approximate straight line is compared with the rectangular shape formed by the pixels having the pixel value corresponding to the third pixel value in the registered form image data (S116). The rectangle formed by the pixels having the pixel value corresponding to the third pixel value in the registered form image data corresponds to the shape of the third rectangle 151 provided in the form. It is determined whether or not the shape of the third rectangle 151 in the registered form image data and the shape formed by the approximate straight line based on the read data are within a predetermined same range (S117). When the shape of the third rectangle 151 in the registered form image data and the shape formed by the approximate straight line based on the read data are within the same predetermined range (S117: YES), the shape is surrounded by the shape formed by the approximate straight line. The region is specified as the third region corresponding to the third pixel value (S118).

The CPU 10 determines whether or not the third area is specified in the area determination process (S94). The CPU 10 determines that the third area is specified when the RAM 12 stores information such as coordinate values indicating the position and size in the read data of the third area. When the area corresponding to the third pixel value is not specified in the area determination process (S94: NO), the CPU 10 returns the process to the information identification process (see FIG. 6). It should be noted that this process takes into consideration that the form may not be filled in arbitrarily and the third area may not be specified.

When the area corresponding to the third pixel value is specified in the area determination process (S94: YES), the CPU 10 stores the third area data indicating the image of the third area in the RAM 12 of the third area. Obtained from the read data based on information such as coordinate values (S95). The third area data corresponds to the image of the first optional entry field 150 in the entry form 101.

Based on the third pixel value, the CPU 10 specifies the position and size of the area surrounded by the figure indicated by the set of pixels of the third pixel value, which is the set pixel value, in the registered form image data (S96). ). In this case, the CPU 10 stores information such as coordinate values indicating the position and size in the registered form image data of the area surrounded by the third rectangle 151 corresponding to the third pixel value in the RAM 12. The CPU 10 acquires the registered third area data indicating the image of the area surrounded by the third rectangle 151 from the registered form image data based on the information such as the coordinate values stored in the RAM 12 (S98). The registered third area data corresponds to the image of the first optional entry field 150 in the blank form 100.

The CPU 10 compares the third area data with the registered third area data (S99). In this process, the CPU 10 acquires, for example, histograms of color elements of the third region data and the registered third region data, respectively. The CPU 10 compares each histogram.

The CPU 10 determines whether there is a difference between the third area data and the registered third area data (S101). This determination is made, for example, by whether there is a difference between the histograms. That is, by the process of S101, it is determined whether or not the user has entered the handwritten characters or the like in the first optional entry field 150 of the entry form 101 with respect to the first optional entry field 150 of the unfilled form 100. In the example of this embodiment, the first optional entry field 150 of the blank form 100 shown in FIG. 2 is blank, but the first optional entry field 150 of the entry form 101 shown in FIG. 3 contains the spouse of the user. The character string indicating the name, etc. of is written by hand. In this case, a difference occurs between the histogram of the third region data and the histogram of the registered third region data.

When there is a difference between the third area data and the registered third area data (S101: YES), the CPU 10 sets the third area data as the execution target of the character recognition process in the process of S21 (see FIG. 6) described later. It is decided (S102). The CPU 10 stores in the RAM 12 that the third area data is to be executed for the character recognition process. On the other hand, when there is no difference between the third region data and the registered third region data (S101: NO), the CPU 10 determines the third region as the brightness change display target (S103). The CPU 10 acquires the average value of the brightness values indicated by the entire third region data from the third region data. The CPU 10 stores the average value of the acquired luminance values in the RAM 12 in association with the fact that the third region is determined as the luminance change display target.

In the present embodiment, there are two columns included in the optional entry column of the form, the first optional entry column 150 and the second optional entry column 160 having different outline colors in the form. There are also two types of set pixel values and set thresholds to be handled. The CPU 10 determines whether there is another set pixel value for which the area determination process (S93) has not yet been performed (S104). The CPU 10 refers to the column corresponding to the optional entry column of the setting table stored in the HDD 33, and when there is a column set in a column lower than the column corresponding to the third pixel value and the third threshold value. , It is determined that there is another set pixel value for which the area determination process has not been performed yet. In the example of the present embodiment, since the second optional area column is set below the first arbitrary area column in the medical insurance setting table 70, if there is another set pixel value for which the area determination process has not yet been performed. Suppose it is judged.

