JP2019101647A - Information processing device, control method therefor, and program - Google Patents

Abstract

To provide a technique for automatically determining whether or not a writing format of before-writing data and a writing format of after-writing data are consistent with each other.SOLUTION: An information processing device for acquiring information obtained from a difference between before-writing data and after-writing data: reads data portions in the before-writing data and the after-writing data; associates positions of feature points existing commonly in the read data portions of the before-writing data and the after-writing data between the read before-writing data and the read after-writing data; alters the positions of the feature points in the read after-writing data so as to be consistent with the before-writing data from which the positions of the associated feature points are read; performs binarization to a gradation indicating a portion including data and a gradation indicating a portion lacking data in the read before-writing data and the read after-writing data after altering; and determines the consistency in writing formats of the read before-writing data and the read after-writing data in accordance with the degree of overlapping of the binarized before-writing data and the binarized after-writing data.SELECTED DRAWING: Figure 3

Description

  The present invention relates to a technique for automatically determining whether or not a set of pre-entry data and post-entry data whose entry form is correct.

  When using OCR (Optical Character Recognition) apparatus that reads characters from a form, setting layout information of the form type (entry form) to be input by scanning etc. in advance to the OCR apparatus is equivalent to OCR It is common.

  Although the type of form to be used can be given explicitly by the user, a method of automatically identifying it by the machine is also conceivable. The automatic identification is convenient because it is not necessary to fix the number of pages and read a form or to determine the order when inputting form sheets in batch units. The main thing of the automatic identification method of form is to extract feature points that are useful for identification from the input form image, and select the closest form point among the candidate group of form type (entry form), It is identified as the type of form (entry form) to be processed.

  However, depending on the layout of the form type (entry form), some feature points may be difficult to obtain. For example, it is a form of a layout configured by underlining.

  Patent Document 1 discloses a technique of identifying based on a logical operation and a mismatch measure as a method of enabling form identification even in the form (entry form) of forms having layouts in which characteristic points are difficult to obtain as described above. ing.

JP, 2013-196611, A

  Thus, Patent Document 1 prepares three types of data: a pre-entry form image (data before entry), an entry example form image, and an identification target form image. Each form image is divided into grids, and it is determined whether each small area is a foreground area (1) or a background area (0). You will then have a grid of 0s and 1s. By performing logical operations between the entry front form grid and the identification target form grid, and between the entry example form grid and the identification target form grid, the amounts of small areas in which 0 and 1 do not match between the grids are obtained. Based on the amount of non-matching small areas, measure the degree of non-match on a specific scale. A technique is described that determines the degree of disagreement between the entry front form image and the identification target form image, and determines whether the identification target form image is generated by the entry front form image based on the degree of mismatch.

However, in this method, in order to improve the accuracy, it is necessary to create an entry example form image in advance, which is cumbersome.
Furthermore, although area division is performed in order to absorb paper displacement and differences in size between images, it is necessary to set a plurality of grain sizes for area division, which is complicated.
Furthermore, it does not correspond to the shift of inclination etc., and there is a problem in accuracy

  Therefore, it is desirable to automatically correct the difference in size, inclination shift, and position shift of the scanned identification target form image without creating the entry example form image.

  The present invention uses both pre-entry data and post-entry data to obtain information from pre-entry data and post-entry data difference, and sets the pre-entry data and post-entry data in the form of entry. It is an object of the present invention to provide a mechanism capable of automatically determining whether or not

  The present invention is an information processing apparatus for acquiring information obtained from a difference between pre-entry data and post-entry data, and a reading means for reading a data part in the pre-entry data and post-entry data; Correspondence means for correlating the position of feature points commonly present in the data portion of the data and post-entry data between the read pre-entry data and the post-entry data, and the position of the associated feature points Deformation means for deforming the position of the feature point of the read post-entry data so as to match the read pre-entry data, and the read pre-entry data and post-entry data after the transformation. Binarization means for binarizing into a gradation indicating a certain part and a gradation indicating a part without data, the data before the binarization and the binarization According multiplicity of filled-out data, characterized in that it comprises, a judging means for judging the consistency of the filling type and the read entry data before and filled-out data.

