JP7435010B2 - Information processing device, information processing method, information processing program - Google Patents

Description

The present invention relates to an information processing device, an information processing method, and an information processing program for determining the authenticity of a document in which information used for identity verification is recorded.

Conventionally, when signing a contract, a person may present an official document to confirm the person's identity, and once the person's identity is verified, the contract can be executed.

Patent Document 1 discloses that a registered character pattern of characters included in a certificate is recorded in association with the issuing area and issuing period of the certificate, and the issuing area and issuing period of the certificate are determined based on the characters included in the image. A technique is described that acquires an associated registered character pattern and determines whether or not characters included in an image match the registered character pattern.

Patent Document 2 exemplifies a driver's license as an official document for identifying the person, and the characteristic data of the driver's license, such as character font color information, along with the issuing organization and date of issue, is managed in a database and presented. When determining the authenticity of a driver's license, the document describes a technology that recognizes the issuing authority and date of issue based on the image data of the driver's license, and extracts characteristic data from a database.

Patent Document 3 describes a technique in which a template is an average image obtained by combining a plurality of images of the same type of coin, and it is determined whether or not a coin to be checked is a counterfeit coin.

Patent Document 4 exemplifies a driver's license as an official document for identifying a person, and describes the external characteristics of multiple driver's licenses (layout: characters, images, seals, symbols, etc. printed on the surface). A technology that extracts (sequence, position, character font), stores it as an authenticity determination logic template, and uses the authenticity determination logic template to determine whether a scanned driver's license is genuine or forged. is listed.

JP 2019-003504 Publication Japanese Patent Application Publication No. 2004-118643 WO2006/006356 publication Japanese Patent Application Publication No. 2011-178075

However, the font used for driver's licenses, which are frequently used as official documents to verify the identity of a person, varies depending on each prefecture's public safety commission, the place of issuance of the public safety commission, and the time of issuance. For example, it is difficult to check the authenticity by comparing with a template.

Furthermore, the character font specifications for driver's licenses have not been officially disclosed by the Public Safety Commission, making it difficult to uniformly check the layout and character characteristics of driver's licenses across the country.

The present invention improves the accuracy of authenticity checks by performing a limited type of numeric format check in addition to basic checks by comparing with templates of character fonts used in official documents to verify the identity of the person. The purpose is to obtain an information processing device, an information processing method, and an information processing program that can be improved.

The information processing device according to the present invention includes an extraction unit that extracts a number used for determining the authenticity of the official document from an official document for verifying a person, and stores characteristic information regarding the dimensions of a specific part of the number. a storage unit, a measurement unit that measures the dimensions of the specific part of the number extracted by the extraction unit, and a verification unit that collates the measurement value measured by the measurement unit with the characteristic information stored in the storage unit. , a determination unit that determines the authenticity of the official document based on the verification result of the verification unit , and the official document for verifying the person is a driver's license containing personal information, and The number is extracted from a license number item field in which only numbers are listed, and the storage unit stores the characteristic information for each limited font applied to the driver's license, and the The present invention is characterized in that the verification unit identifies the font of the driver's license, and collates the characteristic information read based on the identified font with the measured value.
Further, the information processing device according to the present invention includes an OCR processing unit that performs OCR processing on a data item that is a written item for determining consistency from an official document for verifying the person; a consistency check unit that determines the consistency of the data items that have been processed; and a font pattern identification unit that specifies a font pattern for a part of the data item when the consistency check unit determines that there is consistency. , an extraction unit that extracts a number used for determining the authenticity of the official document; a storage unit that stores characteristic information regarding the size of a specific part of the number; and a storage unit that stores characteristic information regarding the size of the specific part of the number extracted by the extraction unit; A measurement unit that performs measurement; a verification unit that collates the measurement value measured by the measurement unit with characteristic information stored in the storage unit; and a verification result of the font pattern identification unit and the verification unit. and a determination unit that determines the authenticity of the official document based on the result.

