JP7151015B1 - Program, computer and information processing method - Google Patents

Abstract

A program, a computer, and an information processing method capable of preventing leakage of personal information due to theft of face images are provided. SOLUTION: An encoding conversion means 14 obtains a character string by performing encoding conversion on a face image, and a dividing means 15 divides the obtained character string into a plurality of character string parts according to a division rule. Then, the transmission means 12 associates each of the divided character string portions with the identification information and transmits them to the plurality of data stores 42, 44, 46, and 48, respectively. In addition, the receiving means 11 receives each character string part linked to the identification information from each data store 42, 44, 46, 48, and the integration means 16 integrates each character string part of the same identification information according to the division rule. The decode conversion means 17 restores the face image by decoding and converting the obtained character string. [Selection drawing] Fig. 1

Description

The present invention relates to programs, computers, and information processing methods.

Conventionally, as a face authentication system that performs authentication based on face data, there has been known a face authentication system that performs personal authentication by matching input face data with pre-registered registered face data (for example, See Patent Documents 1 and 2).

The face authentication system described in Patent Literature 1 authenticates a target person using an image of a face portion in an image captured by a camera. Further, in the face authentication system described in Patent Document 2, for example, when there are four face images to be registered, two face patterns are used for accuracy assurance, one face pattern is used for disturbance component absorption, and one face pattern is used for disturbance component absorption. is the face pattern to be updated. Then, when a new face pattern is registered, the update target face pattern having the second lowest degree of similarity with the newly registered face pattern is deleted from among the four pre-registered face patterns. That is, by leaving the face pattern for disturbance component absorption that has the lowest degree of similarity with the face pattern to be newly registered, face authentication is performed while adapting to variations in the environment.

JP 2019-197426 A Japanese Patent Application Laid-Open No. 2006-72540

It is necessary to store the face image of each user in the face recognition system in order to carry out face recognition. is stolen, it poses a serious problem of leakage of personal information.

SUMMARY OF THE INVENTION The present invention has been made in consideration of such points, and aims to provide a program, a computer, and an information processing method that can prevent personal information from being leaked due to theft of facial images. aim.

The program of the present invention is
A program that causes a computer to function as encoding conversion means, division means, transmission means, reception means, integration means, decoding conversion means, and authentication means,
The encoding conversion means acquires a character string by performing encoding conversion on the face image received from the user terminal by the receiving means,
The dividing means divides the acquired character string into a plurality of character string parts according to a division rule,
issuing user identification information when the receiving means receives face image information from the user terminal;
The transmission means links each of the divided character string parts to the issued user identification information and transmits the character string parts to a plurality of data stores;
The receiving means receives each character string part associated with the user identification information from each data store,
The integration means acquires a character string by integrating each character string part of the same user identification information according to the division rule;
The decoding conversion means restores the face image by decoding the acquired character string,
When the receiving means receives face image information from the authentication engine, it requests each data store to transmit the character string portion, thereby obtaining the character string portion linked to each of the user identification information from each of the data stores. , the authentication means receives from the authentication engine by comparing the face image received from the authentication engine and the face image linked to each of the user identification information restored by the decoding conversion means; It is characterized by specifying the user identification information corresponding to the face image.

In the program of the present invention,
The authentication means converts the face image received from the authentication engine and the face image linked to each of the user identification information restored by the decode conversion means into homomorphic encryption, and converts the converted homomorphic encryption. The user identification information corresponding to the face image received from the authentication engine may be identified by comparing .

Further, the authentication means performs homomorphic encryption in which the face images received from the authentication engine are encrypted, and homomorphic encryption in which the face images of the plurality of users restored by the decoding conversion means are respectively encrypted. When comparing, the Euglid distance is calculated for each of the user identification information, and the identification information of the user whose Euglid distance is the smallest and smaller than a predetermined threshold is the identification information corresponding to the face image received from the authentication engine. It may be specified as user identification information.

Alternatively, homomorphic encryption may be obtained by converting a user's face image into 512-dimensional vector data before being encrypted by the authentication means, and performing encryption processing on the converted 512-dimensional vector data. .

Further, in the authentication means, homomorphic encryption in which the user's face image transmitted from the authentication engine is encrypted, and homomorphic encryption for each user identification information corresponding to the user's face image restored by the decoding conversion means A user may not be authenticated if the Euclidean distance to a cipher is greater than a predetermined threshold for all of the user identities.

Further, the transmission means transmits the division rule in association with the user identification information to a data store different from the plurality of data stores to which each character string part was transmitted,
When the integration means integrates each character string portion of the same user identification information to obtain a character string, the reception means receives the division rule linked to the user identification information, and the received division rule may be used.

The program of the present invention is
A program that causes a computer to function as division means, encoding conversion means, transmission means, reception means, integration means, decoding conversion means, and authentication means,
The dividing means divides the face image received from the user terminal by the receiving means into a plurality of face image portions according to a division rule,
The encoding conversion means acquires a plurality of character string portions by performing encoding conversion on each divided face image portion,
issuing user identification information when the receiving means receives face image information from the user terminal;
The transmitting means links each acquired character string part to the issued user identification information and transmits the acquired character string part to a plurality of data stores;
The receiving means receives each character string part associated with the user identification information from each data store,
The decoding conversion means acquires a plurality of face image portions by performing decoding conversion on each received character string portion,
The integrating means integrates a plurality of face image portions of the same user identification information according to the division rule to restore the face image;
When the receiving means receives face image information from the authentication engine, it requests each data store to transmit the character string portion, thereby obtaining the character string portion linked to each of the user identification information from each of the data stores. , the authentication means compares the face image received from the authentication engine with the face image linked to each of the user identification information restored by the integration means, thereby identifying the face received from the authentication engine; It is characterized by specifying the user identification information corresponding to the image.