  The present invention is a control method of an information processing apparatus for acquiring information obtained from a difference between pre-entry data and post-entry data, and the reading means of the information processing device is data in the pre-entry data and post-entry data. A reading step of reading a position, and a position of feature points common to data positions of the read pre-entry data and the post-entry data, the association means of the information processing apparatus, And an associating step of associating the post-entry data, and the post-entry data of the read-out data so that the transformation means of the information processing apparatus matches the positions of the associated feature points with the read pre-entry data. A transformation step of transforming the position of the feature point, and the read pre-entry data and post-entry data after the transformation by the binarizing means of the information processing apparatus And a binarization step of binarizing to a gradation indicating a portion having data, and a gradation indicating a portion having no data, the pre-filling data binarized by the determination unit of the information processing apparatus, and And determining the consistency of the entry form between the read pre-entry data and the post-entry data according to the degree of duplication of the post-entry data that has been binarized.

  The present invention is a program that can be read and executed by an information processing apparatus that acquires information obtained from the difference between the pre-entry data and the post-entry data, and the information processing apparatus comprises: Correlating the reading means for reading the part and the positions of the feature points common to the data part of the read pre-entry data and post-entry data, between the read pre-entry data and the post-entry data Means, and deformation means for deforming the position of the feature point of the read post-entry data so that the positions of the associated feature points are matched with the read pre-entry data; and In the pre-entry data and post-entry data, binarization is performed to binarize to a tone indicating a location where data is present and a tone indicating a location where data is not present Determining means for determining the consistency of the entry form between the read pre-entry data and the post-entry data according to the stage and the degree of duplication of the binarized pre-entry data and the binarized post-entry data; Make it work.

  According to the present invention, in order to obtain information from the difference between pre-entry data and post-entry data, using both pre-entry data and post-entry data, a set of pre-entry data and post-entry data whose entry form is correct It is possible to provide a mechanism capable of automatically determining whether or not.

It is a figure which shows an example of the system (handwritten character OCR system) structure of form identification in embodiment of this invention. It is a block diagram showing an example of hardware constitutions, such as an OCR device in an embodiment of the present invention. It is a flow chart which shows an example of processing of form identification in an embodiment of the present invention. It is a flow chart which shows an example of processing to fit in an embodiment of the present invention. It is a flowchart which shows an example of calculation of a non-matching rate in embodiment of this invention. It is a figure which shows an example of a pre-entry form information database in embodiment of this invention. It is a figure which shows an example of a pre-entry form image (cover) in the embodiment of the present invention. It is a figure which shows an example of a pre-entry form image (text) in the embodiment of the present invention. It is a figure which shows an example of the identification target slip image in embodiment of this invention. FIG. 7 is a diagram illustrating an example of a post-deformation identification target form image obtained by the fitting process according to the embodiment of the present invention. It is a figure showing an example of the matching processing of the feature point for modification in the embodiment of the present invention. It is a figure showing an example of processing which takes product of a pixel of a pre-entry form image (front cover) and a post-deformation identification target form image (front cover) in the embodiment of the present invention and generates an AND image (1). It is a figure showing an example of processing which takes the product of the pixels of a pre-entry form image (text) and a post-deformation identification target form image (cover) in the embodiment of the present invention and generates an AND image (2). It is a figure showing an example of processing which takes exclusive OR of pixels of AND picture (1) and a discernment subject slip (cover) after modification in an embodiment of the present invention, and generates an XOR picture. It is a figure showing an example of processing which takes exclusive OR of pixels of AND picture (2) and a discernment subject form (cover page) after modification in an embodiment of the present invention, and generates an XOR picture. It is a figure which shows an example which shows the positional relationship of the pixel before deformation | transformation in embodiment of this invention, and a deformation | transformation, and a formula which calculates non-matching rate.

  FIG. 1 will be described.