The information processing method according to the present invention extracts a number used for determining the authenticity of the official document from an official document for verifying the person, measures the dimensions of a specific part of the extracted number, and measures the measured value. and characteristic information regarding the dimensions of a specific part of the number stored in advance, and when determining the authenticity of the official document based on the matching result, it is determined that the official document for confirming the person is an individual. The number is extracted from the driver's license field in which information is written and only numbers are listed, and the characteristics are extracted for each limited font applied to the driver's license. The device stores information, specifies the font of the driver's license, and compares the characteristic information read out based on the specified font with the measured value .

The information processing program according to the present invention is characterized in that it causes a computer to operate as each part of an information processing device.

As explained above, in the present invention, in addition to the basic check by comparison with the template of the character font adopted in official documents for verifying the person, the authenticity can be verified by checking the form of characteristic characters that meet predetermined conditions. This has the effect of improving the accuracy of checking.

(A) is a schematic diagram of the procedural examination system according to the present embodiment, and (B) is a front view of a driver's license check result image displayed on the output device (monitor) of the examination terminal device. FIG. 2 is a functional block diagram in which processes for forgery determination check executed by the forgery determination support server (control unit) according to the present embodiment are classified by function. (A) is a front view of the driver's license image, and (B) is a front view of the driver's license image with overlapping division frames when classified by item. FIG. 3 is a communication protocol diagram executed between an examination terminal device, an examination management control device, and a forgery determination support server. Regarding the example of this embodiment, (A) is a front view of a driver's license image in which the division frame of the item (license number) selected in this example is superimposed, and (B) is a front view of the driver's license image shown in FIG. It is an enlarged view of the division frame (license number) selected in A). It is a list diagram in which numbers to be authenticated are classified into font A, font B, and font C applied to driver's licenses. (A) and (C) are flowcharts of the procedure for determining the authenticity of a license number (number) extracted from FIG. 5(B).

FIG. 1 shows a schematic diagram of a procedural examination system 10 according to the present embodiment.

The procedure examination system 10 is, for example, a system for examining whether or not a requester is the person in question at a stage before performing a procedure at a financial institution.

For example, when the requester 12 launches a dedicated application program from the communication terminal device (smartphone, PC "personal computer", etc.) 14 that he or she is carrying and requests a procedure, information regarding the examination for the procedure is sent to the network. The examination terminal device 18 is notified via the 16 wireless base stations 16A.

The examination terminal device 18 includes an input device (keyboard, mouse, etc.) 18A, an output device (monitor, printer, etc.) 18B, and a control section 18C. Note that a touch panel having both the functions of the input device 18A and the output device 18B may be used. Although not shown, the control unit 18C is mainly composed of a microcomputer including a CPU, RAM, ROM, input/output ports, and a bus connecting these.

A plurality of examination terminal devices 18 are installed, each facing an operator (not shown), and handles procedures from a plurality of clients 12. The operator requests the examination management control device 20 to examine, through the network 16, the data to be examined, which includes official documents for verifying the person, out of the notified information.

In this embodiment, the driver's license image information sent from the client 12 is used as the official document for verifying the person. The driver's license image information is the driver's license image 22 (front side), as shown in FIG. 3(A).

The examination management control device 20 includes a control section 20A and a database 20B. Although not shown, the control unit 20A is mainly composed of a microcomputer including a CPU, RAM, ROM, input/output ports, and a bus connecting these.

Information for executing examination procedures is stored in advance in the database 20B.

The examination management control device 20 performs examination on the assumption that the presented driver's license image 22 is genuine. A request is made to the forgery determination support server 24 via the network 16 to determine whether the forgery determination support server 24 is authentic.

That is, driver's license image information (driver's license image 22) is transmitted from the examination management control device 20 to the forgery determination support server 24.

The forgery determination support server 24 includes a control section 24A and a database 24B.

The database 24B is provided with various storage sections that store check items for determining whether the driver's license image 22 is forged or not (authenticity determination) (details will be described later).

