JP7393960B2 - Information processing device, information processing method, and program - Google Patents

Description

The present invention relates to an information processing apparatus and the like that reads an image of a paper such as a bill, performs OCR on the read image, and obtains information such as the amount written on the paper.

For example, in a financial institution such as a bank, each store has a form containing the number of sheets and the total amount written on each sheet regarding bundles of paper such as bills and checks received from customers. The bundle of paper is then brought to the office center along with the form. At the office center, the number of sheets and the total amount of the bundle of sheets brought in are recalculated and compared with the number of sheets and the total amount written on the form.

At the administrative center, in order to save labor in the administrative work described above, images of two or more sheets are read in using a scanner, etc., and OCR is performed on the read images to obtain the information written on the sheets. An information processing device may be used.

In this type of information processing device, the information obtained from the paper image by OCR generally does not necessarily match the information written on the paper due to the lack of accuracy of OCR, so it does not require the effort of human checking. It takes.

In order to reduce this effort, conventional methods have been used to perform OCR recognition on one image using two or more types of OCR engines, obtain two or more recognition results, and based on the coincidence or mismatch of the two or more recognition results. Therefore, there has been an OCR system that narrows down the sheets to be checked by a person (for example, see Patent Document 1).

However, even when performing OCR recognition using two or more types of OCR engines, if the OCR accuracy of each OCR engine is insufficient, there will be many discrepancies in the recognition results, making it impossible to effectively reduce the effort of checking. Have difficulty. For this reason, the above OCR system corrects the read image in the first OCR engine, and recognizes the corrected image in each of the first and second OCR engines, thereby improving the accuracy of OCR in each OCR engine. I was trying to increase it.

Patent No. 5464474

However, even with the above-described correction, there is a limit to increasing the accuracy of OCR.

Therefore, there is a need for an information processing apparatus that can accurately acquire information written on paper by compensating for the lack of accuracy in OCR.

The information processing device of the first invention includes a statistical information storage unit that stores statistical information that is the result of statistical processing of information written on each of two or more sheets, and an image that is read from each of the two or more sheets. a paper image reception unit that receives a certain paper image; and a first OCR processing unit that performs first OCR processing on each of the two or more paper images and obtains two or more pieces of first OCR information that is information on a predetermined area. an OCR unit, a second OCR unit that performs second OCR processing on each of the two or more paper images and obtains two or more pieces of second OCR information that is information on a predetermined area; Using the above first OCR information and two or more second OCR information, a determining section determines correct OCR information that is OCR information corresponding to two or more sheets, and a determining section determines the correct OCR information, and The information processing apparatus includes an output section that outputs inspection information that is information regarding correct OCR information corresponding to each sheet.

With this configuration, the information written on the paper can be accurately acquired by compensating for the lack of accuracy in OCR.

Further, the information processing apparatus of the second invention is different from the first invention to an information processing apparatus in which the statistical information is the sum of numerical values written on two or more sheets.

With this configuration, the lack of precision in OCR can be easily and effectively compensated for.

Further, in the information processing device of the third invention, in contrast to the first or second invention, the paper image receiving unit receives two or more different paper images for each of the two or more papers, and The OCR unit performs first OCR processing on two or more different paper images for each of the two or more papers, and obtains two or more pieces of first OCR information that is information on a predetermined area. , the second OCR unit performs second OCR processing on two or more different paper images for each of the two or more papers, and generates two or more second OCR information that is information on a predetermined area. The determination unit determines correct OCR information that is OCR information corresponding to each of the two or more sheets using the statistical information, four or more first OCR information, and four or more second OCR information. It is an information processing device.

With this configuration, the lack of accuracy in OCR can be compensated for more effectively.

Further, in the information processing device of the fourth invention, in contrast to any one of the first to third inventions, an identifier is exposed on each of two or more sheets, and The section includes a special identifier storage section that stores one or more special identifiers that are identifiers that acquire identifiers from predetermined areas through OCR processing on each of the two or more paper images and perform predetermined special processing. , a determination unit that determines whether the identifier acquired by the first OCR unit or the second OCR unit is a special identifier; and a determination unit that determines whether the identifier acquired by the first OCR unit or the second OCR unit is a special identifier; The information processing apparatus further includes a special processing unit that performs special processing that is a predetermined process and is a process for increasing the accuracy of the determination process in which the determination unit determines correct OCR information.

With this configuration, by performing special processing only on sheets on which special identifiers are exposed, it is possible to compensate for the lack of accuracy in OCR while suppressing an increase in processing load.

According to the present invention, by compensating for the lack of precision in OCR, an information processing device is realized that can accurately acquire information written on paper.

Block diagram of information processing system in embodiment Flowchart explaining the operation of the information processing device Flowchart explaining the same decision process Flowchart explaining the operation of the terminal device Diagram showing an example of the same paper group Diagram showing an example of the same form This is a data structure diagram of the OCR-related information, and also shows an example of the OCR-related information. Data structure diagram of same judgment information This is a diagram showing the structure of the test results and also shows an example of the test results. Diagram showing an example of the same selection screen Diagram showing another example of the same OCR related information group Diagram showing another example of the same determination information group Diagram showing another example of the same test results A diagram showing another example of the same selection screen Diagram showing another example of the same OCR related information group Diagram showing another example of the same determination information group Data structure diagram of special identifier related information A diagram showing yet another example of the same OCR-related information group Diagram showing an example of the same confirmation screen External view of the computer system Diagram showing an example of the internal configuration of the computer system

Hereinafter, embodiments of an information system including an information processing device and the like will be described with reference to the drawings. Note that in the embodiments, constituent elements with the same reference numerals perform similar operations, and therefore, repeated explanation may be omitted.

FIG. 1 is a block diagram of an information processing system in this embodiment. This information processing system includes an information processing device 1 and one or more terminal devices 2. The information processing device 1 is communicably connected to each of the one or more terminal devices 2 via, for example, a network such as a LAN or the Internet, or a wireless or wired communication line.

The information processing device 1 is, for example, a server of a financial institution such as a bank. The server is, for example, a server within a financial institution, but may also be a cloud server, an ASP server, etc., and its location and type do not matter. The terminal device 2 is, for example, a PC in a financial institution, but it may also be a mobile terminal, and its location and type do not matter. A mobile terminal is a terminal that can be carried by a user, and includes, for example, a tablet terminal, a smartphone, a notebook PC, a mobile phone, etc., but the type thereof does not matter.

However, the information processing device 1 may be standalone. That is, the information processing method of the present invention can be implemented on, for example, a single PC, and does not depend on the hardware configuration.

In addition, in the financial institution of this embodiment, for example, in each store, regarding a bundle of papers such as bills received from a customer, the number of papers and the total amount written on each paper are written down. A form is created, and the bundle of papers is brought to the office center together with the form. At the office center, the number of sheets and the total amount of the bundle of sheets brought in are recalculated and compared with the number of sheets and the total amount written on the form.

Information systems are used to save labor in the above-mentioned clerical work at administrative centers. First, as a basic operation, the information processing device 1 reads two or more images of each paper via the terminal device 2, performs OCR on the read images, and calculates the amount written on the paper, etc. Obtain OCR information including information.

As a characteristic operation, the statistical processing results (for example, the total amount) for the information written on each of two or more sheets are stored in the recording medium, and the information processing device 1 stores The first OCR information obtained by the first OCR processing and the second OCR information obtained by the second OCR processing are acquired, and the stored statistical information, the two or more obtained first OCR information, and Correct OCR information is determined using the two or more pieces of second OCR information. As a result, even without increasing the accuracy of OCR, it is possible to effectively reduce the effort required by a person to compare the written information on the paper with the OCR information acquired from the paper image.

The information processing device 1 includes a storage section 11, a reception section 12, a processing section 13, and an output section 14. The storage unit 11 includes a statistical information storage unit 111 and a special identifier storage unit 112. The reception unit 12 includes a paper image reception unit 121. The processing section 13 includes a first OCR section 131 , a second OCR section 132 , a determining section 133 , a determining section 134 , and a special processing section 135 .

The terminal device 2 includes a terminal storage section 21 , a terminal reception section 22 , a terminal processing section 23 , a terminal transmission section 24 , a terminal reception section 25 , and a terminal output section 26 .

The storage unit 11 that constitutes the information processing device 1 can store various types of information. The various types of information include, for example, statistical information, which will be described later, and special identifiers, which will be described later. Other information will be explained in due course.

The statistical information storage unit 111 stores one or more pieces of statistical information. Statistical information is the result of statistical processing of information written on two or more sheets.

The paper may be, for example, a bill or a check, but its type does not matter. The bill is, for example, a promissory note or a bill of exchange, but the type does not matter. Further, the paper may be, for example, a slip. The slip may be, for example, a deposit slip, a withdrawal slip, a payout slip, etc., but the type thereof does not matter.

The information written on the paper is, for example, a numerical value. The numerical value is usually a monetary amount. In addition, information other than numerical values is usually written on the paper. Information other than numerical values includes, for example, the name of the sender, the name of the recipient, and type information. The name of the drawer is the name of the drawer who issued the paper such as a bill. The drawer is usually a company, but may also be an officer of the company. The beneficiary name is the name of the beneficiary who receives the amount written on paper such as a bill. The recipient is also typically a company, but may also be an individual. Type information is information indicating the type of paper. The type information includes, for example, "promissory note," "cheque," "slip," etc., but the format does not matter.

The statistical processing result is the result of statistical processing. Statistical processing is processing related to statistics. Statistical processing is usually a process of calculating the sum of numerical values written on two or more sheets. The sum of numerical values is the sum of two or more numerical values. The sum of numerical values is, for example, the total amount of two or more amounts, but may be the sum of numerical values other than amounts.

However, the statistical processing may be, for example, processing to obtain representative values of numerical values written on each of two or more sheets, and any type of processing may be used as long as it is related to statistics. A representative value is a value that represents a numerical value of 2 or more. The representative value is, for example, an average value, but may also be a maximum value, a median value, etc., and its type does not matter.