When it is determined that there is another set pixel value for which the area determination process has not yet been performed (S104: YES), the CPU 10 acquires another set pixel value and another set threshold value from the setting table (S105). In the example of the present embodiment, the fourth pixel value and the fourth threshold value are acquired as the set pixel value and the set threshold value from the second arbitrary area column of the medical insurance setting table 70. After that, the CPU 10 executes the processes S93 to S104 based on the fourth pixel value and the fourth threshold value. In the example of the present embodiment, both the second optional entry field 160 of the unfilled form 100 shown in FIG. 2 and the second optional entry field 160 of the entry form 101 shown in FIG. It is assumed that the fourth region corresponding to the fourth pixel value is determined as the brightness change display target.

After that, when it is determined that there is another set pixel value for which the area determination process has not yet been performed (S104: NO), the CPU 10 returns the process to the information identification process (see FIG. 6).

Returning to the description of FIG. The CPU 10 determines whether or not the required entry field of the entry form 101 is not filled in (S19). In the present embodiment, the date entry field 120 is not filled in when the instruction to output the required entry field blank error display 83 (see FIG. 16) is stored in the RAM 12 by the process of S83 (see FIG. 10). Is judged. When the required entry field is not blank (S19: YES), the CPU 10 is targeted for character recognition processing in each of the processes of S62 (see FIG. 9), S82 (see FIG. 10), and S102 (see FIG. 11). Character recognition processing is executed for each of the first region data, the second region data, and the third region data (S21). In the present embodiment, the character recognition process is the OCR process described above. The CPU 10 selects the first character string data, the second character string data, and the third character, which are the execution results of the character recognition process, from the first area data, the second area data, and the third area data that are the execution targets of the character recognition process. Acquire column data (S22). These character string data are acquired for registration in the registration form DB in the process of S26 described later. Therefore, these character string data are also referred to as registered character string data. Since the processing of S22 is performed according to the fact that the required entry field is not blank, the registered character string data includes at least the second character string data. Each of the first character string data, the second character string data, and the third character string data is a character string indicating a character string included in each of the pre-entry field 140, the date entry field 120, and the optional entry field of the entry form 101. Corresponds to data. The CPU 10 stores the acquired registered character string data in the RAM 12.

The CPU 10 generates reading result data indicating a reading result display, which indicates the result of the entry form 101 being read by the reading unit 26 of the scanner 2 (S23). The reading result display is an image displayed on the display unit 16 so that the operator can confirm whether the entry form 101 has been correctly read by the scanning unit 26 of the scanner 2. The reading result data is generated based on the reading data acquired in S12.

The first area corresponding to the pre-filled field 140 and the third and fourth areas corresponding to the optional entry fields in the read data are subject to brightness change display by the processing of S63 (see FIG. 9) and S103 (see FIG. 11). May have been decided. The CPU 10 refers to the information stored in the RAM 12 when generating the reading result data. If the RAM 12 stores information that any area is determined to be the brightness change display target, the CPU 10 determines that the area determined to be the brightness change display target is determined to be the brightness change display target. The read data is processed so that it is displayed with a brightness value lower than the brightness value associated with the information indicating the effect. As a result, for example, the area determined to be the brightness change target is displayed in gray out in the reading result display. In the example of this embodiment, it is assumed that the fourth region is determined to be the brightness change target. The first area, the third area, and the fourth area are determined to be the brightness change targets by handwriting in the pre-entry field 140 and the optional entry field of the entry form 101 corresponding to the first area, the third area, and the fourth area. This is the case when characters etc. are not entered. In such a case, by making it difficult for the operator to see the area determined to be the brightness change target by the grayout display, the operator is shown that it is not necessary to pay attention to the area where the grayout display is displayed. .. In addition, the CPU 10 attaches a display 202 (see FIG. 15) for prompting confirmation of the reading result display, and generates reading result data.