  FIG. 1 is a diagram showing an example of a system configuration of a form identification system according to an embodiment of the present invention.

  An OCR apparatus (information processing apparatus) 1001 for performing main processing of handwritten character extraction and a scanner 3001 for scanning a form and converting it into an image file (image scan) are connected via a communication path 4001. There is.

  The communication path 4001 can take the form of a wired LAN, a wireless LAN, a USB or the like according to the physical interface of the scanner 3001.

  A file server 5001 may be placed on the communication path 4001. As a method of loading the image scanned by the scanner 3001 into the OCR device 1001, a method of transmitting an image directly from the scanner 3001 to the OCR device 1001, storing the image file loaded by the scanner 3001 in the file server 5001 once Either method may be used to extract the image file from the server 5001.

  FIG. 2 will be described.

  FIG. 2 is a block diagram showing an example of the hardware configuration of the OCR apparatus 1001 according to an embodiment of the present invention. The file server 5001 has a similar configuration.

  The CPU 1 centrally controls the devices and controllers connected to the system bus 4. In addition, the ROM 2 or the external memory 11 has a BIOS (Basic Input / Output System) which is a control program of the CPU 1, an operating system program (hereinafter referred to as an OS), and a function executed by each server or each OCR device. The necessary various programs described later are stored.

  The RAM 3 functions as a main memory, a work area, etc. of the CPU 1. The CPU 1 loads programs necessary for execution of processing from the ROM 2 or the external memory 11 to the RAM 3 and executes the loaded programs to realize various operations.

  An input controller 5 controls inputs from a keyboard (KB) 9, a camera device 12 (imaging apparatus), a pointing device such as a mouse (not shown), and the like. A video controller 206 controls display on a display such as a CRT display (CRT) 10 or the like. Although the CRT 10 is described, the display may be not only a CRT but also another display such as a liquid crystal display. These are used by the administrator as needed.

  The memory controller 7 is an external storage device (hard disk (HD)) for storing a boot program, various applications, font data, user files, editing files, various data, etc., a flexible disk (FD), or an OCR device MCIA card slot Control access to an external memory 11 such as a Compact Flash (registered trademark) memory connected via an adapter.

  The communication I / F controller 8 is for connecting and communicating with an external device via a network (for example, the LAN shown in the figure), and executes communication control processing in the network. For example, communication using TCP / IP is possible.

  The CPU 1 enables display on the CRT 10 by executing, for example, outline font rasterization processing on a display information area in the RAM 3. Further, the CPU 1 enables an instruction by the user with a mouse cursor (not shown) on the CRT 10 or the like.

  Various programs to be described later for realizing the present invention are stored in the external memory 11, and are executed by the CPU 1 by being loaded into the RAM 3 as necessary. Furthermore, definition files and various information tables used at the time of execution of the program are also stored in the external memory 11, and a detailed description of these will be described later.

  Furthermore, an executable program according to the present invention is a program that causes a computer to execute the processing of FIGS. 3 to 4. The program in the present invention may be a program for each process of FIGS. 3 to 4.

  As described above, the recording medium recording the program for realizing the functions of the above-described embodiments is supplied to the system or apparatus, and the computer (or CPU or MPU) of the system or apparatus stores the program stored in the recording medium. It goes without saying that the object of the present invention can also be achieved by reading and executing.

  In this case, the program itself read out from the recording medium realizes the novel function of the present invention, and the recording medium recording the program constitutes the present invention.

  As a recording medium for supplying the program, for example, a flexible disk, hard disk, optical disk, magneto-optical disk, CD-ROM, CD-R, DVD-ROM, magnetic tape, non-volatile memory card, ROM, EEPROM, silicon A disk etc. can be used.

  Further, by executing the program read by the computer, not only the functions of the above-described embodiment are realized, but also an operating system (OS) or the like running on the computer is actually executed based on the instructions of the program. It goes without saying that the processing is partially or entirely performed, and the processing realizes the functions of the above-described embodiments.