Statistical information is usually associated with a store identifier. A store identifier is information that identifies a store. A store is, for example, a head office or one or more branches of a financial institution, but the type thereof does not matter. The store identifier is, for example, a store number, a branch name, etc., but any information that can identify the store may be used. Further, the statistical information is usually associated with sheet count information. The number of sheets information is information indicating the number of sheets of paper. Note that the store identifier and the number of sheets information are normally stored in the statistical information storage section 111, but their location does not matter.

The statistical information is, for example, information automatically acquired by OCR processing, which will be described later, from a form in which statistical processing results such as the total amount are entered, but it may also be information input by the user based on the form.

Furthermore, an identifier is exposed on the surface of each of the two or more sheets. The identifier here is information used to determine whether or not the object is subject to special processing, which will be described later. The identifier is usually information that identifies the issuer. The information for identifying the issuer is, for example, the issuer's name described above. However, the information for identifying the payer may be, for example, an account number, a telephone number, etc., or any information that can identify the payer.

For example, the expression may be such that the identifier is printed on the paper in advance, the identifier may be added using a printer, or the identifier may be handwritten.

It is preferable that an identifier such as the payer's name be automatically acquired from each paper by OCR processing.

The special identifier storage unit 112 stores one or more special identifiers. The special identifier is information indicating that special processing is to be performed. The special identifier is usually the drawer's name. The special process is a predetermined special process that is performed only on the paper on which the special identifier is exposed. The special process is, for example, a process for increasing the accuracy of the decision process by the decision unit 133. Note that details of the special processing and determination processing will be described later.

The reception unit 12 receives various types of information. The various types of information include, for example, paper images to be described later. The reception unit 12 can also receive various instructions. Note that various instructions will be explained using specific examples.

The reception unit 12 normally receives information transmitted from the terminal device 2, but for example, it may receive information input from an input device such as a keyboard or touch panel, or receive information via an optical reading device such as a scanner or camera. It may also be possible to receive information read from a recording medium such as a disk or a semiconductor memory, or the like.

The paper image reception unit 121 receives paper images. The paper image receiving unit 121 normally receives a paper image transmitted from the terminal device 2, but for example, it may also accept a paper image read through an optical reading device, or it may receive a paper image read from a recording medium. It is also possible to accept a paper image that has been printed.

A paper image is an image obtained by reading two or more sheets. Specifically, for example, two or more sheets are each a bill, and the first sheet (hereinafter referred to as sheet 1) contains information such as the amount "¥1,000*" and the name of the drawer "BB Holdings." On the second sheet (Paper 2), information such as the amount "¥2,000*" and the name of the payer "AA Corporation" are written. In this case, the paper image receiving unit 121 first receives an image obtained by reading paper 1 with a scanner or the like (hereinafter referred to as "paper image 1"), and then receives an image obtained by reading paper 2 (hereinafter referred to as "paper image 2").

For example, the paper image receiving unit 121 may receive two or more different paper images for each of two or more papers.

Different paper images are usually images obtained by reading one paper in two or more different formats. The different formats include, for example, JPEG, TIF, GIF, BMP, etc., but the types are not limited.

Specifically, for example, paper 1 is read in two formats, JPEG and TIF, and the paper image reception unit 121 receives an image read in JPEG format (hereinafter referred to as paper image 1J) and an image in TIF format. Two types of images may be accepted: the image read in (hereinafter referred to as paper image 1T). Similarly, paper 2 is read in two formats, JPEG and TIF, and the paper image receiving unit 121 receives an image read in JPEG format (hereinafter referred to as paper image 2J) and an image read in TIF format. (hereinafter referred to as paper image 2T), two types of images may be accepted.

Note that the different paper images may be, for example, images obtained by reading one paper twice. Double reading refers to reading one sheet of paper two or more times in the same format. This is because even if the same paper is read in the same format, the information obtained from each paper image is not always the same (for example, fluctuations occur in the OCR results).

Specifically, for example, paper 1 is read twice in the same format (for example, JPEG format), and the paper image reception unit 121 receives the image read the first time (hereinafter, paper image 11). Two types of images may be accepted: and an image read a second time (hereinafter referred to as paper image 12). Similarly, paper 2 is read twice in the same format, and the paper image receiving unit 121 reads the image read the first time (hereinafter referred to as paper image 21) and the image read the second time (hereinafter referred to as paper image 21). Hereinafter, two types of images, the paper image 22), may be accepted.

Note that when one sheet image is received for each of m sheets, there are (2 to the power of m) combinations of matches and mismatches, which will be described later. Further, when n paper images are received for each of m sheets, there are {(2 to the power of m)×n} combinations of matches and mismatches.

The processing unit 13 performs various processes. The various types of processing include, for example, processing performed by the first OCR section 131, the second OCR section 132, the determining section 133, the determining section 134, and the special processing section 135. Furthermore, the processing unit 13 also performs various types of determination described in the flowchart, for example. Note that other processing will be explained in due course.

The first OCR unit 131 performs first OCR processing on each of the two or more paper images and obtains two or more pieces of first OCR information.

OCR processing involves processing a paper image by cutting out an image from a predetermined area and comparing the cut out image with a pre-prepared pattern. It can be said that this is a process of identifying an array (hereinafter referred to as a character string) and acquiring an array of one or more character codes (hereinafter referred to as a character code string) corresponding to the identified character string.

The predetermined area is, for example, an area where a numerical value (for example, an amount of money) is written, but it may also be an area where information other than numerical values (for example, the payer's name) is written, and the type thereof does not matter. Item names are usually associated with each type of area. The item name is information indicating the type of information written in the area. The item name is, for example, "amount" or "payer", but the type of information indicated by the item name does not matter.

Note that the predetermined area is specified, for example, by detecting the item name from the paper image, but may also be specified using position information. Position information is information regarding the position of an area within a paper image. The position information is, for example, a set of two coordinates indicating a starting point and an ending point, such as (x1, y1) and (x2, y2), but its format does not matter.

The first OCR processing is OCR processing using the first method. In contrast, the second OCR process to be described later is an OCR process based on the second method. The method can be said to be a program that implements OCR. Therefore, the difference between the first method and the second method is usually a difference in the program that implements OCR. However, the difference between the first method and the second method may be, for example, a difference in algorithm. Differences in algorithms include, for example, differences in algorithms for implementing processes such as image cropping and pattern matching, differences in cropping positions, and differences in patterns prepared in advance, but the type does not matter.

The first OCR information is information obtained from a predetermined area of the paper image by first OCR processing on each of two or more paper images. The first OCR information is, for example, a code string corresponding to a character string written in a predetermined area of a sheet of paper.

Note that the above-mentioned matters regarding the first OCR unit 131 also apply to the second OCR unit 132, which will be described later, except for the difference in the OCR processing method.

Specifically, the first OCR unit 131 performs the first OCR process on the paper image 1, for example, and generates the first OCR information "1000 yen, Company A, etc." including information such as the amount and the name of the payer.・Get “”. Similarly, the first OCR unit 131 performs the first OCR process on the paper image 2, and the first OCR information including information such as the amount and the name of the payer is "500 yen, B Corporation, . . ." get.

The first OCR unit 131 performs first OCR processing on each of the two or more different paper images received by the paper image receiving unit 121 for each of the two or more papers, and extracts the two or more pieces of first OCR information. You may obtain it.

Specifically, for example, for paper 1, the first OCR unit 131 performs first OCR processing on the paper image 1J read in JPEG format, and the first OCR section 131 performs first OCR processing on the paper image 1J read in JPEG format, and the first OCR section 131 performs first OCR processing on the paper image 1J read in JPEG format. ...", and also performs the first OCR processing on the paper image 1T read in TIF format, and obtains the first OCR information "1000 yen, Company A, ..." .

Similarly, regarding the paper 2, the first OCR unit 131 performs the first OCR processing on the paper image 2J read in JPEG format, and the first OCR information "500 yen, B Corporation, . . . ”, the first OCR process may also be performed on the paper image 2J read in TIF format, and the first OCR information “500 yen, B Corporation, . . .” may be obtained. .

Note that the two or more pieces of first OCR information obtained from each paper image obtained by reading one paper in two or more different formats do not necessarily match. For example, for paper 1, the first OCR information "1000 yen" is acquired from paper image 1J, while the first OCR information "1001 yen" is acquired from paper image 1T, and for paper 2, the first OCR information "1001 yen" is acquired from paper image 1T. There is also a case where the first OCR information "500 yen" is acquired and the first OCR information "500 yen" is also acquired from the paper image 2T.

Alternatively, regarding the paper 1, the first OCR unit 131 performs the first OCR processing on the paper image 11 read the first time, and outputs the first OCR information "1000 yen, A Corporation,..." At the same time, the first OCR processing may also be performed on the paper image 12 read a second time to obtain the first OCR information "1000 yen, Company A, . . .".

Similarly, regarding the paper 2, the first OCR unit 131 performs the first OCR processing on the paper image 21 read the first time, and the first OCR information "500 yen, B Corporation, . . . ”, the first OCR process may also be performed on the paper image 22 read the second time, and the first OCR information “500 yen, B Corporation, . . .” may be obtained. .

Note that the two or more pieces of first OCR information obtained from each paper image obtained by reading one paper twice or more do not necessarily match. For example, for paper 1, the first OCR information "1000 yen" is obtained from paper image 11, the first OCR information "1000 yen" is also obtained from paper image 12, and for paper 2, the first OCR information "1000 yen" is obtained from paper image 21. While the OCR information "505 yen" is acquired, there is also a case where the first OCR information "500 yen" is acquired from the paper image 22.

The second OCR unit 132 performs second OCR processing on each of the two or more paper images and obtains two or more pieces of second OCR information.

The second OCR processing is OCR processing using the second method. The second method is typically a different method than the first method. Note that the difference between the first method and the second method has been described above, so it will not be repeated.

The target of the second OCR process by the second OCR unit 132 is usually the same as the target of the first OCR process by the first OCR unit 131. That is, a first OCR process by the first OCR unit 131 and a second OCR process by the second OCR unit 132 are performed on each of the two or more paper images received by the paper image receiving unit 121. However, the two or more accepted paper images may include a paper image on which only either the first OCR process or the second OCR process is performed.