In the program of the present invention,
The authentication means converts the face image received from the authentication engine and the face image data linked to each of the user identification information restored by the integration means into homomorphic encryption, and converts the converted homomorphic encryption. The user identification information corresponding to the face image received from the authentication engine may be identified by comparing .

The authentication means compares the homomorphic encryption in which the face images received from the authentication engine are encrypted with the homomorphic encryption in which the face images of the plurality of users are respectively encrypted and restored by the integration means. When doing so, a Eugrid distance is calculated for each of the user identification information, and the identification information of the user whose Eugrid distance is the smallest and smaller than a predetermined threshold is obtained from the authentication engine and the user corresponding to the face image received You may specify with identification information.

Alternatively, homomorphic encryption may be obtained by converting a user's face image into 512-dimensional vector data before being encrypted by the authentication means, and performing encryption processing on the converted 512-dimensional vector data. .

Further, in the authentication means, homomorphic encryption in which the user's face image transmitted from the authentication engine is encrypted, and homomorphic encryption for each user identification information corresponding to the user's face image restored by the integration means and is greater than a predetermined threshold for all of the user identification information, authentication of the user may not be performed.

Further, the transmission means transmits the division rule in association with the user identification information to a data store different from the plurality of data stores to which each character string part was transmitted,
When the integration means integrates each character string portion of the same user identification information to obtain a character string, the reception means receives the division rule linked to the user identification information, and the received division rule may be used.

The computer of the present invention
By executing the program, it functions as encoding conversion means, division means, transmission means, reception means, integration means, decoding conversion means, and authentication means,
The encoding conversion means acquires a character string by performing encoding conversion on the face image received from the user terminal by the receiving means,
The dividing means divides the acquired character string into a plurality of character string parts according to a division rule,
issuing user identification information when the receiving means receives face image information from the user terminal;
The transmission means links each of the divided character string parts to the issued user identification information and transmits the character string parts to a plurality of data stores;
The receiving means receives each character string part associated with the user identification information from each data store,
The integration means acquires a character string by integrating each character string part of the same user identification information according to the division rule;
The decoding conversion means restores the face image by decoding the acquired character string,
When the receiving means receives face image information from the authentication engine, it requests each data store to transmit the character string portion, thereby obtaining the character string portion linked to each of the user identification information from each of the data stores. , the authentication means receives from the authentication engine by comparing the face image received from the authentication engine and the face image linked to each of the user identification information restored by the decoding conversion means; It is characterized by specifying the user identification information corresponding to the face image.

The computer of the present invention
By executing the program, it functions as dividing means, encoding conversion means, transmission means, reception means, integration means, decoding conversion means, and authentication means,
The dividing means divides the face image received from the user terminal by the receiving means into a plurality of face image portions according to a division rule,
The encoding conversion means acquires a plurality of character string portions by performing encoding conversion on each divided face image portion,
issuing user identification information when the receiving means receives face image information from the user terminal;
The transmitting means links each acquired character string part to the issued user identification information and transmits the acquired character string part to a plurality of data stores;
The receiving means receives each character string part associated with the user identification information from each data store,
The decoding conversion means acquires a plurality of face image portions by performing decoding conversion on each received character string portion,
The integrating means integrates a plurality of face image portions of the same user identification information according to the division rule to restore the face image;
When the receiving means receives face image information from the authentication engine, it requests each data store to transmit the character string portion, thereby obtaining the character string portion linked to each of the user identification information from each of the data stores. , the authentication means compares the face image received from the authentication engine with the face image linked to each of the user identification information restored by the integration means, thereby identifying the face received from the authentication engine; It is characterized by specifying the user identification information corresponding to the image.

The information processing method of the present invention comprises:
A computer-implemented information processing method comprising:
obtaining a character string by performing encoding conversion on a face image received from a user terminal;
splitting the obtained string into a plurality of string parts according to a splitting rule;
a step of issuing user identification information when face image information is received from the user terminal;
a step of linking each of the divided character string portions to the issued user identification information and transmitting the string to a plurality of data stores;
receiving each character string part associated with the user identification information from each data store;
obtaining a character string by integrating each character string part of the same user identification information according to the division rule;
a step of restoring a face image by decoding the acquired character string;
with
When face image information is received from the authentication engine, each data store is requested to transmit the character string portion, thereby obtaining the character string portion linked to each of the user identification information from each of the data stores, and the authentication engine. Identifying the user identification information corresponding to the face image received from the authentication engine by comparing the face image received from the authentication engine with the face image linked to each of the restored user identification information. Characterized by

The information processing method of the present invention comprises:
A computer-implemented information processing method comprising:
dividing a face image received from a user terminal into a plurality of face image parts according to a division rule;
obtaining a plurality of character string portions by performing encoding conversion on each divided face image portion;
a step of issuing user identification information when face image information is received from the user terminal;
a step of linking each of the obtained character string parts to the issued user identification information and transmitting them to a plurality of data stores;
receiving each character string part associated with the user identification information from each data store;
a step of obtaining a plurality of face image parts by performing decoding conversion on each received character string part;
a step of restoring a face image by integrating a plurality of face image portions of the same user identification information according to the division rule;
with
When face image information is received from the authentication engine, each data store is requested to transmit the character string portion, thereby obtaining the character string portion linked to each of the user identification information from each of the data stores, and the authentication engine. Identifying the user identification information corresponding to the face image received from the authentication engine by comparing the face image received from the authentication engine with the face image linked to each of the restored user identification information. Characterized by

According to the program, computer, and information processing method of the present invention, it is possible to prevent personal information from being leaked due to face image theft.