  Furthermore, after the program read from the recording medium is written to the memory provided to the function expansion board inserted into the computer or the function expansion unit connected to the computer, the function expansion board is read based on the instruction of the program code. It goes without saying that the case where the CPU or the like provided in the function expansion unit performs part or all of the actual processing and the functions of the above-described embodiment are realized by the processing.

  Further, the present invention may be applied to a system constituted by a plurality of devices or to an apparatus comprising a single device. It goes without saying that the present invention can also be applied to the case where it is achieved by supplying a program to a system or apparatus. In this case, by reading a recording medium storing a program for achieving the present invention into the system or apparatus, the system or apparatus can receive the effects of the present invention.

  Further, by downloading and reading out a program for achieving the present invention from a server on a network, a database or the like by a communication program, the system or apparatus can receive the effects of the present invention. In addition, the structure which combined each embodiment mentioned above and its modification is also contained in this invention altogether.

  Next, processing executed by the OCR device 1001 according to the embodiment of the present invention will be described with reference to the flowcharts of FIGS.

  The flowchart in FIG. 3 is processing that the CPU 201 of the OCR device 1001 reads and executes a predetermined control program, and is a flowchart showing processing for form identification.

  As an example of the use case, consider that the identification target form is a school report, and a task to identify the cover or the text.

  The post-entry data described below is composed of continuous pages. Further, the post-entry data is information obtained from the difference with the pre-entry data for each successive page of the pre-entry data of a plurality of types (for the cover and for the text) having different entry forms (handwritten student ID number, handwritten name, It is assumed that the handwritten score etc. has already been entered.

  In step S301, the CPU 201 in FIG. 2 receives a form identification request from the user.

  In step S3011, the number of pages of consecutive pages in the data portion of the data after entry is selected.

In step S302, the CPU 201 in FIG. 2 reads identification target form information into the memory.
It is assumed that the identification target form information includes an identification target form image 501 (this time is a cover) scanned by the scanner 1101 and information that belongs to the group 1 among the information 321 of the group to which the form belongs. Further, it is assumed that the identification target form image (post-entry data) 501 has an inclination associated with the scan. Furthermore, it reads the data part (ruled line, entry frame, item name, handwritten student identification number, handwritten name, handwritten score, etc.) in the data after entry (reading means).

  The database table 301 of FIG. 6 is a table showing an example of storing pre-entry form information candidates. In this example, an image ID, an image file name, an index 321 of a group, and information on whether a cover or a body is included as attributes.

  As described above, the database table 301 stores a plurality of kinds of pre-entry data different in entry form as entry form information candidates (storage means).

  The images shown in FIGS. 7A and 7B are obtained by reading out the pre-entry form image belonging to group 1 in the database table 301. The image of the record 311 corresponds to the entry form image (data before entry) 401 (cover), and the image of the record 312 corresponds to the entry form image 402 (text).

  In step S303, the CPU 201 in FIG. 2 extracts the entry form information candidate from the file server 5001 and stores it in a queue on the memory. The file server 5001 is configured as shown in the database table 301.

  Based on the identification target form information given in step S302, candidates for the pre-entry form information candidate corresponding to the group 1 to which the form belongs are extracted.

  In step S304, the CPU 201 in FIG. 2 refers to the queue prepared in step S303, and determines whether there is a record of pre-entry form information. If the record remains in the queue, the process proceeds to step S305. If not, the process proceeds to step S309.

  In step S305, the CPU 201 in FIG. 2 refers to the queue prepared in step S303, extracts one head record, and stores the record in the memory.

  Furthermore, the data portion (ruled line, entry frame, item name, etc.) in the data before entry is read (reading means).

  In step S306, the CPU 201 in FIG. 2 passes the identification target form information read in step S302 and one record of the pre-entry form information read in step S305 to the "fitting" subprocess, and the identification target form image The post-deformation identification form image 502, which is a fitting target pixel of the entry pre-form image, is received as a return value, and this return value is stored in the memory.