The second OCR information is information obtained from a predetermined area of the paper image by performing second OCR processing on each of two or more paper images.

Note that the predetermined area regarding the second OCR information is usually the same area as the predetermined area regarding the first OCR information. Basically, the same two code strings are obtained. That is, the first OCR information and the second OCR information obtained from the same area (for example, an area where a numerical value such as an amount of money is written: hereinafter referred to as the first area) within one sheet usually match.

However, even if the first OCR information and the second OCR information obtained from the first area within one sheet of paper do not necessarily match, depending on the accuracy of OCR. For example, from the first area of paper image 1, the first OCR information "1000 yen" and the second OCR information "1000 yen" are acquired, and from the first area of paper image 2, the first OCR information "500 yen" is obtained. There is also a possibility that the second OCR information "501 yen" is acquired.

The second OCR unit 132 may also perform second OCR processing on two or more different paper images for each of two or more papers, and obtain two or more pieces of second OCR information. . Note that in this process, the first OCR unit 131 performs first OCR processing on two or more different paper images for each of two or more papers, and obtains two or more first OCR information. This is the same as the processing, and the explanation will be omitted.

The determining unit 133 performs determination processing. The determination process refers to the statistical information stored in the statistical information storage unit 111, two or more pieces of first OCR information acquired by the first OCR unit 131, and two or more pieces of second OCR information acquired by the second OCR unit 132. This is OCR information corresponding to two or more sheets of paper, and is a process of determining correct OCR information using information.

Here, two or more first OCR information means that the first OCR unit 131 performs first OCR processing on each of the two or more paper images received by the paper image receiving unit 121 for each of the two or more papers. This is information obtained by performing. Similarly, two or more pieces of second OCR information means that the second OCR unit 132 performs a second This is information obtained by processing.

Correct OCR information is information that matches the information written on the paper, out of the first OCR information or the second OCR information. The correct OCR information is, for example, "first OCR information", "second OCR information", or "first OCR information and second OCR information", but it may also be an ID, etc., and its format does not matter. .

The determining unit 133 determines correct OCR information for each of the two or more sheets using the statistical information, two or more pieces of first OCR information, and two or more pieces of second OCR information, for example, in accordance with the following procedure.

That is, the determining unit 133 determines, for example, for each of M sheets (where M is an integer of 2 or more), the first OCR information and the second OCR information obtained from the sheet image read from the sheet are , determine whether they match or not. Further, the determination unit 133 also obtains the number of sheets determined to be non-matching among the M sheets (hereinafter, number of non-matching sheets m: m is an integer greater than or equal to 0), and determines that the number of non-matching sheets m is 1. It is further determined whether or not the above is true.

Then, if it is determined that the number m of mismatched sheets is 1 or more, the determination unit 133 determines whether the correct OCR information for each of the M sheets is "first OCR information", "second OCR information", or " Which of the first OCR information and the second OCR information is selected is determined as follows.

In other words, the first OCR information and the second OCR information are (1) determined to be "first OCR information and second OCR information" for sheets that match, and (2) for m sheets of paper that do not match, Among the possible combinations of matches and mismatches (i.e., 2 to the m power), the statistical information obtained based on the combination (e.g., total amount) is the same as the stored statistical information (e.g., total amount). The one corresponding to the matching combination is determined.

Note that if it is determined that the number m of mismatched sheets is 0, the "first OCR information and second OCR information" are determined for any of the M sheets.

Further, details of the determination process using statistical information, two or more first OCR information, and two or more second OCR information will be explained with a flowchart and a specific example.

Alternatively, the determining unit 133 may, for example, use the statistical information stored in the statistical information storage unit 111, four or more pieces of first OCR information acquired by the first OCR unit 131, and four or more pieces of first OCR information acquired by the second OCR unit 132. The second OCR information may be used to determine correct OCR information that corresponds to two or more sheets.

Here, the first OCR information of four or more means that the first OCR section 131 performs first OCR information on each of two or more different paper images received by the paper image reception section 121 for each of two or more sheets. This is information obtained by processing. Similarly, four or more pieces of second OCR information means that the second OCR unit 132 performs a second This is information obtained by performing OCR processing.

Note that the determination process using statistical information, 4 or more first OCR information, and 4 or more second OCR information uses statistical information, 2 or more first OCR information, and 2 or more second OCR information. The process is basically the same as the determination process (described above) except that (1) and (2) above are changed to, for example, (1a) and (1b) below.

In other words, assuming that n different paper images (where n is an integer of 2 or more) are accepted for each of M sheets, the first OCR unit 131 receives the n different paper images for each of the n different paper images. The first OCR process is performed to obtain n pieces of first OCR information, and the n pieces of first OCR information are stored in the storage unit 11, while the second OCR unit 132 processes each of the n pieces of different paper. Second OCR processing is performed on the image, n pieces of second OCR information are acquired, and the n pieces of second OCR information are stored in the storage unit 11 in association with the above n pieces of first OCR information. do. As a result, n pairs of first OCR information and second OCR information (hereinafter sometimes referred to as "pair information") are stored in the storage unit 11 for each of the M sheets.

Then, for each of the M sheets, the first OCR unit 131 determines that (1a) at least one pair of information in which the first OCR information and the second OCR information match exists among the n pieces of pair information. (2a) For sheets that do not exist, the possible combinations of matches and mismatches (i.e., {(2 to the power of m) x n }), the statistical information (e.g., total amount) acquired based on the combination is determined to correspond to the combination that matches the stored statistical information (e.g., total amount).

The determining unit 134 determines whether the identifier acquired by the first OCR unit 131 or the second OCR unit 132 is a special identifier. The determination unit 134 determines that the acquired identifier is a special identifier, for example, when it matches one or more special identifiers stored in the special identifier storage unit 112.

Note that although a match regarding information such as an amount is usually a perfect match, a match regarding an identifier does not necessarily have to be a perfect match. A match regarding an identifier may be, for example, that the number or proportion of matching characters among one or more characters constituting the identifier exceeds a threshold value or is equal to or greater than a threshold value.

The special processing unit 135 performs predetermined special processing on each of the one or more sheets on which the identifier determined by the determination unit 134 to be a special identifier appears. As mentioned above, the special process is a special process that is performed only on the paper on which the special identifier is exposed, and is, for example, a process to improve the accuracy of the decision process by the decision unit 133.

Specifically, the special process may be, for example, a process of accepting two or more different paper images for one paper, as described above. As described above, the different paper images are, for example, images obtained by reading one paper in different formats, but may also be images obtained by reading one paper twice in the same format. That is, the special processing unit 135 creates two or more different paper images, such as JPEG format and TIF format, only for each of the one or more papers on which the identifier determined by the determination unit 134 is a special identifier. It is preferable to carry out special processing for accepting other sheets of paper, and to perform normal acceptance processing for accepting only images in one format, such as JPEG, for example, without performing such special processing on other sheets.

Note that special processing can be used to convert two or more different paper images by, for example, converting the format of one received paper image into another format (for example, converting from JPEG format to TIF format or GIF format). It may also be a process of acquiring it.

Alternatively, the special process may be, for example, a process that increases the accuracy of at least one of the first or second OCR process, and any type of special process may be used as long as it increases the accuracy of the determination process by the determination unit 133. .

Processing that increases the accuracy of OCR processing includes, for example, processing that accepts paper images with higher resolution, processing that corrects accepted paper images, etc., but any processing that increases the accuracy of OCR processing as a result , the type does not matter.

The special processing unit 135 performs the above-mentioned special processing only on one or more sheets on which an identifier determined to be a special identifier by the determination unit 134 appears, so that the determination processing by the determination unit 133 is performed. The accuracy of can be increased while suppressing an increase in processing load.

Specifically, if the special identifier is, for example, the drawer name of a drawer who issues a large number of papers such as bills that are likely to be misrecognized (e.g., "AA Co., Ltd."), the drawer name that is the special identifier is By performing special processing on the exposed paper, the occurrence of erroneous recognition is suppressed. On the other hand, no special processing is performed on sheets with a payer's name that is not a special identifier (for example, "BB Holdings", etc.), so an increase in processing load is suppressed.

The output unit 14 outputs various information. The various types of information include, for example, test information that will be described later. The output by the output unit 14 is usually displayed on a display, but for example, it can be printed out on a printer, output sound from a speaker, handed over the processing result to another processing device or other program, or transferred to a recording medium. The information may be stored, transmitted to an external device, etc., and the format does not matter.

The inspection information is information regarding correct OCR information corresponding to two or more sheets determined by the determination unit 133. The inspection information may be, for example, a set of pairs of paper identifiers and correct OCR information (hereinafter referred to as pair information). The paper identifier is information that identifies paper. The paper identifier is, for example, a number such as "0001" or "0002", but may also be "first sheet" or "second sheet", or any information that can identify the sheet.

Note that the set information may also include, for example, information corresponding to correct OCR information, and its format does not matter. For example, when the correct OCR information is "k-th OCR information" (k is a natural number), the corresponding information is the amount of money, etc. obtained by the k-th OCR process corresponding to "k-th OCR information". It is a numerical value. Information corresponding to correct OCR information usually matches the information written on the paper.

Further, it is preferable that the output unit 14 outputs the paper identifier group of the m mismatched sheets, for example, when correct OCR information cannot be obtained for the m mismatched sheets mentioned above. This allows the user to efficiently check only one or more sheets corresponding to the output sheet identifier group.

The terminal storage unit 21 that constitutes the terminal device 2 can store various types of information. The various types of information are, for example, terminal identifiers. The terminal identifier is information that identifies the terminal device 2. The terminal identifier is, for example, an address such as a MAC address or an IP address, a telephone number, an ID, etc., but it may also be a user identifier that identifies the user of the terminal device 2, or any information that can identify the terminal device 2. . The user identifier may be, for example, an e-mail address, telephone number, address/name, ID, etc., but any information that can identify the user of the terminal device 2 may be used.