The forgery determination support server 24 broadly categorizes the received driver's license image 22 and performs two types of matters (consistency and image identification).

The forgery determination support server 24 sends back to the examination management control device 20 the overall check results of consistency and presence or absence of forgery in image identification.

The examination management control device 20 returns the authenticity determination information of the examination target data received from the examination terminal device 18.

In the examination terminal device 18, the authenticity determination information sent from the examination management control device 20 is displayed on the output device 18B (monitor).

The information displayed on the output device 18B (monitor) does not specify authenticity, but rather indicates the degree of possibility of forgery for each check item executed by the forgery identification support server 24, as shown in FIG. 1(B). It is now displayed.

For example, in FIG. 1(B), the check items and the results of each check item are displayed as "o" or "x" on the output device 18B (monitor) of the examination terminal device 18. Therefore, since "○" or "x" is displayed for each check item, the final determination of authenticity is left to the operator who operates the examination terminal device 18. In other words, the forgery identification support server 24 provides information for the operator who is operating the examination terminal device 18 to determine the authenticity, regardless of the operator's skill level. It becomes possible to average the judgments.

Note that the results to be displayed are not limited to "○" and "x", but may be classified into three or more levels, and the possibility of forgery may be displayed using a numerical value (percentage).

Further, in addition to the display screen shown in FIG. 1B, the degree of detail of the authenticity determination may be changed over a plurality of pages, and the screen may be switched depending on the skill level of the operator.

(Details of check by forgery identification support server 24)

FIG. 2 is a functional block diagram in which the processing for checking forgery determination performed by the control unit 24A of the forgery determination support server 24 is classified by function.

Note that each block does not limit the hardware configuration of the control unit 24A of the forgery identification support server 24, and each function may be executed as software by a processing program executed by a CPU of a microcomputer. .

In addition, instead of processing programs in a unified manner like a microcomputer, the control unit 24A is constructed of IC chips such as ASICs configured for each function divided as appropriate, and the IC chips cooperate with each other to verify authenticity. The configuration may be such that a determination operation is executed.

As shown in FIG. 2, the forgery determination support server 24 (control unit 24A) includes an image information reception unit 50, and receives the driver's license image 22 (FIG. 3(A)) as image data from the examination management control device 20. ).

The driver's license image 22 received by the image information reception unit 50 is temporarily stored in the reception image information storage unit 52.

The received image information storage section 52 is connected to an OCR processing section 54 and an image identification processing section 56.

The OCR processing unit 54 converts character images of each item (10 items (a) to (j), described later) of the driver's license image 22 stored in the reception image information storage unit 52 into character data by OCR processing. get.

The OCR processing unit 54 is connected to the consistency checking unit 58 and sends the acquired character data to the consistency checking unit 58.

The consistency confirmation unit 58 reads information regarding a single data check item from the single data check item storage unit 60 of the database 24B, and also reads information regarding an inter-data correlation check item from the inter-data correlation check item storage unit 62. It will be done.

The consistency checking unit 58 executes the following consistency processing to check consistency.

(1) Consistency processing

As shown in FIG. 3(B), the driver's license image 22 is classified into the following 11 items (a) to (k).
(a) Name (b) Address (c) Date of birth (d) Date of issue (e) Expiration date (f) License number (g) Public Safety Commission type (h) License conditions, etc. (i) License acquisition date (j) License type column (k) Photo (image)

In addition, in ``(h) License conditions'', as an example, ``glasses, etc.'' may be written, but there are also cases where the field is left blank.

As items for determining consistency, 10 items (a) to (j) excluding item (k) are converted into data by performing OCR processing in the OCR processing unit 54, and are checked using a single data item. Perform a correlation check on two or more data items.

For example, the check based on a single data item includes checking for the existence of a date, checking for a future date, checking for being under 16 years old, etc. in the date of birth of item (c).

Also, for example, a correlation check with two data is a correlation between the date of birth in item (c) and the date of license acquisition in item (i). A check is performed to see if a year has passed.