The CPU 10 transmits an instruction to output a reading result display based on the reading result data generated in the process of S23 to the display unit 16 via the input / output I / F 14 (S24). FIG. 15 shows a reading result display 200 according to an example of this embodiment. The reading result display 200 includes a result display 201 in which an image based on the reading data is displayed, and the above-mentioned display 202. In the result display 201, the operator can visually recognize the reading result in the pre-entry column 140, the date entry column 120, and the first optional entry column 150 corresponding to the first area, the second area, and the third area of the read data. Displayed in mode. Of the result display 201, the second optional entry field 160 corresponding to the fourth area of the read data is determined to be the brightness change target, so that the brightness is reduced and grayed out.

The operator who confirmed the reading result display on the display unit 16 can input an instruction to the CPU 10 via the operation unit 17. The CPU 10 determines whether an instruction to register each of the first character string data, the second character string data, and the third character string data acquired in the process of S22 in the registration form DB is input via the operation unit 17. (S25).

When a predetermined instruction to register in the registration form DB is input (S25: YES), the CPU 10 determines the registration character string data acquired in the process of S22 in the process of S13 of the registration form DB. Registered in the registration form DB corresponding to the form type in association with the user ID determined in the process of S14 (S26). In the example of this embodiment, the first character string data is registered in each item of the personal information column of the medical insurance registration form DB70. The second character string data is registered in the item of the application date in the form management information column of the medical insurance registration form DB70. The third character string data is registered in each item of the spouse information of the medical insurance registration form DB70. At this time, if the first character string data includes the character string data corresponding to the character string indicating the changed address, etc. added by handwritten characters, the information registered in the personal information column is added. The character string data corresponding to the character string may be updated and registered. The CPU 10 shifts the process to the determination of S29.

On the other hand, the CPU 10 generates reading result data (S27) when the required entry field is not filled in (S19: NO). The reading result data generated in this process is generated based on the reading data and with the display 202 attached, similarly to the reading result data generated in the process of S23. Further, if there is an area determined as the brightness change display target, the CPU 10 determines the brightness value associated with the information indicating that the area determined as the brightness change display target is determined as the brightness change display target. Process the read data so that it is displayed with a lower brightness value. Further, the CPU 10 further refers to the information stored in the RAM 12 when generating the reading result data. If the instruction to output the required entry field blank error display 83 (see FIG. 16) is stored in the RAM 12, the reading result data is displayed so that the reading result display with the required entry field blank error display 83 is displayed. Generate.

The CPU 10 transmits an instruction to output a reading result display based on the reading result data generated in the process of S27 to the display unit 16 via the input / output I / F 14 (S28). FIG. 16 shows a reading result display 300 with a mandatory entry field blank error display 83. The reading result display 300 includes a result display 301 in which an image based on the reading data is displayed, and the above-mentioned display 202. In the present embodiment, when the date entry field 120 of the entry form 101 shown in FIG. 3 is left blank, such a reading result display 300 is displayed on the display unit 16. In the mandatory entry field blank error display 83 of the present embodiment, a predetermined message is written in a graphic frame having a balloon drawn from the date entry field 120, and a so-called pop-up is displayed in front of the result display 301. It is provided as a display. As a result, the mandatory entry field blank error display 83 can clearly indicate to the operator that the required entry is not made in the date entry field 120.

In this case, since the necessary items are not entered in the date entry field 120, the character recognition process like S21 is not performed, and the character string data registration process like S26 is not performed in the registration voting DB. The CPU 10 shifts the process to the determination of S29.

The CPU 10 determines whether to end the information identification process (S29). For example, when an instruction to end the information identification process is input via the operation unit 17, the CPU 10 determines that the information identification process is ended (S29: YES), and ends the information identification process. For example, when an instruction for executing the information identification process for the next form is input via the operation unit 17 via the operation unit 17, the CPU 10 does not end the information identification process (S29: NO). , The process is returned to S11.