1 is a diagram schematically showing a face authentication system according to an embodiment of the invention; FIG. FIG. 2 is an explanatory diagram showing a state in which a face image is divided into four face image parts in the face authentication system shown in FIG. 1; 4 is a flowchart showing an operation of converting a face image into a character string, dividing it into a plurality of character string portions, and storing the character string in each data store in the face authentication system shown in FIG. 1; 2 is a flow chart showing an operation of restoring a face image by converting a character string obtained by integrating a plurality of character string portions transmitted from each data store in the face authentication system shown in FIG. 1;

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments of the present invention will be described with reference to the drawings. 1 to 4 are diagrams showing a face authentication system according to this embodiment.

The face authentication system 1 according to the present embodiment includes a face authentication server 2, an authentication engine 30, a first data store 42, a second data store 44, a third data store 46, a fourth data store 48, and a fifth data store 50 . The face authentication server 2 and each of the authentication engine 30, the first data store 42, the second data store 44, the third data store 46, the fourth data store 48 and the fifth data store 50 are connected via a network such as the Internet line. are communicatively connected to each other via Also, the face authentication server 2 is communicably connected to a user terminal 20 owned by the user via a network such as the Internet line.

The face authentication server 2 is a computer or the like installed at a system company. The face authentication server 2 has a control section 10 , a storage section 18 and a communication interface 19 . The storage unit 18 and the communication interface 19 are each connected to the control unit 10 via buses. The control unit 10 is a CPU (central processing unit) or the like, and by executing a program stored in the storage unit 18, the reception unit 11, the transmission unit 12, the authentication unit 13, the encoding conversion unit 14, the division unit 15 , as integration means 16 and decoding conversion means 17 . The storage unit 18 stores user registration information in association with user identification information. The user's registration information includes the user's name, date of birth, user name, e-mail address, telephone number of the user terminal 20, credit card information, password, and the like. The communication interface 19 transmits and receives signals to and from other devices wirelessly or by wire through a network such as the Internet line. Note that the programs executed by the control unit 10 are not limited to those stored in the storage unit 18 . When the control unit 10 executes a program stored in a recording medium attached to the face authentication server 2 or a program transmitted from an external device to the face authentication server 2, the control unit 10 sends the information to the reception unit 11. It may function as means 12 , authentication means 13 , encoding conversion means 14 , division means 15 , integration means 16 and decoding conversion means 17 .

The accepting unit 11 accepts information transmitted to the face authentication server 2 from an external device (eg, the user terminal 20, the authentication engine 30, each data store 42, 44, 46, 48, 50). The transmission means 12 transmits various information through the communication interface 19 to the external device. The authentication means 13 authenticates the user by comparing the face image transmitted from the authentication engine 30 and the face image restored by the decoding conversion means 17, which will be described later. The encoding conversion means 14 obtains a character string by performing encoding conversion on the face image. A dividing unit 15 divides the acquired character string into a plurality of character string parts according to the division rule. The integration means 16 integrates each character string part of the same user identification information according to the division rule to obtain a character string. The decode conversion means 17 restores the face image by decoding the obtained character string. The details of the functions of these means will be described later.

The authentication engine 30 captures a face image of the user and transmits the captured face image to the face authentication server 2 . Further, when the authentication information of the user is sent from the face authentication server 2, the authentication engine 30 performs various processes for the authenticated user. Specifically, as shown in FIG. 1, the authentication engine 30 has a control section 32 such as a CPU (Central Processing Unit), an imaging section 34 such as a camera, and a communication interface 39 . Here, if the authentication engine 30 is located in an office building, a collective housing facility, or the like, the control unit 32, when the user has been authenticated, is placed at the entrance of the office building, collective housing facility, or the like. Unlock the door to be placed. Also, when the authentication engine 30 is installed in restaurants, hotels, transportation facilities, convenience stores, etc., cashless settlement is possible when paying fees at these service facilities. In this case, the payment information is sent from the authentication engine 30 to the server of the financial institution or credit card company, whereby the payment amount is automatically deducted from the user's bank account or added to the credit card statement. will come to Also, the control unit 32 transmits and receives signals to and from the face authentication server 2 via a network such as an Internet line through a communication interface 39 . The imaging unit 34 is adapted to capture an image of the user's face or the like.

The first data store 42, the second data store 44, the third data store 46, the fourth data store 48, and the fifth data store 50 are each connected to the face authentication server 2 so as to be communicable via a network such as the Internet line. It is Each data store 42, 44, 46, 48, 50 stores the character string portion transmitted from the face authentication server 2 in association with the identification information of the user.

A user terminal 20 is owned by a user. The user terminal 20 includes, for example, a personal computer, a PC tablet terminal, a smart phone, and the like. The user terminal 20 has a control section 22 , a touch panel 24 , an imaging section 26 , a storage section 28 and a communication interface 29 . The control unit 22 is a CPU (Central Processing Unit) or the like. Various screens are displayed on the touch panel 24 , and the user can input various information to the control section 22 through the touch panel 24 . The imaging unit 26 is, for example, a camera, and is designed to capture an image of the user's face or the like. The storage unit 28 stores various information. The communication interface 29 transmits and receives signals to and from other devices (specifically, the face authentication server 2 and the like) wirelessly or by wire through a network such as the Internet line. In the user terminal 20, a face authentication application can be installed from an online store, the website of the management company of the face authentication server 2, or the like. When such a face authentication application is installed, the user uses the user terminal 20 to register a face image and various registration information (for example, name, date of birth, user name, e-mail address, telephone number of the user terminal 20, credit card information, etc.). information, passwords, etc.).

Next, the flow of processing by the face authentication system 1 of this embodiment will be described. FIG. 3 is a flow chart showing the operation of converting a face image into a character string in the face authentication system 1 shown in FIG. . FIG. 4 is a facial image obtained by converting a character string obtained by integrating a plurality of character string portions transmitted from the data stores 42, 44, 46, and 48 in the face authentication system 1 shown in FIG. is a flow chart showing the operation of restoring the .