The terminal reception unit 22 receives various types of information. The various types of information include, for example, the above-mentioned paper image. For example, two or more images of each sheet are read through an optical reading device such as a scanner or a camera, and the terminal reception unit 22 receives the two or more images of each of the read sheets.

Note that the two or more read paper images are stored in, for example, the terminal storage unit 21, and the terminal reception unit 22 receives the two or more paper images (hereinafter referred to as a paper image group) stored in the terminal storage unit 21. may be accepted all at once.

The terminal processing unit 23 performs various processes. The various types of processing include, for example, various types of determination described in flowcharts. The terminal processing unit 23 also converts received information into transmitted information, for example.

The terminal transmitter 24 transmits various types of information. The various types of information include, for example, paper images. The terminal transmitting unit 24 transmits, for example, each of the two or more paper images received by the terminal receiving unit 22 to the information processing device 1 in pairs with the terminal identifier stored in the terminal storage unit 21.

The terminal receiving unit 25 receives various information. The various types of information include, for example, the above-mentioned test results, the above-mentioned sheet identifier group, and the like. The terminal receiving unit 25 receives test results or a group of paper identifiers from the information processing device 1, for example, in response to the terminal transmitting unit 24 transmitting two or more paper images or the like to the information processing device 1.

The terminal output unit 26 outputs various information. The various types of information include, for example, test results, paper identifier groups, and the like. The output by the terminal output unit 26 is usually displayed on a display, but it may also be stored in a recording medium, for example, and the format is not limited. The terminal output section 26 displays, for example, information such as the test results or paper identifier group received by the terminal reception section 25 on a display.

Based on the inspection results, the user can confirm that information has been accurately acquired from each of the two or more paper images. What has been done accurately is, for example, for all sheets of 2 or more, information such as the amount written on the sheet and the amount etc. obtained by OCR processing from the sheet image read from the sheet. The information must match.

Further, when a group of paper identifiers is output, the user can efficiently check only one or more papers for which there is a high possibility that information was not acquired accurately. What was not done accurately is, for example, OCR processing of at least one of two or more sheets of paper from the information such as the amount written on the sheet and the paper image read from the sheet. This is because the information such as the amount obtained by the above information does not match.

The storage unit 11, the statistical information storage unit 111, the special identifier storage unit 112, and the terminal storage unit 21 are preferably nonvolatile recording media such as a hard disk or flash memory, but they can also be realized by volatile recording media such as RAM. It is possible.

The process by which information is stored in the storage unit 11 or the like does not matter. For example, information may be stored in the storage unit 11 or the like via a recording medium, or information transmitted via a network or communication line may be stored in the storage unit 11 or the like. Alternatively, information input via an input device may be stored in the storage unit 11 or the like. The input device may be anything, such as a keyboard, mouse, touch panel, camera, etc.

The reception unit 12, paper image reception unit 121, and terminal reception unit 22 may or may not include input devices. The reception unit 12 and the like can be realized by the driver software of the input device, or by the input device and its driver software.

The processing section 13, the first OCR section 131, the second OCR section 132, the determination section 133, the judgment section 134, the special processing section 135, and the terminal processing section 23 can usually be realized by an MPU, a memory, or the like. The processing procedures of the processing unit 13 and the like are usually realized by software, and the software is recorded on a recording medium such as a ROM. However, the processing procedure may be realized by hardware (dedicated circuit).

The output unit 14 and the terminal output unit 26 may or may not include output devices such as a display and a speaker. The output unit 14 and the like can be realized by the driver software of the output device, or by the output device and its driver software.

The transmission functions of the terminal transmitting unit 24 and the output unit 14 are usually realized by wireless or wired communication means (for example, a communication module such as a NIC (Network Interface Controller) or a modem), but are not realized by broadcasting means (for example, broadcasting). module).

The receiving functions of the terminal receiving section 25, receiving section 12, etc. are usually realized by wireless or wired communication means, but may also be realized by means for receiving broadcasts (for example, a broadcast receiving module).

Next, the operation of the information processing system will be explained using flowcharts shown in FIGS. 2 to 4.

FIG. 2 is a flowchart illustrating the operation of the information processing device 1. Note that in this flowchart, the reception by the paper image reception unit 121 is usually reception from the terminal device 2, and the output by the output unit 14 is usually transmission to the terminal device 2. It is assumed that the paper image receiving unit 121 receives a group of paper images paired with a terminal identifier from the terminal device 2 .

(Step S201) The processing unit 13 determines whether the paper image receiving unit 121 has received a group of paper images or the like from the terminal device 2. If it is determined that the paper image receiving unit 121 has received the paper image group etc. from the terminal device 2, the process advances to step S202, and if it is determined that the paper image group has not been received, the process returns to step S201.

(Step S202) The processing unit 13 counts the number of paper images forming the paper image group received in step S201, and sets the counting result in variable M. The variable M is a variable indicating the number of sheets of paper that has been read.

(Step S203) The processing unit 13 determines whether the variable M matches the number of sheets information associated with the statistical information stored in the statistical information storage unit 111. If it is determined that the variable M matches the sheet number information, the process advances to step S205, and if it is determined that they do not match, the process advances to step S204.

(Step S204) The output unit 14 outputs the "number of sheets mismatch" to the terminal device 2 (hereinafter sometimes simply referred to as "terminal device 2") identified by the terminal identifier received in pair with the paper image group in step S201. ). After that, the process returns to step S201.

(Step S205) The first OCR unit 131 acquires a first OCR information group having M pieces of first OCR information by performing the first OCR processing on the paper image group received in step S201, and stores it in the storage unit 11. do.

(Step S206) The second OCR unit 132 acquires a second OCR information group having M pieces of second OCR information by performing a second OCR process on the paper image group received in step S201, and obtains a second OCR information group having M pieces of second OCR information. The first OCR information group is stored in the storage unit 11 in association with the first OCR information group. As a result, the storage unit 11 stores M pairs of first OCR information and second OCR information.

(Step S207) The processing unit 13 sets the initial value "1" to the variable i. The variable i is a variable for sequentially selecting unselected pairs among the M pairs stored in the storage unit 11. Note that the variable i can be said to be a variable for sequentially selecting unselected sheets among the M sheets.

(Step S208) The processing unit 13 determines whether there is an i-th pair. If it is determined that there is an i-th pair, the process advances to step S209, and if it is determined that there is no i-th pair, the process advances to step S213.

(Step S209) The processing unit 13 determines whether the first OCR information and the second OCR information forming the i-th pair match. If it is determined that the first OCR information and the second OCR information constituting the i-th pair match, the process proceeds to step S210, and if it is determined that they do not match, the process proceeds to step S211.

(Step S210) The processing unit 13 sets "match" to the i-th determination result, and proceeds to step S212.

(Step S211) The processing unit 13 sets "mismatch" to the i-th determination result, and proceeds to step S212.

(Step S212) The processing unit 13 increments the variable i, and returns to step S208.

(Step S213) The determining unit 133 selects the statistical information (for example, total amount) stored in the statistical information storage unit 111, the first OCR information group acquired in step S205, and the second OCR information group acquired in step S206. A determination process for determining the correct OCR information group is executed using the OCR information group. Note that the determination process will be explained using the flowchart of FIG. 3.

(Step S214) The processing unit 13 determines whether the correct OCR information group is "Null". Note that "Null" is a flag indicating the absence of the information. If it is determined that the correct OCR information group is "Null", the process advances to step S215; if it is determined that the correct OCR information group is not "Null", the process advances to step S216.

(Step S215) The output unit 14 transmits to the terminal device 2 a group of paper identifiers corresponding to one or more pairs (hereinafter referred to as pair groups) for which the determination result in step S209 is "inconsistent". after that. Return to step S201.

(Step S216) The output unit 14 transmits the inspection results regarding the correct OCR information group to the terminal device 2. after that. Return to step S201.

Note that in the flowchart of FIG. 2, the process starts when the information processing device 1 is powered on or a program is started, and ends when the power is turned off or an interrupt to end the process occurs. However, the trigger for starting or ending the process does not matter.

FIG. 3 is a flowchart illustrating the determination process in step S213.

(Step S301) The determining unit 133 counts the pairs for which the determination result in step S209 is "unmatched", and sets the count result to the variable m. The variable m is a variable that indicates the number of pairs in which the first OCR information and the second OCR information do not match among the M pairs. Note that the variable m can be said to be a variable that indicates the number of sheets of paper for which the first OCR information and the second OCR information do not match among the M sheets.

(Step S302) The determining unit 133 determines whether the variable m is 1 or more. If it is determined that the variable m is greater than or equal to 1, the process proceeds to step S304; if it is determined that the variable m is not greater than or equal to 1 (that is, it is 0), the process proceeds to step S303.

(Step S303) The determining unit 133 determines the first OCR information group and the second OCR information group as correct OCR information groups. After that, the process returns to the upper level processing.

(Step S304) The determining unit 133 creates 2 to the m power of the m pairs for which the determination result in step S209 is "inconsistent."

(Step S305) The determining unit 133 sets the initial value "1" to the variable j. The variable j is a variable for sequentially selecting unselected combinations from among the "2 m combinations" created in step S304.

(Step S306) The determining unit 133 determines whether there is a j-th combination. If it is determined that there is a j-th combination, the process advances to step S307; if it is determined that there is no combination, the process advances to step S312.

(Step S307) The determining unit 133 acquires statistical information (for example, the j-th total amount) corresponding to the j-th combination.

(Step S308) The determining unit 133 determines whether the statistical information acquired in step S307 matches the statistical information (for example, the total amount) stored in the statistical information storage unit 111. If it is determined that the acquired statistical information matches the stored statistical information, the process advances to step S309, and if it is determined that they do not match, the process advances to step S310.

(Step S309) The determining unit 133 sets "match" to the j-th determination result, and proceeds to step S311.

(Step S310) The determining unit 133 sets "mismatch" to the j-th determination result, and proceeds to step S311.

(Step S311) The determining unit 133 increments the variable j, and returns to step S306.