As described above, the CPU 10 determines the form type of the entry form based on the read data obtained by reading the image of the entry form 101 by the process of S11 (S13). The CPU 10 acquires the registered form image data according to the user ID from the registered form DB corresponding to the form type determined by the process of S13 (S115). Further, the CPU 10 has a shape formed by an approximate straight line based on the coordinate position of a pixel having a pixel value corresponding to the second pixel value in the read data in the area determination process (S73) executed in the required entry area determination process (S16). The area surrounded by is specified as the second area corresponding to the date entry field 120 (S118). The CPU 10 acquires the second area data indicating the image of the second area from the read data (S75). The second area data corresponds to the image of the date entry field 120 in the entry form 101. The CPU 10 determines the position and size of the area surrounded by the first rectangle 121 in the registered form image data based on the second pixel value corresponding to the color of the first rectangle 121 indicating the outline of the date entry field 120. Identify (S76). The CPU 10 acquires the registration second area data indicating the image of the specified area from the registration form image data (S78). The registration first area data corresponds to the image of the date entry field 120 in the blank form 100. The CPU 10 compares the second area data with the registered second area data (S79), and when there is a difference between the second area data and the registered second area data (S81: YES), character recognition for the second area data The registered character string data including the second character string data which is the execution result of the process is acquired (S22). On the other hand, when there is no difference between the second area data and the registered second area data (S81: NO), the reading result data is generated without performing the character recognition process (S27), and is based on the generated reading result data. An instruction to output the reading result display is transmitted to the display unit 16 (S28). That is, in the entry form 101, when the date entry field 120, which is a required entry field, is filled in with handwritten characters or the like, at least the second character string data corresponding to the character string entered in the date entry field 120 is included. Registration character string data is acquired. On the other hand, if there is no entry in the date entry field 120, the registered character string data is not acquired. For example, by storing the position of the date entry field 120 in association with the registered form image data in advance, it is conceivable that the registered character string data can be acquired without specifying the first rectangle 121 based on the color. In this case, for example, in order to specify the position of the date entry field 120 from the scanned data, it is necessary to specify a reference position such as an edge or a corner of the document. However, if the edges and corners of the document are similar, for example, to the background color of the document and the color of the background member provided in the reading unit 26, the detection accuracy may decrease. In the present embodiment, by specifying the first rectangle 121 based on the color, it is possible to accurately acquire the registered character string data without being influenced by the background color of the document and the color of the background member. In this way, the CPU 10 can identify the types of a plurality of entry fields provided in the form and execute processing according to the types of the entry fields.

When the registered character string data including the second character string data is acquired in the process of S22, the CPU 10 generates the reading result data (S23). The CPU 10 transmits an instruction to output a reading result display based on the generated reading result data to the display unit 16 (S24). Therefore, the operator can confirm by displaying the reading result whether the entry form 101 is correctly read by the reading unit 26 of the scanner 2. Further, when the instruction to be registered in the registration form DB is input via the operation unit 17 (S25: YES) in response to the reading result display being displayed on the display unit 16, the CPU 10 acquires the data in the process of S22. The registered character string data is registered in the registration form DB corresponding to the form type determined in the process of S13 in the registration form DB (S26). As a result, the registration character string data is properly managed in the registration form DB.

When there is no difference between the second area data and the registered second area data (S81: NO), the CPU 10 stores an instruction to output the required entry field blank error display 83 in the RAM 12 (S83). In response to the storage of this instruction, the CPU 10 generates reading result data indicating the reading result display with the required entry field blank error display 83 (S27). The CPU 10 transmits an instruction to output a reading result display with an error display 83 in which the required entry field is not entered to the display unit 16 (S28). Therefore, the reading result display with the required entry field blank error display 83 is displayed on the display unit 16, and the operator is informed that the handwritten characters are not entered in the required entry field of the entry form 101. Will be done.

In the area determination process (S93) executed in the arbitrary entry area determination process (S18), the CPU 10 is surrounded by a shape formed by an approximate straight line based on the coordinate position of a pixel having a pixel value corresponding to the third pixel value in the read data. The designated area is specified as the third area corresponding to the optional entry field (S118). The CPU 10 acquires the third area data indicating the image of the third area from the read data (S95). The third area data corresponds to the image of the first optional entry field 150 in the entry form 101. The CPU 10 determines the position and size of the area surrounded by the third rectangle 151 in the registered form image data based on the third pixel value corresponding to the color of the third rectangle 151 indicating the outline of the first optional entry field 150. Specify the value (S96). The CPU 10 acquires the registered third area data indicating the image of the specified area from the registered form image data (S98). The registered third area data corresponds to the image of the first optional entry field 150 in the blank form 100. The CPU 10 compares the third area data with the registered third area data (S99), and when there is a difference between the third area data and the registered third area data (S101: YES), character recognition for the third area data The registered character string data including the third character string data which is the execution result of the process is acquired (S22). On the other hand, when there is no difference between the third region data and the registered third region data (S101: NO), the CPU 10 determines the third region as the brightness change display target (S103). That is, in the entry form 101, when handwritten characters or the like are entered in the optional entry field, the registered character string data including the character string data corresponding to the character string entered in the optional entry field is acquired. On the other hand, when there is no entry in the optional entry field, an instruction to output a reading result display based on the reading result data generated so that the area corresponding to the optional entry field is grayed out is transmitted to the display unit 16. .. In this case, if the area corresponding to the optional entry field is grayed out, the operator is informed that handwritten characters or the like are not entered in the optional entry field and detailed confirmation in the reading result display is unnecessary. Shown. In this way, the CPU 10 can identify the types of a plurality of entry fields provided in the form and execute processing according to the types of the entry fields.