First, the user registers a face image using the user terminal 20 . Specifically, the user first installs the face authentication application on the user terminal 20 . Then, when the user activates the face authentication application on the user terminal 20 , a user registration screen is displayed on the touch panel 24 . On such a user registration screen, the user inputs registration information such as name (Kanji), name (Kana), date of birth, user name, telephone number of user terminal 20, e-mail address of user terminal 20, password, etc. When the registration button is pressed after checking the box for agreeing to the terms, the touch panel 24 displays a face image pickup screen. When the user captures a facial image with the imaging unit 26 on such an imaging screen, various registration information and information on the user's facial image input on the user registration screen are transmitted from the user terminal 20 to the face authentication server 2 . In this manner, the reception means 11 of the face authentication server 2 receives registration information and face image information from the user terminal 20 . The identification information of the user terminal 20 is also received by the receiving means 11 of the face authentication server 2 . The control unit 10 of the face authentication server 2 issues user identification information (specifically, a user identification number, for example) when the reception unit 11 receives registration information and face image information from the user terminal 20 .

Next, a method for managing the face image received by the receiving means 11 of the face authentication server 2 in the face authentication system 1 will be described with reference to the flowchart shown in FIG. First, a character string is acquired by performing encoding conversion on the face image received by the receiving means 11 by the encoding conversion means 14 (STEP 1). Various known methods are used as the encoding method. For example, the encoding method is base64 encode. According to such base64 encoding, even in an environment where multi-byte characters such as Japanese cannot be handled, face images can be handled as data by converting them into 64 kinds of English characters.

Then, the character string acquired by the encoding conversion means 14 is divided into a plurality of character string parts according to the division rule by the dividing means 15 (STEP 2). Various rules can be used as splitting rules for splitting a string into multiple string parts. Dividing rules include, for example, dividing a character string every time it reaches a predetermined number of characters counted from the beginning, creating a plurality of character string parts corresponding to the number of characters obtained by dividing the character string by a predetermined number, and the like. Alternatively, a different division rule may be generated by the division means 15 each time. Specifically, several division rules are stored in the storage unit 18, and the division means 15 selects one division rule, for example, at random, from among the plurality of division rules stored in the storage unit 18. use.

After that, the transmitting means 12 links each of the divided character string portions to the user identification information and transmits them to the data stores 42, 44, 46, and 48 in a one-to-one manner (STEP 3). Here, the one-to-one mode means that each character string part is divided into each data so that each divided character string part and each of the plurality of data stores 42, 44, 46, 48 have a one-to-one relationship. It is sent to the stores 42,44,46,48. For example, if the dividing means 15 divides the character string into four character string parts, each of the divided character string parts is stored in the first data store 42, the second data store 44, the third data store 46, and the fourth data store, respectively. sent to the store 48; It should be noted that instead of each split string portion being sent to each data store 42, 44, 46, 48 in a one-to-one fashion, multiple string portions of each split string portion may contain the same data. may be sent to the store. The character string portion sent to each data store 42, 44, 46, 48 is stored in each data store 42, 44, 46, 48 while being linked to the user identification information. Further, the transmitting means 12 associates the dividing rule used when the dividing means 15 divides the character string with the user identification information, and stores it in another data store 50 (that is, for example, the fifth data store where each character string part is not transmitted). data store 50) (STEP 4). The division rule sent to the fifth data store 50 is stored in this fifth data store 50 while being linked to the user identification information.

In this manner, the face image accepted by the accepting means 11 of the face authentication server 2 is managed by the face authentication system 1 . In this case, even if a third party illegally accesses a certain data store 42, 44, 46, 48 and steals the character string portion, the information of the character string portion stored in the other data stores 42, 44, 46, 48 is Since the face image cannot be restored without it, it is possible to prevent the personal information (specifically, the user's face image) from being leaked.

Next, the operation when the imaging unit 34 of the authentication engine 30 captures the face image of the user in the face authentication system 1 will be described. FIG. 4 is a flow chart showing a method of reconstructing facial images from character string portions stored in each data store 42, 44, 46, 48. As shown in FIG.

When a user's face image is captured by an imaging unit 34 of an authentication engine 30 installed in various service organizations, the captured face image data of the user and identification information of the authentication engine 30 are sent from the authentication engine 30 to the face authentication server 2 . sent to. Then, when the receiving means 11 of the face authentication server 2 receives the face image data of the user and the identification information of the authentication engine 30 transmitted from the authentication engine 30, the transmitting means 12 sends data to the data stores 42, 44, 46, 48. Send a request signal for the string part to When each data store 42, 44, 46, 48 receives a request signal for a character string portion from the face authentication server 2, each data store 42, 44, 46, 48 stores each user identification information and each user identification. The character string portion associated with the information is transmitted from each data store 42 , 44 , 46 , 48 to the face authentication server 2 and accepted by the acceptance means 11 of the face authentication server 2 . In this way, the face authentication server 2 acquires the user identification information and the character string portion linked to this user identification information from each data store 42, 44, 46, 48 (STEP 11). Further, when the receiving means 11 of the face authentication server 2 receives the face image data of the user and the identification information of the authentication engine 30 transmitted from the authentication engine 30, the transmitting means 12 requests the data store 50 for the division rule. Send a signal. When the data store 50 receives a division rule request signal from the face authentication server 2 , each user identification information stored in the data store 50 and the division rule linked to each user identification information are transferred to the fifth data store 50 . to the face authentication server 2 and accepted by the acceptance means 11 of the face authentication server 2 . In this way, the face authentication server 2 acquires the division rule linked to the user identification information from the fifth data store 50 (STEP 12).