It is preferable to perform a total of about 20 types of consistency checks using all of the above 10 items.

The consistency check unit 58 is connected to the result management unit 64 and sends out consistency check result information.

Further, the consistency confirmation section 58 is connected to the image identification processing section 56. Based on the consistency check result, the consistency check unit 58 outputs information regarding whether or not image recognition processing is necessary to be performed, to the image recognition processing unit 56.

That is, if it is determined that there is no forgery and the image is normal at the time of the consistency check, information that requires image identification processing is output in order to further strictly determine forgery.

Note that the image identification process may be executed regardless of the consistency check result.

When the image identification processing section 56 receives information requiring image identification processing from the consistency confirmation section 58, it reads the driver's license image 22 from the received image information storage section 52 and executes the image identification processing described below.

(2) Image identification processing

In the above-mentioned consistency (1), if it is determined that consistency is achieved, image identification is performed. Note that this does not negate performing image identification regardless of the presence or absence of consistency.

The image identification processing unit 56 first reads font pattern template information from the font pattern template storage unit 66 of the database 24B, and performs template determination of the font pattern to be adopted for the driver's license image 22 as a whole.

The present applicant has discovered that the font patterns adopted for driver's licenses are aggregated into several types of font patterns. Therefore, using the template of the font pattern, the font used in the driver's license image 22 to be determined for forgery is specified.

As shown in FIG. 3(B), the items to be determined based on image identification are: (a) name, (b) address, and (f) driver's license image 22 among the above ten items. number, and (k) apply the photo (image).

The image identification processing section 56 reads the image data check information from the check item storage section 68 of the database 24B, and executes each check.

(No. 1) Space spacing check for single characters

This is a check item adopted as a feature of the present invention, and the characteristic character stored in the characteristic character information storage section 74 of the database 24B is extracted from among the characters of the address displayed in item (b), and the characteristic character is The distance between lines forming characters is measured (“pixel”), and the measured value is sent to the forgery confirmation unit 70.

The forgery confirmation unit 70 reads out the frequency distribution of the font pattern, which is stored in a database in advance, from the frequency distribution storage unit 72 of the database 24B, and determines whether the measured value is within an acceptable range. Characteristic characters are selected from characters that frequently appear in each public safety commission.

(No. 2) License number font check

It is determined which of the three types of font patterns adopted for driver's licenses each digit of the license number displayed in item (f) belongs to.

There are 10 types of numbers from "0" to "9", and the driver's license number uses a different font from the numbers used for other items, and the font is different from the numbers used for other items. Easy to compare differences. The details of the check procedure in "(No. 2) License number font check" will be described later.

The forgery confirmation section 70 is connected to the result integration section 64 and sends the forgery confirmation result based on the image recognition process executed by the image identification processing section 56 to the result integration section 64 .

The result integrating unit 64 integrates the results of the consistency checking unit 58 and forgery checking unit 70 and sends them to the examination management control device 20 via the check result sending unit 78.

Thereby, the check result is sent from the examination management control device 20 to the examination terminal device 18, and the check result is displayed on the output device 18B (monitor) as shown in FIG. 1(B).

The operation of this embodiment will be explained below based on the communication protocol executed between the examination terminal device 18, examination management control device 20, and forgery determination support server 24 shown in FIG. Note that the following processing procedure is an example of executing the consistency check and image identification processing, and this processing procedure is not limited.

When a request for an examination procedure is received from the client 12, the examination terminal device 18 first collates unexamined data (step 100), and then selects examination target data (step 102).

If the data to be examined exists, the examination terminal device 18 transmits the data to be examined to the examination management control device 20 (step 104). Upon receiving this transmission, the examination management control device 20 transmits the driver's license image 22, which is an official document for verifying the person, included in the examination target data, to the forgery determination support server 24 (step 106).

The forgery identification support server 24 executes a consistency check (single data check and inter-data correlation check) based on each item of the received driver's license image 22 (step 108), and determines whether or not image identification processing is necessary. The judgment is made and the judgment result is sent to the result management section 64 (step 110).