  In step S307, the CPU 201 in FIG. 2 passes one record of the entry form information read in step S305 and the post-deformation identification form image 502 received in step S306 to the "calculation of mismatch rate" sub-process. Calculate the mismatch rate of the two images.

  In step S308, the CPU 201 in FIG. 2 stores the non-matching rate calculated in step S308 in the memory in association with the record of the entry form information. Thereafter, the process is passed immediately before step S303.

  In step S309, the CPU 201 in FIG. 2 extracts the one with the smallest mismatch rate out of the pair of the record of the entry form information stored in step S308 and the mismatch rate, and stores it in the memory.

  That is, here, reading is performed by comparing the multiplicity (mismatch rate) between the plural kinds of pre-entry data different in the binarized entry form and the binarized post-entry data which have already been determined in step S506. The pre-entry data in which the entry form is matched among the plurality of kinds of pre-entry data which are different from the post-entry data and the entry form will be selected (selection means).

  In step S310, the CPU 201 in FIG. 2 compares the minimum mismatch rate stored in step S309 with a threshold value set in advance. This threshold is used to determine whether the form is or not. If the non-decision rate is less than the threshold value, the process proceeds to step S311, otherwise proceeds to S312.

  In step S311, the CPU 201 in FIG. 2 outputs a recognition result that the image of the entry pre-form information stored in step S308 is the corresponding form. At this time, the noncoincidence rates for a plurality of candidates are compared, and the client terminal 7001 is controlled to be able to display the result (incoincidence rates for a plurality of candidates).

  Furthermore, here, since the image of the entry pre-form information stored in step S308 is the corresponding form, it is obtained from the difference between the binarized pre-entry data selected in step S309 and the binarized post-entry data. Information (handwritten student identification number, handwritten name, handwritten score, etc.) may be acquired (acquisition means).

  In step S312, the CPU 201 in FIG. 2 determines that the image of the identification target form information read in step S302 does not match any of the pre-entry form information registered in the file server 5001, "OTHER (not registered) And causes the client terminal 7001 to display a recognition result of “OTHER (unregistered)”.

  In step S313, it is determined whether there is an unprocessed page in the post-entry data. If it is determined that there is an unprocessed page, the process returns to S3011, and the subsequent processes are repeated. If it is determined that there is no unprocessed page, the process ends.

  As described above, when the post-entry data is a continuous page (cover, text, etc.), the pre-entry data having good consistency between the data location of the post-entry data read for each page and the entry form is entered by the determination means. According to the determination result of the form consistency, it will be selected in page units from a plurality of types of pre-filling data different in the form of entry.

The flowchart of FIG. 4 is a process in which the CPU 201 of the OCR device 1001 reads and executes a predetermined control program, and is a flowchart showing a process of fitting two images to one another in pixel units. In step S401, the CPU 201 of FIG. In step S302, the identification target form image 501 in the identification target form information extracted in step S302, and the feature point group 602 is extracted. The feature point group 602 is a part of the identification target form image 501, and feature points are extracted from the entire image. The feature point is preferably an end point of a line segment in an image, an intersection point, or a point on a line which is a boundary of luminance or color. KAZE, ORB (Oriented-BRIEF), SIFT (Scale-invariant feature transform), etc. are known as methods for extracting points meeting such conditions. The quality of feature points does not have to be a quality that is useful for identification because the purpose is to be able to correspond feature points.

  In step S402, the CPU 201 in FIG. 2 extracts the feature point group 601 from the entry form image 401 in the same manner as in step S401.

  In step S403, the CPU 201 in FIG. 2 associates feature points that are estimated to be detected from the same part of the form before and after entry between the feature point group 601 and the feature point group 602. As an algorithm of feature point correspondence, KNN (K-Nearest Neighbor algorithm), Blute-Force, FLAN (Fast Library for Approximate Neighbors Neighbors), etc. are known.

  In other words, here, the positions of the feature points that exist in common between the read pre-entry data and the post-entry data point are associated between the read pre-entry data and the post-entry data (Correspondence Attached means).