After that, the integrating means 16 acquires a character string by integrating each character string part of the same user identification information for each user identification information according to the division rule transmitted from the fifth data store 50 (STEP 13). . Then, the decode conversion means 17 restores the face image by decoding the acquired character string (STEP 14). In this manner, the authentication means 13 can authenticate the user by comparing the face image transmitted from the authentication engine 30 and the face image restored by the decoding conversion means 17.例文帳に追加Note that the restored face image of the user is not stored in the storage unit 18 of the face authentication server 2 . As a result, it is possible to prevent personal information from being leaked by a third party illegally accessing the face authentication server 2 and stealing the user's face image from the storage unit 18 .

Next, the details of the user authentication method by the authentication means 13 will be described. The authentication means 13 develops the face image of the user transmitted from the authentication engine 30 and accepted by the acceptance means 11 into 512-dimensional vector data. Also, the authentication means 13 converts the developed 512-dimensional vector data into homomorphic encryption. Homomorphic encryption is a cryptography that allows operations on encrypted data without decryption. It can perform binary operations such as , subtraction, and multiplication. In addition, the authentication means 13 expands the face image linked to each user identification information restored by the decoding conversion means 17 into 512-dimensional vector data, and further converts the expanded 512-dimensional vector data into homomorphic encryption. Convert. Then, the authentication means 13 converts the 512-dimensional vector data relating to the user's face image transmitted from the authentication engine 30 to the homomorphic encryption, and the 512-dimensional vector data relating to the face image of each user restored by the decoding conversion means 17. and the converted homomorphic encryption are calculated for each user's identification information. Here, the Euclidean distance is the square root of the sum of squares of the difference between two points in the 512-dimensional Euclidean space. The homomorphic _encryption (a ₁ , a ₂ , a ₃ , . When the homomorphic encryption obtained by converting the 512-dimensional vector data related to each user's face image is (b ₁ , b ₂ , b ₃ , . . . , b ₅₁₂ ), the Eugrid distance L is represented by the following formula be.

Then, the authentication unit 13 determines the user identification information corresponding to the homomorphic encryption with the smallest calculated Eugrid distance smaller than a predetermined threshold as the user identification information imaged by the imaging unit 34 . When the user identification information corresponding to the user's face image captured by the imaging unit 34 of the authentication engine 30 is determined by the authentication means 13 in this manner, the user identification information is transmitted to the authentication engine 30 by the transmission means 12. be done. When the authentication information of the user is sent from the face authentication server 2, the authentication engine 30 performs various processes for the authenticated user. Specifically, when the authentication engine 30 is placed at the entrance of an office building, a collective housing facility, or the like, the image of the user's face is captured by the imaging unit 34, and the authentication means 13 of the face authentication server 2 recognizes the user's face. When the authentication is performed, the authentication engine 30 unlocks the doors arranged at the entrances and the like of these office buildings and collective housing facilities. Further, when the authentication engine 30 is installed in a taxi, a restaurant, a hotel, a transportation facility, a convenience store, or the like, the user's face can be detected by the imaging unit 34 when paying a fee at these service facilities. When the image is captured and the authentication means 13 of the face authentication server 2 authenticates the user, the authentication engine 30 enables cashless payment.

The homomorphic encryption in which the 512-dimensional vector data relating to the user's facial image transmitted from the authentication engine 30 is converted, and the homomorphic encryption in which the 512-dimensional vector data relating to the user's facial image restored by the decoding conversion means 17 are converted. If the Eugrid distance between the ciphertext and the ciphertext is greater than a predetermined threshold for all user identification information, the user's face image captured by the imaging unit 34 of the authentication engine 30 is stored in each data store 42, 44, 46, 48 is determined not to be registered, and user authentication is not performed. In this case, the identification information of the user is not transmitted from the face authentication server 2 to the authentication engine 30, and the authentication engine 30 does not process the user.

According to the program, the computer (face authentication server 2), and the information processing method of this embodiment configured as described above, the encoding conversion means 14 obtains a character string by performing encoding conversion on the face image. Then, the dividing means 15 divides the acquired character string into a plurality of character string parts according to the division rule, and the transmitting means 12 associates each divided character string part with the user identification information to form a plurality of character string parts. Send to data stores 42, 44, 46, 48 respectively. Further, the receiving means 11 receives each character string part linked to the user identification information from each data store 42, 44, 46, 48, and the integration means 16 receives each character string part of the same user identification information according to the division rule. are integrated to obtain a character string, and the decode conversion means 17 decodes and converts the obtained character string to restore the face image. Further, when the receiving means 11 receives face image information from the authentication engine 30, each data store 42, 44, 46, 48 is requested to transmit a character string portion, and each data store 42, 44, 46, 48 The character string part linked to the user identification information is acquired, and the authentication means 13 extracts the face image received from the authentication engine 30 and the face image linked to each user identification information restored by the decoding conversion means 17. By comparing with, the user identification information corresponding to the face image received from the authentication engine 30 is specified. According to such a program, computer, and information processing method, the user's face image is not stored in the storage unit 18 of the face authentication server 2, and each character string part is stored in a plurality of data stores 42, 44, 46, and 48. shared and managed. Therefore, even if a third party illegally accesses a certain data store 42, 44, 46, 48 and steals the character string portion, the information of the character string portion stored in the other data stores 42, 44, 46, 48 is Since the face image cannot be restored without it, it is possible to prevent the personal information (specifically, the user's face image) from being leaked. Further, when the receiving means 11 receives face image information from the authentication engine 30, each data store 42, 44, 46, 48 is requested to transmit a character string portion, and each data store 42, 44, 46, 48 After obtaining the character string part linked to the user identification information and restoring the face image by the decoding conversion means 17 for each user identification information, the face image received from the authentication engine 30 and the decoding conversion means 17 restored. Since the user identification information is compared with a plurality of face images linked to each user identification information, the user identification information corresponding to the face image received from the authentication engine 30 can be specified more reliably.