The CPU 10 extracts the user ID area data corresponding to the user ID field 130 of the entry form 101 from the read data (S41). The CPU 10 generates ID character string data from the user ID area data, and compares the ID character string data with the registered user ID registered in the registration form DB (S45). The CPU 10 acquires a registered user ID that matches the character string indicated by the ID character string data as a user ID corresponding to the read data (S50). The CPU 10 registers the registered character string data in the registered form DB corresponding to the form type in association with the user ID acquired in this way (S26). Therefore, the registered character string data is appropriately managed in association with the form type and the user ID.

When the area corresponding to the second pixel value is not specified in the area determination process (S74: NO), the CPU 10 transmits an instruction to output the entry form reading error display 82 to the display unit 16 (S84). In this case, the entry form reading error display 82 is displayed on the display unit 16, and the operator is notified that the second area corresponding to the date entry field 120 is not specified in the area determination process.

In the above embodiment, the process of S12 corresponds to the "first acquisition step" of the first and second embodiments . Processing S118 in the area determination process performed in mandatory entry region determination process corresponds to "first region specifying step" in the first and second embodiments. The processing of S 81 , S 22 and S 28 corresponds to the "first execution step" of the first and second embodiments. The process of S118 in the area determination process executed in the optional entry area determination process corresponds to the "second area identification step" of the first and second embodiments. The process of S95 corresponds to the "second acquisition step" of the first and second embodiments. The processing of S101, S11, S24, S28 and S103 corresponds to the "second execution step" of the first and second embodiments.

The processing of S23 and S27 corresponds to the "generation step" of the first embodiment. The process of S24 corresponds to the "first output step" of the first embodiment. The process of S26 corresponds to the "memory step" of the first embodiment. The HDD 33 corresponds to the "storage means" of the first embodiment . Processing of S 46 corresponds to the "determination step" of the first embodiment. The process of S50 corresponds to the " third acquisition step" of the first embodiment. The process of S84 corresponds to the " second output step" of the first embodiment.

The reading unit 26 corresponds to the "reading unit" of the third embodiment. The CPU 10 that executes the process of S12 functions as the "first acquisition means" of the third embodiment . CPU10 that performs the processing of S118 in the area determination process performed in mandatory entry region determination process, functions as a "first region specifying part" of the third embodiment. The CPU 10 that executes the processes of S 81 , S 22 and S 28 functions as the "first execution means" of the third embodiment. The CPU 10 that executes the process of S118 in the area determination process executed in the optional entry area determination process functions as the "second area specifying means" of the third embodiment. The CPU 10 that executes the process of S95 functions as the "second acquisition means" of the third embodiment. The CPU 10 that executes the processes of S101, S11, S24, S28, and S103 functions as the "second execution means" of the third embodiment.

The embodiments illustrated herein are not limited to the above reporting, various modifications are possible. In the above embodiment, in each process of S24, S28, S36, S47, S48, S83, and S84, the CPU 10 transmits an instruction to output an image such as a display based on the data corresponding to each process to the display unit 16. There is. The CPU 10 may transmit some or all of these output instructions to the external display unit 19. Further, some or all of these output instructions transmitted by the CPU 10 may be transmitted to the scanner 2. In this case, the CPU 20 of the scanner 2 that has received the output instruction transmits an instruction to output an image such as a display based on the data corresponding to each process to the display unit 27 or the external display unit 29 via the input / output I / F 24. You may.