  FIG. 10 is a diagram showing an example of the process of associating feature points in step 403. Feature points detected from handwritten characters on the identification target form image 501 are not detected from the entry front form image 401, so they are excluded from association here, and feature point correspondence information corresponding to ruled lines and headings originally present in the form 603 is obtained.

  In step S404, the CPU 201 in FIG. 2 calculates a transformation matrix for transforming the identification target form image 501 so as to overlap with the entry form image 401 from the coordinates of corresponding feature points. As a method for calculating a transformation matrix, RANSAC, Least Median, etc. are known.

  In step S405, a transformation matrix T is applied to the identification target form image 501 to perform transformation. The transformation performed here is to apply a single transformation matrix to the entire image. By this deformation, there is an advantage that correction can be made even when the difference in size, the shift, the large inclination generated at the time of scanning, or the reverse scanning of the top and bottom. The image obtained here is referred to as a post-deformation identification target form image 502.

  That is, here, the positions of the feature points of the read-in data are corrected so as to match the positions of the feature points associated in step S403 with the read data before writing (deformation means).

  Expression 1 is an expression representing the relationship between an arbitrary point (x, y) on the identification target form image 502 after deformation and a corresponding point (x ′, y ′) on the identification target form image 501 before deformation.

  Expression 2 is an expression representing the color C2 (x, y) at an arbitrary point (x, y) on the post-deformation identification form image 502. C2 (x, y) is C1 (x ', y') at a point (x ', y') on the identification target form image 501 before deformation.

  The flowchart of FIG. 5 is a process executed by the CPU 201 of the OCR apparatus 1001 for reading and executing a predetermined control program, and is a flowchart showing a process of calculating a mismatch rate of an identification target image with an original form.

  In step S501, the CPU 201 in FIG. 2 applies a binarization process to the pre-entry form image 401 and the post-transformation identification target form image 502, and the pre-form image 701 for binary entry and the post-binary identification target form An image 702 is generated.

  That is, here, after the transformation in step S405, the read pre-entry data and the post-entry data are binarized into a gradation indicating a portion having data and a gradation indicating a portion not having data. (Binarization means).

  In step S502, the CPU 201 in FIG. 2 generates an AND image 703 in which the logical product (AND) of the pixels is obtained with respect to the binarized entry previous form image 701 and the binarized deformed identification target form image 702. In the AND image, as the identification target form image 702 after binarization contains more information with the binarized entry pre-form image 701, the binarized entry pre-form image 701 and the AND image 703 become closer. Conversely, when ANDing the different binarized form image 711 with the identification target form image 702 after binarization transformation, an AND image 712 different from the binarized form image 711 is formed.

  In step S 503, the CPU 201 in FIG. 2 generates an XOR image 704 obtained by taking an exclusive OR (XOR) of pixels with respect to the binarized entry previous form image 701 and the AND image 703. In the XOR operation, when the pixels are different from each other in the two images, they are regarded as non-matching pixels (in this example, black pixels). Therefore, the smaller the number of unmatched pixels, the closer to the binarized entry form image 701. When XOR is generated with a different AND image 712, many unmatched pixels remain.

  In step S504, the CPU 201 in FIG. 2 aggregates the amounts of all foreground pixels (black pixels in this example) in the binarized entry form image 701, and stores the amounts in the memory.

  In step S505, the CPU 201 in FIG. 2 aggregates the amounts of all unmatched pixels (black pixels in this example) in the XOR image 704 obtained in step S503, and stores the amounts in the memory.

  In step S506, the CPU 201 in FIG. 2 calculates the amounts of all foreground pixels in the binarized entry form image 701 obtained in step S504 and the amounts of all unmatched pixels in the XOR image 704 obtained in step S505. The mismatch rate is determined by Equation 3 using In this example, by dividing the latter by the former, the ratio of how many error pixels remain as a logical operation result is determined.

  That is, here, the consistency of the entry format between the read pre-entry data and the post-entry data is determined according to the duplication degree (mismatch rate) of the pre-entry data binarized in step S501 and the binarized post-entry data. (Determination means).