Further, in the program, computer, and information processing method of the present embodiment, as described above, the authentication means 13 links the face image received from the authentication engine 30 and each user identification information restored by the decode conversion means 17. The user identification information corresponding to the face image received from the authentication engine 30 is specified by converting each attached face image data into homomorphic encryption and comparing the converted homomorphic encryption. In this case, the security can be further improved by using homomorphic encryption. In the program, computer, and information processing method of the present embodiment, the face image received from the authentication engine 30 and the face image linked to each user identification information restored by the decoding conversion means 17 are each homomorphically encrypted. , and the transformed homomorphic encryption is not limited to what is compared. The face image received from the authentication engine 30 and the face image linked to each user identification information restored by the decoding conversion means 17 may be directly compared.

Further, the authentication means 13 performs homomorphic encryption in which the face images received from the authentication engine 30 are encrypted, and homomorphic encryption in which the face images of the plurality of users restored by the decoding conversion means 17 are encrypted. When comparing, the Euglid distance is calculated for each piece of user identification information, and the identification information of the user whose Euglid distance is the smallest and smaller than a predetermined threshold is determined as the user identification information corresponding to the face image received from the authentication engine 30. and specify.

Alternatively, homomorphic encryption may be obtained by converting the face image of the user into 512-dimensional vector data before being encrypted by the authentication means 13, and performing encryption processing on the converted 512-dimensional vector data. .

In the authentication means 13, the homomorphic encryption in which the user's face image transmitted from the authentication engine 30 is encrypted, and the homomorphic encryption for each user identification information corresponding to the user's face image restored by the decoding conversion means 17 If the Eugrid distance to the cipher is greater than a predetermined threshold for all user identification information, user authentication may not be performed.

In addition, in the program, computer, and information processing method of the present embodiment, as described above, the transmitting means 12 stores data stored in data stores other than the plurality of data stores 42, 44, 46, and 48 to which the respective character string portions have been transmitted. When the division rule is linked to the user identification information and transmitted to the data store (fifth data store 50), and the integration means 16 integrates each character string part of the same user identification information to obtain a character string, the reception means 11 receives the division rule linked to the user identification information, and uses the received division rule. In this case, even if the information of all the character string parts is stolen by unauthorized access by a third party, the face image cannot be restored because these character string parts cannot be integrated without the division rule. Therefore, it is possible to prevent leakage of personal information.

Note that the program, computer, and information processing method according to the present embodiment are not limited to the aspects described above, and various modifications can be made.

For example, when dividing a face image and storing a plurality of character string portions in a plurality of data stores 42, 44, 46, and 48, instead of the method described above, the face image is first divided into a plurality of face image portions as described below. a plurality of character string portions are generated by encoding conversion of each divided face image portion, and the generated plurality of character string portions are stored in a plurality of data stores 42, 44, 46, 48; good too. Such methods are described below.

Also in the face authentication system 1 according to the modified example, the control unit 10 of the face authentication server 2 executes the program stored in the storage unit 18 so that the reception unit 11, the transmission unit 12, the authentication unit 13, the encoding conversion unit 14 , division means 15 , integration means 16 and decoding conversion means 17 . A dividing means 15 divides the face image into a plurality of face image parts according to the division rule. The encoding conversion means 14 obtains a plurality of character string portions by performing encoding conversion on each divided face image portion. The transmission unit 12 associates each acquired character string part with the identification information and transmits it to a plurality of data stores 42 , 44 , 46 , 48 . The accepting unit 11 accepts each character string part associated with the identification information from each data store 42 , 44 , 46 , 48 . The decode conversion means 17 acquires a plurality of face image parts by performing decode conversion on each received character string part. The integration means 16 restores the face image by integrating a plurality of face image portions of the same identification information according to the division rule.

Next, the flow of processing by the face authentication system 1 according to the modified example described above will be described.

First, a method of managing a face image received by the receiving means 11 of the face authentication server 2 in the face authentication system 1 will be described with reference to FIG. First, the dividing means 15 divides the face image received by the receiving means 11 into a plurality of face image portions according to the dividing rule. Specifically, as shown in FIG. 2, the face image is divided into four areas A, B, C, and D with two vertical columns and two horizontal columns. The division rule is, for example, to "divide the face image into four areas A, B, C, and D in two vertical and two horizontal lines". Note that the division rule is not limited to such a division method, and the face image may be divided into four columns vertically or four columns horizontally. Alternatively, a different division rule may be generated by the division means 15 each time. Specifically, several division rules are stored in the storage unit 18, and the division means 15 selects one division rule, for example, at random, from among the plurality of division rules stored in the storage unit 18. use.

Next, the encoding conversion means 14 performs encoding conversion on each of the plurality of face image portions divided by the dividing means 15 (specifically, the images indicated by areas A, B, C, and D in FIG. 2). Get multiple string parts by doing Various known methods are used as the encoding method. For example, the encoding method is base64 encode.

After that, the transmitting means 12 links each encoded character string part to the user identification information and transmits it to each data store 42 , 44 , 46 , 48 . For example, when the dividing means 15 divides the user's facial image into four facial image parts, each character string part obtained by encoding the divided facial image part is stored in the first data store 42 and the second data store 44, respectively. , third data store 46 and fourth data store 48 . The character string portion sent to each data store 42, 44, 46, 48 is stored in each data store 42, 44, 46, 48 while being linked to the user identification information. Further, the transmitting means 12 associates the dividing rule used by the dividing means 15 to divide the face image with the user identification information, and stores it in another data store 50 (that is, for example, a fifth data store in which each character string portion is not transmitted). data store 50). The division rule sent to the fifth data store 50 is stored in this fifth data store 50 while being linked to the user identification information.