Further, the scanner 2 may function as the information identification device 1, and the CPU 20 of the scanner 2 may execute the information identification process. In this case, the output instruction in each process of S24, S28, S36, S47, S48, S83, and S84 may be transmitted by the CPU 20 to the display unit 27 or the external display unit 29 via the external communication I / F25. In this case, the scanner 2 corresponds to the "image reader" of the third embodiment. The CPU 10 that executes the processes of S23 and S27 functions as the "generation means" of the third embodiment. The external communication I / F 25 corresponds to the "communication unit" of the third embodiment. In addition, the management server 3 may function as the information identification device 1.

For each area determined to be the brightness change target in the processing of S63 and S103, the reading result data is generated so as to be displayed with a brightness lower than the brightness at the time of acquisition of each area data in the processing of S23 and S27. However, it is not limited to this. For example, the reading result data may be generated so that each region determined to be the brightness change target is displayed with a brightness higher than the brightness at the time of acquiring each region data. Even in this case, it becomes difficult for the operator to visually recognize the area determined to be the brightness change target, and the operator is shown that it is not necessary to pay attention to the corresponding area.

When the processing of S84 is performed, there are various reasons why the read data cannot be properly acquired. The entry form reading error display 82 may have display contents such as "The date entry field cannot be recognized" that simply indicates that the area corresponding to the date entry field 120 has not been specified. In addition, the entry form reading error display 82 specifically informs the operator of the reason why the read data has not been properly acquired, "It is a form scanning error", and "Check the connection status between the scanner and the PC." It may display a message such as ". As described above, the entry form reading error display 82 may include various message displays for notifying that the area corresponding to the date entry field 120 is not specified.

In the above embodiment, the mandatory entry field blank error display 83 is configured by a predetermined pop-up display, but the present invention is not limited to this. The required entry field blank error display 83 is displayed in association with the second area corresponding to the required entry field in the reading result display 300, and informs the operator that the required entry field is not filled in. May include indications of various aspects.

In the above embodiment, the character recognition process is executed in the processes of S21 and S42, but the character recognition process does not necessarily have to be performed in the information identification device 1. For example, in the management server 3 and other external devices connected to the information identification device 1 via the network, character recognition processing is executed for the character recognition target, the information identification device 1 receives the character recognition result, and the subsequent information. It may be configured to perform identification processing.

The CPU 10 functions as a processor that executes information identification processing by expanding the program stored in the HDD 13 on the RAM 12. A general-purpose processor may be used as the CPU 10. A microcomputer, ASIC, FPGA or the like may be used as a processor instead of the CPU 10. When the scanner 2 functions as the information identification device 1, the CPU 20 functions as a processor that executes the information identification process by expanding the program stored in the ROM 21 on the RAM 22. When the management server 3 functions as the information identification device 1, the CPU 30 functions as a processor that executes the information identification process by expanding the program stored in the HDD 33 on the RAM 32. Similar to the above, a microcomputer, ASIC, FPGA or the like may be used as a processor instead of the CPU 30.

The information identification process may be distributed by a plurality of electronic devices (that is, a plurality of CPUs). For example, a part of the information identification process may be executed by another server such as the management server 3 connected to the network 9. The program may be downloaded from, for example, another server such as the management server 3 connected to the network 9 and stored in the HDD 13 of the information identification device 1.

1 Information identification device 2 Scanner 3 Management server 9 Network 10, 20, 30 CPU
13,33 HDD
16,27 Display unit 19,29 External display unit 21 ROM
26 Reading unit 82 Entry form reading error display 83 Required entry field blank error display 101 Entry form 200, 300 Reading result display

Claims (7)