If it is determined that image identification processing is necessary (step 112), image identification processing is executed (step 114). In the image identification process, a character space check, a single character shape check, a number font check, a registered image comparison check, and a modification check are executed.

Furthermore, if it is determined that the image identification process is unnecessary (step 116), the image identification process is skipped (step 118).

The result management unit 64 manages only the check results of the consistency check or the check results of the consistency check and image identification processing (step 118).

The forgery determination support server 24 returns the check result to the examination management control device 20 (step 120), and then the examination management control device 20 returns the authenticity determination information of the data to be examined to the examination terminal device 18 (step 122). ).

In the examination terminal device 18, the check result is displayed on the output device 18B (monitor), and the operator confirms this display (step 124) and determines the final authenticity based on the check result information (step 126). .

As explained above, in this embodiment, the information necessary for determining the authenticity of a driver's license for identity verification, which has been inconsistent in the judgment of each operator facing the examination terminal device 18, is transferred to the forgery determination support server. 24. In other words, the forgery identification support server 24 measures the integrity of the character data stored in the driver's license, which is an official document used to verify the identity of the person, and the format of the image based on the font patterns and check items stored in the database. The presence or absence of forgery is checked based on the frequency distribution of the values, etc., and the check result is displayed on the output device 18B of the examination terminal device 18. Thereby, it is possible to average the authenticity judgments of a plurality of operators regardless of their skill levels.

(Example)

In the above embodiment, the general authenticity determination process (consistency confirmation and image identification process) in the forgery determination support server 24 has been described. In particular, in the image identification process, five types of image data check information were read out and each check was executed.

In this example, detailed check procedures for "(No. 2) Driver's license number font check", which is a characteristic check requirement of the present invention among the check items of image identification processing, will be explained with reference to FIGS. 5 to 7. do.

As shown in FIG. 5A, in "(No. 2) License number font check", the "(f) Driving license" item is selected.

The reason for choosing "(f) Driver's license" is that each public safety commission that issues driver's licenses only requires numbers to be written, and there are up to 10 types of numbers (from "0" to "9"). It is a list of numbers, and it is only necessary to determine the authenticity of the 10 patterns of numbers.

Further, although the typeface (font) applied to driver's licenses differs depending on the public safety commission, it was found that the typefaces for driver's licenses can be classified into fonts such as the example shown in FIG. Therefore, in this embodiment, as shown in FIG. 5(B), "(f) License number" of a specific public safety commission is selected, and the image identification process is performed based on the procedure for determining the authenticity of the license number. , ``(No. 2) License number font check'' was executed.

As shown in FIG. 7A, in the first step 150 for "(No. 2) License number font check", a template is read from the font pattern template storage section 66 of the database 24B and the extracted template is read out from the font pattern template storage section 66 of the database 24B. The font is identified from the entire license number (see FIG. 5(B)) (step 152).

In the next step 154, authenticity is determined for each number, that is, for each of the 10 types of numbers from "0" to "9".
This determination of authenticity is performed based on an average image of a plurality of numbers classified for each typeface, as shown in FIG.

As an example, the database 24B of the forgery discrimination support server 24 is provided with a frequency distribution storage section 72, and the frequency of each measurement value of a specific part of the number of each typeface is stored in this frequency distribution storage section 72. Read out the distribution characteristic diagram (big data) (see FIG. 7(B)). As shown in FIG. 7(B), in the case of a normal distribution, the peak value of the characteristic diagram often becomes the average value.

Next, the authenticity is determined based on where the measured value of the specific part of the number to be determined belongs in the characteristic diagram of the frequency distribution shown in FIG. 7(B) (step 154). For example, a permissible range is set in advance in FIG. 7(B), and the presence or absence of forgery is determined based on whether or not the permissible range is within the permissible range (step 156).