In this manner, the face image accepted by the accepting means 11 of the face authentication server 2 is managed by the face authentication system 1 . Even in the face authentication system 1 according to the modified example, even if a third party illegally accesses a certain data store 42, 44, 46, 48 and steals the character string portion, it is stored in the other data store 42, 44, 46, 48. Since the face image cannot be restored without the information of the character string portion that is included in the character string, it is possible to prevent the personal information (specifically, the user's face image) from being leaked.

Next, the operation when the imaging unit 34 of the authentication engine 30 captures the face image of the user in the face authentication system 1 will be described.

When a user's face image is captured by an imaging unit 34 of an authentication engine 30 installed in various service organizations, the captured face image data of the user and identification information of the authentication engine 30 are sent from the authentication engine 30 to the face authentication server 2 . sent to. Then, when the receiving means 11 of the face authentication server 2 receives the face image data of the user and the identification information of the authentication engine 30 transmitted from the authentication engine 30, the transmitting means 12 sends data to the data stores 42, 44, 46, 48. Send a request signal for the string part to When each data store 42, 44, 46, 48 receives a request signal for a character string portion from the face authentication server 2, each data store 42, 44, 46, 48 stores each user identification information and each user identification. The character string portion associated with the information is transmitted from each data store 42 , 44 , 46 , 48 to the face authentication server 2 and accepted by the acceptance means 11 of the face authentication server 2 . In this way, the face authentication server 2 acquires the user identification information and the character string portion associated with the user identification information from each of the data stores 42, 44, 46, and 48. Further, when the receiving means 11 of the face authentication server 2 receives the face image data of the user and the identification information of the authentication engine 30 transmitted from the authentication engine 30, the transmitting means 12 requests the data store 50 for the division rule. Send a signal. When the data store 50 receives the division rule request signal from the face authentication server 2, the user identification information linked to each user identification information and each user identification information stored in the data store 50 is associated with A division rule is transmitted from the fifth data store 50 to the face authentication server 2 and accepted by the acceptance means 11 of the face authentication server 2 . In this way, the face authentication server 2 acquires the division rule linked to the user identification information from the fifth data store 50. FIG.

After that, the decode conversion means 17 restores a plurality of face image parts by decoding and converting each character string part linked to each user identification information accepted by the acceptance means 11 . Then, the integrating means 16 restores the face image by integrating the plurality of face image portions for each user identification information according to the division rule transmitted from the fifth data store 50 . In this way, the face authentication server 2 can perform authentication using the user's face image. Note that the restored face image of the user is not stored in the storage unit 18 of the face authentication server 2 . As a result, it is possible to prevent personal information from being leaked by a third party illegally accessing the face authentication server 2 and stealing the user's face image from the storage unit 18 .

Next, the details of the user authentication method by the authentication means 13 will be described. The authentication means 13 develops the face image of the user transmitted from the authentication engine 30 and accepted by the acceptance means 11 into 512-dimensional vector data. Also, the authentication means 13 converts the developed 512-dimensional vector data into homomorphic encryption. Further, the authentication means 13 expands the face image linked to each user identification information restored by the integrating means 16 into 512-dimensional vector data, and further converts the expanded 512-dimensional vector data into homomorphic encryption. do. Then, the authentication means 13 stores the homomorphic encryption into which the 512-dimensional vector data relating to the user's face image transmitted from the authentication engine 30 is converted, and the 512-dimensional vector data relating to the face image of each user restored by the integration means 16. The Eugrid distance from the converted homomorphic encryption is calculated for each user's identification information. The method of calculating the Eugrid distance is the homomorphic encryption in which the 512-dimensional vector data regarding the user's face image transmitted from the authentication engine 30 as described above is converted, and the face image of each user restored by the decode conversion means 17. Since the calculation method of the Eugrid distance between the 512-dimensional vector data of and the converted homomorphic encryption is substantially the same, the explanation is omitted.

Then, the authentication means 13 captures the identification information of the user corresponding to the homomorphic encryption whose calculated Eugrid distance is the smallest and smaller than a predetermined threshold value for each user's face image restored by the integrating means 16. It is determined to be the user's identification information imaged by the unit 34 . When the user's identification information corresponding to the user's face image captured by the imaging unit 34 of the authentication engine 30 is determined by the authentication means 13 in this way, this user's identification information is sent to the authentication engine 30 by the transmission means 12. sent. When the authentication information of the user is sent from the face authentication server 2, the authentication engine 30 performs various processes for the authenticated user.

Homomorphic encryption in which the 512-dimensional vector data relating to the user's facial image transmitted from the authentication engine 30 is converted, and homomorphic encryption in which the 512-dimensional vector data relating to the user's facial image restored by the integrating means 16 are converted. When the Eugrid distance between is greater than a predetermined threshold for all user identification information, the user's face image data captured by the imaging unit 34 of the authentication engine 30 is stored in each data store 42, 44, 46, 48 is determined not to be registered, and user authentication is not performed. In this case, the identification information of the user is not transmitted from the face authentication server 2 to the authentication engine 30, and the authentication engine 30 does not process the user.