In the present embodiment, the following five points become possible. Each student has different number of pages, and can input a bundle of sheets with mixed cover and text. There is no need to prepare an entry example form image 3. At the time of scanning of the identification target form image, there may be a difference in inclination, deviation, or size.
4. It is not necessary to extract features required for identification accurately. Compare non-match rates in multiple layouts and warn on errors

1001 OCR main unit (information processing unit)
3001 Scanner body (image reader)
4001 Network Path 5001 File Server (Reading Image Management Device)
6001 server (OCR result management device)
7001 client terminal

Claims (5)

An information processing apparatus for acquiring information obtained from a difference between pre-entry data and post-entry data,
Reading means for reading data locations in the pre-entry data and post-entry data;
Association means for associating the positions of feature points commonly present in the data locations of the read pre-entry data and post-entry data, between the read pre-entry data and the post-entry data;
Deforming means for deforming the positions of the feature points of the read post-entry data so as to align the positions of the correlated feature points with the read pre-entry data;
A binarizing means for binarizing to a gradation indicating a portion having data and a gradation indicating a portion having no data in the read pre-entry data and the post-entry data after the transformation;
A determination unit that determines the consistency of the entry format between the read pre-entry data and the entry data according to the degree of duplication of the pre-digitized data and the binarized post-entry data;
An information processing apparatus comprising: Storage means for storing a plurality of types of pre-entry data different in entry form;
The read-out data and the entry format are different from each other by comparing the duplication degree between the plurality of kinds of pre-entry data different in the entry format and the binarized entry format and the data after the entry after the binarization Selection means for selecting the pre-entry data matching the entry form in the pre-entry data;
Acquisition means for acquiring information obtained from the difference between the selected pre-binarized data before entry and the binarized post-entry data;
The information processing apparatus according to claim 1, further comprising: The post-entry data is composed of continuous pages in which information obtained from the difference is described in a plurality of kinds of pre-entry data different in the entry form,
The reading unit reads the data portion of the post-entry data in page units;
The data location of the post-entry data read for each page and the pre-entry data having good consistency with the entry form, according to the result of the determination of the entry form integrity of the decision means, a plurality of types of entry with different entry forms The information processing apparatus according to claim 2, wherein the information is selected in the page unit from the previous data. A control method of an information processing apparatus for acquiring information obtained from a difference between data before entry and data after entry,
A reading step of reading a data portion in the pre-entry data and the post-entry data by a reading unit of the information processing apparatus;
The association means of the information processing apparatus associates the positions of feature points commonly present in the data locations of the read pre-entry data and post-entry data between the read pre-entry data and the post-entry data Matching step,
A deforming step of deforming the positions of the feature points of the read-in data, so that the deformation means of the information processing apparatus aligns the positions of the correlated feature points with the read data before writing;
In the pre-writing data and the post-writing data read after the transformation, the binarization means of the information processing apparatus is binary at the gradation indicating the location with data and the gradation indicating the location without data The binarization process to
The determination means of the information processing apparatus determines the consistency of the entry format between the read pre-entry data and the post-entry data according to the duplication degree of the binarized pre-entry data and the binarized post-entry data Judgment step to
A control method characterized by including. It is a program that can be read and executed by an information processing apparatus that acquires information obtained from the difference between pre-entry data and post-entry data,
The information processing apparatus
Reading means for reading data locations in the pre-entry data and post-entry data;
Association means for associating the positions of feature points commonly present in the data locations of the read pre-entry data and post-entry data, between the read pre-entry data and the post-entry data;
Deforming means for deforming the positions of the feature points of the read post-entry data so as to align the positions of the correlated feature points with the read pre-entry data;
A binarizing means for binarizing to a gradation indicating a portion having data and a gradation indicating a portion having no data in the read pre-entry data and the post-entry data after the transformation;
A determination unit that determines the consistency of the entry format between the read pre-entry data and the entry data according to the degree of duplication of the pre-digitized data and the binarized post-entry data;
Program to make it work.

Images