(Step S312) The determining unit 133 determines whether there is a combination for which the determination result in step S308 is a "match". If it is determined that there is a combination whose determination result in step S308 is "match", the process advances to step S313, and if it is determined that there is no combination, the process advances to step S315.

(Step S313) The determining unit 133 determines whether there is only one combination for which the determination result in step S308 is "match". If it is determined that there is only one combination in which the determination result in step S308 is "matching", the process advances to step S314; if it is determined that there is not only one combination (that is, there are two or more), then step Proceed to S315.

(Step S314) The determining unit 133 sets the only combination for which the determination result in step S308 is "match" in the correct OCR information group. After that, the process returns to the upper level processing.

(Step S315) The determining unit 133 sets "Null" to the correct OCR information group. After that, the process returns to the upper level processing.

FIG. 4 is a flowchart illustrating the operation of the terminal device 2. In this flowchart, it is assumed that the terminal reception section 22 receives the paper image group stored in the terminal storage section 21 all at once.

(Step S401) The terminal processing unit 23 determines whether the terminal reception unit 22 has accepted the paper image group. If it is determined that the terminal receiving unit 22 has accepted the paper image group, the process advances to step S402, and if it is determined that the group of paper images has not been received, the process advances to step S403.

(Step S402) The terminal transmitting unit 24 transmits the paper image group received in step S401 to the information processing device 1 in pairs with the terminal identifier stored in the terminal storage unit 21. After that, the process returns to step S401.

(Step S403) The terminal processing unit 23 determines whether the terminal receiving unit 25 has received “number of sheets mismatch”. If it is determined that the terminal receiving unit 25 has received "number of sheets mismatch", the process advances to step S404, and if it is determined that the message has not been received, the process advances to step S405.

(Step S404) The terminal output unit 26 outputs the "number of sheets mismatch" received in step S403 via an output device such as a display. After that, the process returns to step S401.

(Step S405) The terminal processing unit 23 determines whether the terminal receiving unit 25 has received the test result. If it is determined that the terminal receiving unit 25 has received the test results, the process advances to step S406, and if it is determined that the test results have not been received, the process advances to step S407.

(Step S406) The terminal output unit 26 outputs the test result received in step S405 via an output device such as a display. After that, the process returns to step S401.

(Step S407) The terminal processing unit 23 determines whether the terminal receiving unit 25 has received the paper identifier group. If it is determined that the terminal receiving unit 25 has received the sheet identifier group, the process advances to step S408, and if it is determined that the paper identifier group has not been received, the process returns to step S401.

(Step S408) The terminal output unit 26 outputs the paper identifier group received in step S407 via an output device such as a display. After that, the process returns to step S401.

Note that in the flowchart of FIG. 4, the process starts when the terminal device 2 is powered on or a program is started, and ends when the power is turned off or an interrupt to end the process occurs. However, the trigger for starting or ending the process does not matter.

Note that when n paper images are prepared for each of M sheets and OCR processing is performed on each of the n paper images, some steps (S202, S205 to S205) in the flowchart of FIG. S209, S213) are changed as follows.

The paper image receiving unit 121 receives the paper image group and the variable n in pairs with the terminal identifier from the terminal device 2 . The variable n is a variable indicating the number of paper images prepared for one paper.

(Step S202a) The processing unit 13 counts the number of paper images constituting the paper image group received in step S201, divides the number of paper images by the variable n, and divides the resultant “number of paper images/n” into the variable M Set to .

(Step S205a) The first OCR unit 131 acquires a first OCR information group having (M×n) pieces of first OCR information by performing the first OCR processing on the paper image group received in step S201, and stores it. It is stored in section 11.

(Step S206a) The second OCR unit 132 acquires a second OCR information group having (M×n) pieces of second OCR information by performing a second OCR process on the sheet image group received in step S201, and It is stored in the storage unit 11 in association with the first OCR information group acquired in S205a. As a result, n pairs (pair information) of the first OCR information and the second OCR information are stored in the storage unit 11 for each of the M sheets.

(Step S207a) The processing unit 13 sets the initial value "1" to the variable i. The variable i here is a variable for sequentially selecting unselected sheets among the M sheets.

(Step S208a) The processing unit 13 determines whether or not there is an i-th sheet. If it is determined that there is an i-th sheet, the process advances to step S209a, and if it is determined that there is no i-th sheet, the process advances to step S213a.

(Step S209a) The processing unit 13 determines whether there is at least one pair of information in which the first OCR information and the second OCR information match among the n pieces of pair information corresponding to the i-th sheet. . If it is determined that there is at least one pair of information in which the first OCR information and the second OCR information match among the n pieces of pair information corresponding to the i-th sheet, the process advances to step S210, and there is no pair of information. If it is determined that this is the case, the process advances to step S211.

(Step S213a) The determining unit 133 selects the statistical information (for example, the total amount) stored in the statistical information storage unit 111, the first OCR information group acquired in step S205a, and the second OCR information group acquired in step S206a. A determination process for determining the correct OCR information group is executed using the OCR information group.

Also, in the flowchart of FIG. 3, some steps (S301, S304) are changed as follows.

(Step S301a) The determining unit 133 counts the number of sheets for which the determination result in step S209a is "unmatched", and sets the count result to a variable m. The variable m here is a variable indicating the number of sheets for which the first OCR information and the second OCR information do not match among the M sheets.

(Step S304a) The determining unit 133 creates {(2 to the power of m)×n} combinations for each of the m sheets for which the determination result in step S209a is “unmatched”.

Furthermore, in the flowchart of FIG. 4 as well, some steps (S401, S402) are changed as follows.

(Step S401a) The terminal processing unit 23 determines whether the terminal reception unit 22 has received the paper image group and the variable n. If it is determined that the terminal reception unit 22 has received the paper image group and the variable n, the process advances to step S402a, and if it is determined that the terminal reception unit 22 has not received the paper image group and the variable n, the process advances to step S403.

(Step S402a) The terminal transmitting unit 24 transmits the paper image group and variable n received in step S401 to the information processing device 1 in pairs with the terminal identifier stored in the terminal storage unit 21. After that, the process returns to step S401a.

A specific example of the operation of the information processing system in this embodiment will be described below. Note that the following description can be modified in various ways and is not intended to limit the scope of the present invention in any way.

(Specific example 1)
The paper in this example is a bill. Suppose now that three sheets of paper as shown in FIG. 5 are brought into the office center from each store along with a form as shown in FIG. 6.

FIG. 5 is a diagram showing an example of a paper group. A sheet group is a collection of two or more sheets. This paper group has three sheets. On the first sheet of paper, the amount "¥1,000*" and the drawer "BB Holdings" are written. Furthermore, on the second sheet of paper, the amount "¥2,000*" and the drawer "AA Co., Ltd." are written. Furthermore, on the third sheet, the amount "¥3,000*" and the drawer "AA Co., Ltd." are written.

FIG. 6 is a diagram showing an example of a form. In this form, statistical information such as the total amount "6,000 yen" and the number of sheets "3" are written.

When the images of the three sheets mentioned above are read by a scanner, and information such as the total amount "6,000 yen" written on the form is inputted via an input device such as a keyboard, the terminal device 2 displays the 4, steps S401 and S402 are executed. That is, the terminal reception unit 22 receives information such as a paper image group including the three read paper images and the read total amount “6000 yen”, and the terminal transmission unit 24 receives the received paper images. Information such as the group and total amount is transmitted to the information processing device 1.

In the information processing device 1, the paper image reception unit 121 receives the paper image group and information such as the total amount, and the processing unit 13 etc. accumulates the received information such as the total amount in the statistical information storage unit 111. At the same time, the processes of steps S201 to S216 in FIG. 3 are executed using the received sheet image group and information such as the total amount stored in the statistical information storage unit 111. As a result, two or more (in this example, three) pieces of OCR-related information, as shown in FIG. 7, are accumulated in the storage unit 11, for example.

FIG. 7 is a data structure diagram of OCR related information. The OCR-related information is information regarding the results of OCR performed on a group of paper images. The OCR-related information includes first OCR information, second OCR information, a determination result, an identifier (in this example, the name of the issuer), and the like.

Further, two or more pieces of OCR related information stored in the storage unit 11 are associated with mismatch information. The mismatch information is information regarding one or more sheets for which the first OCR information and the second OCR information do not match. The mismatch information includes one or more paper identifiers. Specifically, the mismatch information is, for example, "first page", "second page, third page", etc., but its format does not matter.

Further, each of the two or more pieces of OCR related information stored is associated with an ID (for example, "1", "2", etc.). Note that, hereinafter, two or more pieces of OCR-related information stored may be referred to as an "OCR-related information group."

Note that FIG. 7 also shows an example of an OCR-related information group. In this example, there is only one sheet of paper for which the determination results do not match. For example, OCR related information associated with ID "1" (hereinafter referred to as OCR related information 1) includes the first OCR information "1000 yen", the second OCR information "1000 yen", the determination result "match", and the identifier " BB Holdings, etc. In addition, the OCR related information (OCR related information 2) associated with ID "2" includes the first OCR information "2000 yen", the second OCR information "2100 yen", the determination result "mismatch", and the identifier "AA stock". Company” etc. Similarly, the OCR related information 3 includes first OCR information "3000 yen", second OCR information "3000 yen", determination result "match", and identifier "AA Corporation". Then, the OCR related information 1 to 3 are associated with the mismatch information "second sheet".

In the information processing device 1, the determining unit 133 executes the determining process in step S213 in FIG. 3 (specifically, steps S301 to S315 in FIG. 3) using the OCR-related information group in FIG. 7, so that, for example, One or more pieces of decision information as shown in FIG. 8 are stored in the storage unit 11.

FIG. 8 is a data structure diagram of determination information. The determination information is information regarding the progress and results of the determination process. The determination information includes mismatch information, combination, total amount determination result, and correct OCR information group. The combinations are all possible combinations of matches and mismatches regarding the paper group identified by one or more paper identifiers included in the mismatch information. As mentioned above, the number of combinations is {2 m The result is as follows. The determination result is the result of the determination in step S308.