An information identification program executed by the computers that make up the information identification device.
On the computer
The first acquisition step of acquiring the reading data showing the image of reading the entry form, which is a form in which characters including handwritten characters are entered, and
A first area identification step for specifying the first area surrounded by the outline corresponding to the first pixel value in the read data as a target area for character recognition processing, and
The first area data showing the image of the first area and the form number corresponding to the first area in the form image data showing the image of the form in which the handwritten characters are not entered, which is the form corresponding to the entry form. Based on the first data of the form showing the image of one area, the first execution step of acquiring the first character string data indicating the character string including at least the first character string included in the first area based on the character recognition process. ,
A second area specifying step for specifying a second area surrounded by a contour corresponding to a second pixel value different from the first pixel value in the read data as a target area for character recognition processing, and a second area specifying step.
A second acquisition step of acquiring second region data showing an image of the second region from the read data, and
The second area data is compared with the second form data showing the image of the second form area corresponding to the second area in the form image data, and the second area data and the second form data are different from each other. In this case , if the second character string data indicating the character string including at least the second character string included in the second area is acquired based on the character recognition process, while the second area data and the form second data are different. If not, the feature is that the second execution step of outputting the instruction to display the second area data is executed in a mode different from the mode indicated by the second area data acquired in the second acquisition step. Information identification program. On the computer
A generation step of generating reading result data including at least the first area data, which is data indicating the reading result of the entry form, and
When the first character string data is acquired in the first execution step, the first output step for outputting an instruction to display the reading result data and the first output step.
The first aspect of claim 1, wherein when a predetermined instruction is received in response to the output of the instruction for displaying the reading result data, the storage step of storing the first character string data in the storage means is executed. Described information identification program. The generation step
When it is determined that the first area data and the first form data are not different, the reading result data in which the information for giving a predetermined notification regarding the first area is associated with the first area data is generated. The information identification program according to claim 2. On the computer
A determination step for determining the user of the entry form based on the read data, and
The third acquisition step of acquiring the user identification information that identifies the user according to the determination result in the determination step is executed.
The memory step
The information identification program according to claim 2, wherein the first character string data is associated with the user-specific information and stored in the storage means. On the computer
The claim is characterized in that when the first region is not specified in the first region specifying step, a second output step for outputting an instruction to execute a notification based on the fact that the first region is not specified is executed. The information identification program according to any one of 1 to 4. It is a method that can be executed by the computers that make up the information identification device.
The first acquisition step of acquiring the reading data showing the image of reading the entry form, which is a form in which characters including handwritten characters are entered, and
A first area identification step for specifying the first area surrounded by the outline corresponding to the first pixel value in the read data as a target area for character recognition processing, and
The first area data showing the image of the first area and the form number corresponding to the first area in the form image data showing the image of the form in which the handwritten characters are not entered, which is the form corresponding to the entry form. Based on the first data of the form showing the image of one area, the first execution step of acquiring the first character string data indicating the character string including at least the first character string included in the first area based on the character recognition process. ,
A second area specifying step for specifying a second area surrounded by a contour corresponding to a second pixel value different from the first pixel value in the read data as a target area for character recognition processing, and a second area specifying step.
A second acquisition step of acquiring second region data showing an image of the second region from the read data, and
The second area data is compared with the second form data showing the image of the second form area corresponding to the second area in the form image data, and the second area data and the second form data are different from each other. In this case , if the second character string data indicating the character string including at least the second character string included in the second area is acquired based on the character recognition process, while the second area data and the form second data are different. If not, it is characterized by including a second execution step that outputs an instruction to display the second area data in a mode different from the mode indicated by the second area data acquired in the second acquisition step. Information identification method to be performed. A reader that reads the image of the form and
A first acquisition means for acquiring read data indicating an image obtained by reading an entry form, which is a form in which characters including handwritten characters are entered, which is read by the reading unit.
A first area specifying means for specifying a first area surrounded by a contour corresponding to the first pixel value in the read data as a target area for character recognition processing, and
The first area data showing the image of the first area and the form number corresponding to the first area in the form image data showing the image of the form in which the handwritten characters are not entered, which is the form corresponding to the entry form. Based on the first data of the form showing the image of one area, the first execution means for acquiring the first character string data indicating the character string including at least the first character string included in the first area based on the character recognition process. ,
A second area specifying means for specifying a second area surrounded by a contour corresponding to a second pixel value different from the first pixel value in the read data as a target area for character recognition processing, and
A second acquisition means for acquiring second region data indicating an image of the second region from the read data, and
The second area data is compared with the second form data showing the image of the second form area corresponding to the second area in the form image data, and the second area data and the second form data are different from each other. In this case , if the second character string data indicating the character string including at least the second character string included in the second area is acquired based on the character recognition process, while the second area data and the form second data are different. If not, it is characterized by providing a second execution means for outputting an instruction to display the second area data in a mode different from the mode indicated by the second area data acquired by the second acquisition means. Image reader.

Images