Here, as shown in Figure 6, among the numbers classified by typeface, the parts surrounded by dotted circles indicate deviations from the allowable range of the frequency distribution in Figure 7 (A). ing. In other words, it can be determined that there is a possibility that the number has been manipulated, and in the authenticity determination, it is determined that it is a "counterfeit". In the subsequent forgery determination, it may be determined as a forgery ("x"), but depending on the degree of deviation from the allowable range, it may be determined as "△", the degree of forgery is 60%, etc. Further, counterfeiting may be determined based on the authenticity ratio of all the numbers in the license number.

In this embodiment, all digits of the driver's license number are targeted, but numbers with diagonal lines in the background may be excluded because they tend to cause errors in determining authenticity.

In addition, the data for verification stored in the database 24B of the forgery identification support server 24 may differ from the driver's license, as the font applied by each public safety commission may change depending on when the driver's license is issued. The verification data to be read may be selected based on the stated date of issue.

10 Procedure examination system 12 Client 14 Communication terminal device 16 Network 16A Wireless base station 18 Examination terminal device 18A Input device 18B Output device 18C Control section 20 Examination management control device 22 Driver's license image 20A Control section 20B Database 24 Forgery determination support server 24A Control unit 24B Database 50 Image information reception unit 52 Received image information storage unit 54 OCR processing unit 56 Image identification processing unit 58 Consistency confirmation unit 60 Single data check item storage unit 62 Inter-data correlation check item storage unit 64 Result management unit 66 Font pattern template storage section 68 Check item storage section 70 Forgery confirmation section 72 Frequency distribution storage section 74 Characteristic character information storage section 76 Registered image storage section 78 Check result sending section

Claims (5)

an extraction unit that extracts a number used for determining the authenticity of the official document from the official document for verifying the person;
a storage unit that stores characteristic information regarding the dimensions of the specific part of the number;
a measurement unit that measures the dimensions of the specific part of the number extracted by the extraction unit;
a collation unit that collates the measurement value measured by the measurement unit and the characteristic information stored in the storage unit;
and a determination unit that determines the authenticity of the official document based on the verification result of the verification unit ,
The official document for verifying the person is a driver's license on which personal information is written, and the number is extracted from the license number item column in which only numbers are listed,
The storage unit stores the characteristic information for each limited font applied to the driver's license,
The verification unit identifies the font of the driver's license, and collates the characteristic information read based on the identified font and the measured value.
Information processing device. an OCR processing unit that performs OCR processing on data items that are entry items for determining consistency from official documents for verifying the person;
a consistency checking unit that determines the consistency of the data item subjected to the OCR processing;
a font pattern identifying unit that identifies a font pattern for a part of the data item when the consistency checking unit determines that there is consistency;
an extraction unit that extracts numbers used to determine the authenticity of the official document;
a storage unit that stores characteristic information regarding the dimensions of the specific part of the number;
a measurement unit that measures the dimensions of the specific part of the number extracted by the extraction unit;
a collation unit that collates the measurement value measured by the measurement unit and the characteristic information stored in the storage unit;
a determination unit that determines the authenticity of the official document based on the identification result of the font pattern identification unit and the verification result of the verification unit;
An information processing device having : Claim 1 or 2, wherein the characteristic information stored in the storage unit is stored separately for each condition under which the applicable number may be changed, including the delivery time and delivery location of the official document. 2. The information processing device according to 2 . Extracting a number used for determining the authenticity of the official document from an official document for verifying the person,
Measure the dimensions of a specific part of the extracted number,
Compare the measured value with characteristic information about the dimensions of the specific part of the number stored in advance,
When determining the authenticity of the official document based on the verification results,
The official document for verifying the person is a driver's license on which personal information is written, and the number is extracted from the license number item column in which only numbers are listed,
The characteristic information is stored for each limited font applied to the driver's license,
identifying the font of the driver's license, and comparing the characteristic information read based on the identified font with the measured value;
Information processing method. computer,
operating as each part of the information processing device according to any one of claims 1 to 3;
Information processing program.