Note that each of the one or more pieces of stored determination information is associated with an ID (for example, "1", "2", etc.). Hereinafter, one or more pieces of stored determination information may be referred to as a "determination information group", and individual determination information may be referred to as "determination information 1", "determination information 2", etc.

Note that FIG. 8 also shows an example of the determination information group. This example is a case of {mismatch information=2nd sheet, m=1}. The determination information group has only one determination information 1.

Determination information 1 includes mismatch information "second sheet" and two combinations. The first combination is that the first OCR information is correct and the first OCR information is incorrect (hereinafter sometimes written as {○, ×}), and the second combination is {×,○} It is.

In addition, determination information 1 includes the total amount “6000 yen” corresponding to the first combination {○, ×} and the determination result “match”, and the total amount “6100 yen” corresponding to the second combination {×, ○} ” and the judgment result “inconsistency” are also included. Furthermore, the determination information 1 also includes a correct OCR information group "second sheet: first OCR information".

Note that the total amount “6,000 yen” corresponding to the first combination {◯, ×} is obtained by, for example, the determining unit 133 as follows using the OCR information group shown in FIG. That is, the determining unit 133 selects the first or second OCR information “1000 yen” that OCR information 1 has, the first OCR information “2000 yen” that OCR information 1 has, and the first or second OCR information that OCR information 3 has. The second OCR information "3000 yen" is added to obtain the addition result "6000 yen".

Similarly, the total amount “6100 yen” corresponding to the second combination {x, ○} is calculated by adding “1000” yen of the first or second OCR information that OCR information 1 has and the first OCR information that OCR information 2 has. It is obtained by adding the information "2100 yen" and the first or second OCR information "3000" yen included in the OCR information 3.

Further, the determination result “match” corresponding to the first combination {○, ×} is the total amount “6000 yen” corresponding to the first combination {○, ×} and the total amount “ 6,000 yen” match or not. Similarly, the determination result “match” corresponding to the second combination {○, ×} is the total amount “6100 yen” corresponding to the second combination {×, ×} and the total amount in the statistical information storage unit 111. It is obtained as a result of determining whether or not "6000 yen" matches.

Furthermore, the correct OCR information group “2nd sheet, match” is the “2nd sheet” of mismatch information and the “2nd sheet” corresponding to “○” of the first combination {○, ×} whose judgment result is “match”. "1 OCR information".

Further, in the information processing device 1, the processing unit 13 and the output unit 14 execute the processing in steps S214 to S215 in FIG. 3 using the determination result in FIG. By executing the processes of S405 to S408, one or more test results as shown in FIG. 9 are displayed on the display of the terminal device 2, for example.

FIG. 9 is a structural diagram of the test results that are output. The inspection result includes three items: "mismatch", "first OCR", and "second OCR". "Mismatch" is an item corresponding to mismatch information included in the judgment information, and "First OCR" and "Second OCR" correspond to the first OCR information and second OCR information that constitute the combination included in the judgment information. This is an item to do.

“Unmatched” indicates one or more page identifiers included in the unmatched information (see FIG. 8, etc.). "First OCR" and "Second OCR" include two recognition results obtained by the first and second OCR processes (for example, {1000 yen, 1000 yen }, {2000 yen, 2100 yen}, etc.), and two determination results (either {◯, x} or {x, ○}) corresponding to the two recognition results are written.

Note that FIG. 9 also shows an example of the test results displayed on the display. In this inspection result, "2nd sheet" is displayed in the "mismatch" item, and {2000 yen, 2100 yen} and {○, ×} are displayed in the "1st OCR" and "2nd OCR" items. Displayed.

As a result, the user notices that there is a discrepancy between the two recognition results obtained by the first OCR and the second OCR on the second sheet of the three sheets, and the recognition results obtained by the first OCR and the second OCR are , 2000 yen and 2100 yen, respectively, and it can be seen that the former was determined to be correct and the latter was determined to be incorrect.

Note that an OK button and the like are also displayed on the display, and the above determination result is confirmed in response to pressing the OK button.

In addition, in the determination information of FIG. 8, if both of the two determination results are "inconsistent," for example, the selection screen shown in FIG. 10 may be displayed. FIG. 10 is a diagram showing an example of the selection screen. The selection screen is a screen for presenting the two recognition results by the first OCR and the second OCR and accepting a selection by the user when the determination results are "inconsistent". On the selection screen, for example, along with a character string such as "It cannot be determined which is correct. Please select the correct one," the judgment result {○, The changed information (for example, {?, ?}, etc.) is displayed. However, the structure and expression of the selection screen does not matter.

This allows the user to understand that there is a discrepancy between the two recognition results obtained by the first OCR and the second OCR for the second sheet of paper, and that it was not possible to determine whether it is correct or incorrect. Therefore, the user checks the second sheet of paper, and since "¥2,000*" is written thereon, the user selects the first OCR corresponding to the recognition result "¥2,000". This indicates that it is impossible to determine {? ，? } is replaced with the determination result {◯, ×} based on the user's selection, and the amount of money for the second sheet is determined to be “2000 yen”.

Note that when the number of sheets in which mismatches occur is two (that is, m = 2), the OCR-related information group, determination information group, inspection results, and selection screen are as shown in FIGS. 11 to 14, for example. . In this example, there is a mismatch between the second and third images.

FIG. 11 is a diagram showing another example of the OCR-related information group (in the case of m=2). Of the OCR related information 1 to 3 that make up this OCR related information group, OCR related information 1 and 2 are the same as the example in FIG. 7, but in OCR related information 3, the first OCR information is "3001 yen". , and the determination result is “no match”. Therefore, the mismatch information is "2nd and 3rd sheets".

FIG. 12 is a diagram showing another example of the determination information group (m=2, mismatch information=2nd, 3rd sheet). This judgment information group has four pieces of judgment information 1 to 4 regarding four possible combinations (that is, the square of 2) of the second and third sheets. Among these, in judgment information 2, the total amount is "6000 yen" and the judgment result is "match", so the correct OCR information group is "2nd sheet: first OCR information" and "3rd sheet: "Second OCR Information".

FIG. 13 is a diagram showing another example of the inspection results (m=2, mismatch information=2nd, 3rd sheet). This inspection result is obtained by adding information regarding the third sheet to the inspection result of FIG. 9. Note that the information regarding the second sheet is the same as that in FIG. In other words, in the inspection results of FIG. 13, "3rd sheet" is added to the "mismatch" item, and the "3rd sheet" corresponding to {3001 yen] is added to the "1st OCR" and "2nd OCR" items. , 3,000 yen} and {x, ○} are added.

As a result, the user notices that even for the third sheet, there is a discrepancy between the two recognition results obtained by the first OCR and the second OCR, and for the third sheet, the recognition results obtained by the first OCR and the second OCR are , 3001 yen and 3000 yen, respectively, indicating that the former was determined to be incorrect and the latter was determined to be correct. Note that the determination result regarding the third sheet is also confirmed together with the determination result regarding the third sheet in response to pressing the OK button.

FIG. 14 is a diagram showing another example of the selection screen (m=2, mismatch information=2nd, 3rd image). This selection screen is obtained by adding information regarding the third sheet to the selection screen of FIG. 10. Note that the information regarding the second sheet is the same as that in FIG. 10. That is, on the selection screen in FIG. 14, "3rd sheet" is added to the "unmatched" item, and the "3rd sheet" corresponding to {3001 yen , 300 yen} and {? ，? } has been added.

This allows the user to understand that even for the third sheet, there is a discrepancy between the two recognition results obtained by the first OCR and the second OCR, and it was not possible to determine whether the sheet is correct or incorrect. Therefore, the user checks the third sheet of paper, and since "¥3,000*" is written thereon, the user selects the second OCR corresponding to the recognition result "¥3,000". As a result, the third image has been added {? ，? } and the like are replaced with the determination result {×, ○} based on the user's selection, and the amount of money for the third sheet is determined to be “3000 yen”.

Note that when m=3, the OCR-related information group and the determination information group are as shown in FIGS. 15 and 16, for example. In this example, mismatches occur in all of the first to third sheets. Note that illustration of the test results and selection screen is omitted.

FIG. 15 is a diagram showing another example of the OCR-related information group (in the case where m=3). Of the OCR related information 1 to 3 that make up this OCR related information group, OCR related information 2 and 3 are the same as the example in FIG. 11, but in OCR related information 1, the first OCR information is "1050 yen". , and the determination result is “no match”. Therefore, the mismatch information is "1st to 3rd sheets".

FIG. 16 is a diagram showing another example of the determination information group (m=3, mismatch information=first to third sheets). This judgment information group has eight pieces of judgment information 1 to 8 regarding eight possible combinations (that is, 2 to the third power) of the first to third sheets. Among these, in judgment information 6, the total amount is "6000 yen" and the judgment result is "match", so the correct OCR information group is "1st sheet: 2nd OCR information", 2nd sheet: 1st sheet "OCR information" and "Third sheet: second OCR information."

Therefore, the inspection result in this case is the one in which information regarding the first sheet is further added to the inspection result in FIG. That is, in the inspection results in this case, "1st sheet" is added to the "mismatch" item, and the "1st sheet" corresponding to "1st sheet" is added to the "1st OCR" and "2nd OCR" items. , 1000 yen} and {x, ○} are added. Note that the user's instructions regarding this test result and the processing performed by the processing unit 13 and the like in response to the instructions are the same as described above.

Further, the selection screen in this case is the selection screen shown in FIG. 14 with information regarding the first sheet added. In other words, on the selection screen in this case, "1st sheet" is added to the "unmatched" item, and the "1st sheet" corresponding to "1st sheet" is added to the "1st OCR" and "2nd OCR" items. , 1000 yen} and {? ，? } is added. Note that the user's instructions on this selection screen and the processing performed by the processing unit 13 and the like in response to the instructions are the same as described above.

(Specific example 2)
In addition, in the above specific example 1, two images (for example, two types of JPEG and TIF images) are prepared for each of the three sheets, and a sheet image group having a total of six sheet images is used. , a determination process may be performed. A detailed explanation of the second specific example will be omitted.

(Specific example 3)
Alternatively, in the first specific example, the special identifier storage unit 112 may store one or more special identifiers (for example, the name of a drawer who issues a large number of papers such as bills that are likely to be misrecognized). The first OCR unit 131 or the second OCR unit 132 also acquires the issuer's name through OCR processing on each paper image, and the determining unit 134 determines that the issuer's name obtained from each of the three paper images is stored in a special identifier storage. It may be determined whether the name of the drawer matches the name of the drawer stored in 112. If the obtained issuer's name is similar enough to be determined to match the stored issuer's name, special processing is performed to increase recognition accuracy (for example, correction to the paper image, (e.g., preparing two or more paper images) may also be performed.

(Specific example 4)
In the above specific example 3, the special process is a process of preparing two or more paper images for one paper, and the special identifier storage unit 112 stores information about the drawer whose number of misreadings is determined by the processing unit 13 to be greater than the threshold value. may be stored. The number of misreadings refers to the number of times or the ratio of the number of times the determination unit 133 determines that the recognition result is incorrect regarding two or more sheets corresponding to one payer's name.

The ratio is the ratio of the number of times that the determination unit 133 determines that there is an error to the number of times that the first OCR unit 131 or the second OCR unit 132 recognizes the error. %) etc., but the format does not matter.

The special identifier storage unit 112 may store one or more pieces of special identifier related information as shown in FIG. 17, for example. FIG. 17 is a data structure diagram of special identifier related information. Special identifier related information is information related to a special identifier. The special identifier related information includes an identifier (in this example, the drawer's name), the number of misreads, and a special identifier flag. The number of misreadings here is the number of times it was determined to be an error.

The special identifier flag is information indicating whether the identifier is a special identifier. The special identifier flag switches between, for example, "ON" indicating that the identifier is a special identifier, and "OFF" indicating that the identifier is not a separate identifier. Identifiers associated with the special identifier flag "ON" are considered to be special identifiers, and identifiers associated with the special identifier flag "OFF" are considered not to be special identifiers.

The special identifier related information 1 associated with the ID "1" has the identifier "AA Corporation", the number of misreadings "3", and the special identifier flag "ON". Similarly, the special identifier related information 2 has the identifier "BB Holdings", the number of misreadings "1", and the special identifier flag "OFF".

Specifically, the initial values of the number of misreads and the special identifier flag stored in the special identifier storage unit 112 are "0" and "OFF", respectively. The first OCR unit 131 or the second OCR unit 132 acquires information such as the amount from the paper image in which any of the identifiers stored in the special identifier storage unit 112 is displayed, and determines the acquired information. Each time the unit 133 determines an "error", the processing unit 13 increments the number of misreadings corresponding to the identifier. Then, the processing unit 13 determines whether or not the number of misreadings after the increment has exceeded a threshold (here, "2"), and when determining that it has exceeded the threshold, sets the special identifier flag corresponding to the identifier. Switch from “OFF” to “ON”.

Assuming that the special identifier storage unit 112 stores two pieces of special identifier related information 1 and 2 shown in FIG. Only for this case, two paper images are prepared for each sheet, and only one paper image is prepared for the sheet on which "BB Holdings" whose special identifier flag is "OFF" is displayed.

In this case, the OCR-related information 2 included in the OCR information group of FIG. 7 described above becomes, for example, the OCR-related information 2 shown in FIG. 18. FIG. 18 is a diagram showing still another example of the OCR-related information group. The OCR-related information 2 included in the OCR-related information group shown in FIG. Contains two pieces of OCR related information.

Among the above two OCR related information, the first OCR related information includes the first OCR information "2000 yen", the second OCR information "2100 yen", and the determination result "mismatch", and the second OCR related information includes the first OCR information "2000 yen", the second OCR information "2000 yen", and the determination result "match".

Therefore, in OCR related information 2, at least one determination result is "match", so the mismatch information associated with the OCR information group in FIG. 18 is "Null". Thereby, the accuracy of OCR can be improved while suppressing an increase in processing load.

Note that if the mismatch information is "Null", a confirmation screen as shown in FIG. 19 may be displayed on the display of the terminal device 2, for example. FIG. 19 is a diagram showing an example of a confirmation screen. The confirmation screen is a screen for confirming that the two recognition results obtained by the first and second OCR processes match for all paper images constituting the paper image group. The confirmation screen includes a character string such as "Two OCRs matched for all sheets..." and an OK button. When the OK button is pressed, all recognition results for the paper image group are confirmed.

As described above, according to the present embodiment, the statistical information storage unit 111 stores statistical information that is the result of statistical processing on information written on each of two or more sheets, and the information processing device 1 stores each of the two or more sheets. Accepting a paper image that is an image read from the paper, performing first OCR processing on each of the two or more paper images, and acquiring two or more first OCR information that is information on a predetermined area; A second OCR process is performed on each of the two or more paper images, and two or more pieces of second OCR information, which is information on a predetermined area, are obtained, and statistical information, two or more pieces of first OCR information, and two or more pieces of second OCR information are obtained. The second OCR information is used to determine correct OCR information that is OCR information corresponding to each of two or more sheets, and inspection information that is information regarding correct OCR information corresponding to each of the determined two or more sheets is determined. Output.

In this way, for each of two or more sheets, the first OCR information by the first OCR process and the second OCR information by the second OCR process are acquired, and the stored statistical information and the acquired two By determining correct OCR information using the above first OCR information and two or more second OCR information, the lack of accuracy of OCR is compensated for, and as a result, the information processing device 1 can correct the information written on the paper. Accurate information can be obtained. Note that what can be accurately acquired is that information that matches the information written on the paper can be acquired from the paper image of the paper by OCR.

Therefore, it is possible to effectively reduce the effort required for a person to check the information written on the paper and the information obtained from the paper image by OCR.

Furthermore, in the above configuration, since the statistical information is the sum of the numerical values written on two or more sheets, the information processing device 1 can easily and effectively compensate for the lack of precision in OCR.

The information processing device 1 also receives two or more different paper images for each of the two or more papers, and performs first OCR processing on each of the two or more different paper images for each of the two or more papers. , obtain two or more pieces of first OCR information that is information of a predetermined area, and perform a second OCR process on two or more different paper images for each of the two or more papers, Obtain two or more pieces of second OCR information that are information on a predetermined area, and use the statistical information, four or more pieces of first OCR information, and four or more pieces of second OCR information to correspond to each piece of paper. The correct OCR information is determined.

In this way, two or more different paper images are acquired for each of the two or more paper images, and for each of the two or more paper images, the first OCR information is obtained by the first OCR processing, and the second OCR information is obtained by the second OCR processing. The accuracy of OCR can be improved by determining the correct OCR information using the stored statistical information and the acquired 4 or more first OCR information and 4 or more second OCR information. Deficiencies can be compensated for more effectively.

Further, in the above configuration, an identifier is exposed on each of the two or more sheets of paper, and the information processing device 1 acquires the identifier from a predetermined area by OCR processing on each of the two or more paper images and specially The identifier storage unit 112 stores one or more special identifiers that are identifiers that perform predetermined special processing, and determines whether the identifier acquired by the first OCR processing or the second OCR processing is a special identifier. This is a predetermined process for the paper on which the identifier judged to be a special identifier appears, and is a process to improve the accuracy of the first OCR process or the second OCR process. Perform processing.

In this way, by performing special processing only on the paper on which the special identifier is exposed, the information processing apparatus 1 can compensate for the lack of OCR accuracy while suppressing an increase in processing load.

Furthermore, in the above configuration, since each of the two or more sheets is a bill, the information processing device 1 can accurately acquire information such as the amount written on the bill.

Furthermore, the processing in this embodiment may be realized by software. Then, this software may be distributed by software download or the like. Furthermore, this software may be recorded on a recording medium such as a CD-ROM and distributed. Note that this also applies to other embodiments in this specification.

Note that the software that implements the information processing device 1 in this embodiment is, for example, the following program. In other words, this program causes a computer that can access the statistical information storage unit 111, which stores statistical information that is the result of statistical processing of information written on each of two or more sheets, to read each of the two or more sheets. A paper image receiving unit 121 that receives a paper image as an image performs first OCR processing on each of the two or more paper images, and obtains two or more pieces of first OCR information that is information on a predetermined area. a first OCR unit 131 that performs second OCR processing on each of the two or more paper images and obtains two or more pieces of second OCR information that is information on a predetermined area; 132, and determining correct OCR information that is OCR information corresponding to each of the two or more sheets using the statistical information, the two or more first OCR information, and the two or more second OCR information. This is a program for making the output unit 14 function as the unit 133 and inspection information that is information regarding correct OCR information corresponding to two or more sheets determined by the determining unit 133.

Note that although the information processing device 1 in this embodiment includes two OCR units (131 and 132), a first and a second OCR unit, it may further include a third OCR unit (not shown), for example. .

The third OCR unit performs third OCR processing on each of the two or more paper images, and obtains two or more pieces of third OCR information that is information on a predetermined area. The determining unit 133 selects the statistical information stored in the statistical information storage unit 111, two or more pieces of first OCR information acquired by the first OCR unit 131, and two or more pieces of second OCR information acquired by the second OCR unit 132. information and two or more pieces of third OCR information acquired by the third OCR unit to determine correct OCR information that corresponds to each of the two or more sheets.

This makes it possible to more effectively compensate for the lack of accuracy in OCR, and as a result, the accuracy of reading the information written on each sheet is improved.

Furthermore, by further increasing the number of OCR units, it is expected that reading accuracy will further improve.

FIG. 20 is an external view of a computer system 900 that implements the information processing device 1 and the like by executing the program in this embodiment. This embodiment can be realized by computer hardware and a computer program executed on the computer hardware. In FIG. 20, computer system 900 includes a computer 901 including a disk drive 905, a keyboard 902, a mouse 903, a display 904, and an optical reading device 917. Note that the entire system including the keyboard 902, mouse 903, display 904, and optical reading device 917 may also be called a computer.

FIG. 21 is a diagram showing an example of the internal configuration of the computer system 900. In FIG. 21, in addition to a disk drive 905, a computer 901 is connected to an MPU 911, a ROM 912 for storing programs such as a boot-up program, and a ROM 912 for temporarily storing instructions of an application program. A RAM 913 that provides storage space, a storage 914 that stores application programs, system programs, and data, a bus 915 that interconnects the MPU 911, ROM 912, etc., and provides connections to networks such as external networks and internal networks. A network card 916 and an optical reading device 917 are provided. The storage 914 is, for example, a hard disk, SSD, flash memory, or the like. The optical reading device 917 is, for example, a camera, a scanner, or the like.

A program that causes the computer system 900 to execute the functions of the information processing device 1 and the like may be stored on a disk 921 such as a DVD or CD-ROM, inserted into the disk drive 905, and transferred to the storage 914. Alternatively, the program may be sent to computer 901 via the network and stored in storage 914. The program is loaded into RAM 913 during execution. Note that the program may be loaded directly from the disk 921 or the network. Further, instead of the disk 921, the program may be read into the computer system 900 via another removable recording medium (for example, a DVD or a memory card).

The program does not necessarily include an operating system (OS) that causes the information processing device 1 or the like to execute functions, a third party program, etc. in 901 indicating details of the computer. A program may include only those portions of instructions that call appropriate functions or modules in a controlled manner to achieve desired results. How computer system 900 operates is well known and will not be described in detail.

Note that although the computer system 900 described above is a server or a stationary PC, the terminal device 2 may be implemented as a mobile terminal such as a tablet terminal, a smartphone, or a notebook PC. In this case, for example, it is desirable that the keyboard 902 and mouse 903 be replaced with a touch panel, the disk drive 905 with a memory card slot, and the disk 921 with a memory card.

However, the above is just an example, and the hardware configuration of the computer that implements the information processing device 1 etc. does not matter.

Note that in the above program, in the transmission step for transmitting information and the reception step for receiving information, processing is performed by hardware, such as processing performed by the modem or interface card in the transmission step (processing that can only be performed by hardware). Processing that is not carried out) is not included.

Further, the number of computers that execute the above program may be one or more. That is, one computer may perform centralized processing, or multiple computers may perform distributed processing.

Furthermore, in the embodiment described above, it goes without saying that two or more communication means (terminal transmitter 24, terminal receiver 25, etc.) existing in one device may be physically realized by one medium. .

Furthermore, in the above embodiments, each process (each function) may be realized by being centrally processed by a single device (system), or by being distributed by a plurality of devices. It's okay.

It goes without saying that the present invention is not limited to the above-described embodiments, and that various modifications can be made, and these are also included within the scope of the present invention.

As described above, the information processing device according to the present invention has the effect of accurately acquiring information written on paper by compensating for the lack of accuracy in OCR, and is useful as an information processing device, etc. .

1 Information processing device 2 Terminal device 11 Storage section 12 Reception section 13 Processing section 14 Output section 21 Terminal storage section 22 Terminal reception section 23 Terminal processing section 24 Terminal transmission section 25 Terminal reception section 26 Terminal output section 111 Statistical information storage section 112 Special Identifier storage section 121 Paper image reception section 131 First OCR section 132 Second OCR section 133 Determination section 134 Judgment section 135 Special processing section

Claims (6)

a statistical information storage unit that stores statistical information that is the result of statistical processing of the information written on each of the two or more sheets;
a paper image reception unit that receives a paper image that is an image obtained by reading each of the two or more papers;
a first OCR unit that performs first OCR processing on each of the two or more paper images and obtains two or more pieces of first OCR information that is information on a predetermined area;
a second OCR unit that performs second OCR processing on each of the two or more paper images and obtains two or more pieces of second OCR information that is information on a predetermined area;
a determining unit that determines correct OCR information that is OCR information corresponding to each of the two or more sheets using the statistical information, the two or more first OCR information, and the two or more second OCR information;
an output unit that outputs inspection information that is information regarding correct OCR information corresponding to each of the two or more sheets determined by the determination unit,
The information processing device, wherein the statistical information is a sum of numerical values written on each of the two or more sheets. a statistical information storage unit that stores statistical information that is the result of statistical processing of the information written on each of the two or more sheets;
a paper image reception unit that receives a paper image that is an image obtained by reading each of the two or more papers;
a first OCR unit that performs first OCR processing on each of the two or more paper images and obtains two or more pieces of first OCR information that is information on a predetermined area;
a second OCR unit that performs second OCR processing on each of the two or more paper images and obtains two or more pieces of second OCR information that is information on a predetermined area;
a determining unit that determines correct OCR information that is OCR information corresponding to each of the two or more sheets using the statistical information, the two or more first OCR information, and the two or more second OCR information;
an output unit that outputs inspection information that is information regarding correct OCR information corresponding to each of the two or more sheets determined by the determination unit,
The paper image reception unit includes:
accepting two or more different paper images for each of the two or more papers;
The first OCR section includes:
Performing a first OCR process on each of the two or more different paper images for each of the two or more sheets to obtain two or more pieces of first OCR information that is information on a predetermined area;
The second OCR unit includes:
performing a second OCR process on each of the two or more different paper images for each of the two or more sheets to obtain two or more second OCR information that is information on a predetermined area;
The determining unit is
An information processing device that uses the statistical information, the four or more first OCR information, and the four or more second OCR information to determine correct OCR information that is OCR information that corresponds to each of the two or more sheets . . a statistical information storage unit that stores statistical information that is the result of statistical processing of the information written on each of the two or more sheets;
a paper image reception unit that receives a paper image that is an image obtained by reading each of the two or more papers;
a first OCR unit that performs first OCR processing on each of the two or more paper images and obtains two or more pieces of first OCR information that is information on a predetermined area;
a second OCR unit that performs second OCR processing on each of the two or more paper images and obtains two or more pieces of second OCR information that is information on a predetermined area;
a determining unit that determines correct OCR information that is OCR information corresponding to each of the two or more sheets using the statistical information, the two or more first OCR information, and the two or more second OCR information;
an output unit that outputs inspection information that is information regarding correct OCR information corresponding to each of the two or more sheets determined by the determination unit,
An identifier is exposed on each of the two or more sheets,
The first OCR unit or the second OCR unit,
Obtaining an identifier from a predetermined area by OCR processing on each of the two or more paper images,
a special identifier storage unit storing one or more special identifiers that are identifiers for performing predetermined special processing;
a determining unit that determines whether the identifier acquired by the first OCR unit or the second OCR unit is a special identifier;
This is a predetermined process for a sheet on which an identifier that has been determined by the determining unit to be a special identifier is displayed, and is a process for increasing the accuracy of the determining process in which the determining unit determines correct OCR information. An information processing device further comprising a special processing unit that performs special processing. A statistical information storage section in which statistical information that is the result of statistical processing on information written on two or more sheets is stored, a paper image reception section, a first OCR section, a second OCR section, a determination section, and an output section. An information processing method realized,
a paper image receiving step in which the paper image receiving unit receives a paper image that is an image obtained by reading each of the two or more papers;
a first OCR step in which the first OCR unit performs a first OCR process on each of the two or more paper images to obtain two or more pieces of first OCR information that is information on a predetermined area; ,
a second OCR step in which the second OCR unit performs a second OCR process on each of the two or more paper images to obtain two or more pieces of second OCR information that is information on a predetermined area;
The determining unit determines correct OCR information that is OCR information corresponding to each of the two or more sheets using the statistical information, the two or more first OCR information, and the two or more second OCR information. a decision step to
an output step in which the output unit outputs inspection information that is information regarding correct OCR information corresponding to each of the two or more sheets determined by the determination unit,
The information processing method , wherein the statistical information is the sum of numerical values written on each of the two or more sheets . A computer that can access the statistical information storage unit where statistical information that is the result of statistical processing of the information written on each of the two or more sheets is stored,
a paper image reception unit that receives a paper image that is an image obtained by reading each of the two or more papers;
a first OCR unit that performs first OCR processing on each of the two or more paper images and obtains two or more pieces of first OCR information that is information on a predetermined area;
a second OCR unit that performs second OCR processing on each of the two or more paper images and obtains two or more pieces of second OCR information that is information on a predetermined area;
a determining unit that determines correct OCR information that is OCR information corresponding to each of the two or more sheets using the statistical information, the two or more first OCR information, and the two or more second OCR information;
A program for functioning as an output unit that outputs inspection information that is information regarding correct OCR information corresponding to two or more sheets determined by the determination unit,
The program, wherein the statistical information is the sum of numerical values written on each of the two or more sheets . A computer that can access the statistical information storage unit where statistical information that is the result of statistical processing of the information written on each of the two or more sheets is stored,
a paper image reception unit that receives a paper image that is an image obtained by reading each of the two or more papers;
a first OCR unit that performs first OCR processing on each of the two or more paper images and obtains two or more pieces of first OCR information that is information on a predetermined area;
a second OCR unit that performs second OCR processing on each of the two or more paper images and obtains two or more pieces of second OCR information that is information on a predetermined area;
a determining unit that determines correct OCR information that is OCR information corresponding to each of the two or more sheets using the statistical information, the two or more first OCR information, and the two or more second OCR information;
A program for functioning as an output unit that outputs inspection information that is information regarding correct OCR information corresponding to two or more sheets determined by the determination unit,
The paper image reception unit includes:
accepting two or more different paper images for each of the two or more papers;
The first OCR section includes:
Performing a first OCR process on each of the two or more different paper images for each of the two or more sheets to obtain two or more pieces of first OCR information that is information on a predetermined area;
The second OCR unit includes:
performing a second OCR process on each of the two or more different paper images for each of the two or more sheets to obtain two or more second OCR information that is information on a predetermined area;
The determining unit is
The statistical information, the four or more first OCR information, and the four or more second OCR information are used to determine correct OCR information that corresponds to each of the two or more sheets. A program that allows a computer to function.

Images