According to the program, computer (face authentication server 2), and information processing method according to such a modification, the dividing means 15 divides the face image into a plurality of face image parts according to the dividing rule, and the encoding conversion means 14 acquires a plurality of character string portions by performing encoding conversion on each of the divided face image portions, and transmitting means 12 transmits each of the divided character string portions with identification information (specifically, user identification information) and transmitted to a plurality of data stores 42 , 44 , 46 , 48 . Further, the receiving means 11 receives each character string part linked to the identification information from each of the data stores 42, 44, 46, and 48, and the decoding conversion means 17 decodes and converts each received character string part. Thus, a plurality of face image portions are obtained, and the integrating means 16 restores the face image by integrating the plurality of face image portions of the same identification information according to the division rule. Further, when the receiving means 11 receives face image information from the authentication engine 30, each data store 42, 44, 46, 48 is requested to transmit a character string portion, and each data store 42, 44, 46, 48 The character string part linked to the user identification information is acquired, and the authentication means 13 combines the face image received from the authentication engine 30 and the face image linked to each user identification information restored by the integration means 16. By comparing , the user identification information corresponding to the face image received from the authentication engine 30 is specified. Even with such a program, computer, and information processing method, the user's face image is not stored in the storage unit 18 of the face authentication server 2, and each character string portion is shared by a plurality of data stores 42, 44, 46, and 48. managed by Therefore, even if a third party illegally accesses a certain data store 42, 44, 46, 48 and steals the character string portion, the information of the character string portion stored in the other data stores 42, 44, 46, 48 is Since the face image cannot be restored without it, it is possible to prevent the leakage of personal information. Further, when the receiving means 11 receives face image information from the authentication engine 30, each data store 42, 44, 46, 48 is requested to transmit a character string portion, and each data store 42, 44, 46, 48 After acquiring the character string part linked to the user identification information and restoring the face image for each user identification information by the integrating means 16, the face image received from the authentication engine 30 and each restored by the integrating means 16 Since multiple face images linked to the user identification information are compared, the user identification information corresponding to the face image received from the authentication engine 30 can be specified more reliably.

Further, in the program, computer, and information processing method according to the modification, as described above, the authentication means 13 is associated with each user identification information restored by the face image received from the authentication engine 30 and the integration means 16. The user identification information corresponding to the face image received from the authentication engine 30 is specified by converting each face image data into homomorphic encryption and comparing the converted homomorphic encryption. In this case, the security can be further improved by using homomorphic encryption. In the program, computer, and information processing method of the present embodiment, the face image received from the authentication engine 30 and the face image linked to each user identification information restored by the integrating means 16 are each homomorphically encrypted. The transformed and transformed homomorphic encryptions are not limited to what they are compared to. The face image received from the authentication engine 30 and the face image linked to each user identification information restored by the integrating means 16 may be directly compared.

Further, the authentication means 13 compares the homomorphic encryption in which the face images received from the authentication engine 30 are encrypted and the homomorphic encryption in which the face images of the plurality of users respectively encrypted by the integration means 16 are encrypted. When doing so, the Euglid distance is calculated for each user identification information, and the identification information of the user whose Eugrid distance is the smallest and smaller than a predetermined threshold is used as the user identification information corresponding to the face image received from the authentication engine 30 Identify.

Alternatively, homomorphic encryption may be obtained by converting the face image of the user into 512-dimensional vector data before being encrypted by the authentication means 13, and performing encryption processing on the converted 512-dimensional vector data. .

In the authentication means 13, homomorphic encryption in which the user's face image transmitted from the authentication engine 30 is encrypted, and homomorphic encryption for each user identification information corresponding to the user's face image restored by the integration means 16 If the Eugrid distance between is greater than a predetermined threshold for all pieces of user identification information, user authentication may not be performed.

In addition, in the program, computer, and information processing method according to the modification, as described above, the transmission unit 12 transmits data stored in a data store different from the plurality of data stores 42, 44, 46, 48 to which each character string part is transmitted. When the division rule is linked to the user identification information and transmitted to the (fifth data store 50), and the integration unit 16 integrates a plurality of face image portions of the same user identification information to restore the face image, the reception unit 11 receives the division rule linked to the user identification information by and uses the received division rule. In this case, even if all the character string information is stolen by a third party through unauthorized access, these character string parts cannot be integrated, making it impossible to restore the face image. Information leakage can be prevented.

Further, in the above description, each character string part is linked to the user identification information and transmitted to each data store 42, 44, 46, 48 by the transmission means 12, and the division rule is linked to the user identification information. Thus, the mode of transmission to the fifth data store 50 by the transmission means 12 has been described. However, the program, computer, and information processing method according to the present invention are not limited to such aspects. As another aspect, each character string part is linked to identification information other than the user identification information (for example, identification information of the user terminal 20, etc.) and is transmitted to each data store 42, 44, 46, 48 by the transmission means 12. Alternatively, the division rule may be linked to identification information other than the user identification information (for example, identification information of the user terminal 20 or the like) and transmitted to the fifth data store 50 by the transmission means 12 . Even in this case, when the facial image is restored in the face authentication server 2, a plurality of character string parts and facial image parts linked to the same identification information are integrated using the division rule linked to the identification information. be able to.

Further, in the above description, the division rule for dividing character strings and face images is transmitted to a data store (fifth data store 50) different from the data stores 42, 44, 46, and 48 in which the character string portions are stored. We have described how to However, the program, computer, and information processing method according to the present invention are not limited to such aspects. As another aspect, a division rule for dividing a character string or a face image is transmitted to one of the data stores 42, 44, 46, and 48 in which character string portions are stored and stored in this data store. may In still another aspect, a division rule for dividing a character string or a face image may be stored in the storage unit 18 of the face authentication server 2 .

Further, in the program, computer, and information processing method according to the present invention, the encoding conversion method of the face image and the face image part by the encoding conversion means and the decoding conversion method of the character string and the character string part by the decoding conversion means are described above. The method is not limited. Various other encoding conversions are possible as long as they can convert face images and face image parts into character strings and character string parts, and restore face images and face image parts from character strings and character string parts. Any method or decoding transform method may be used.

1 face authentication system 2 face authentication server 10 control unit 11 acceptance unit 12 transmission unit 13 authentication unit 14 encoding conversion unit 15 division unit 16 integration unit 17 decoding conversion unit 18 storage unit 19 communication interface 20 user terminal 22 control unit 24 touch panel 26 imaging Unit 28 Storage unit 29 Communication interface 30 Authentication engine 32 Control unit 34 Imaging unit 39 Communication interface 42 First data store 44 Second data store 46 Third data store 48 Fourth data store 50 Fifth data store

Claims (16)

A program that causes a computer to function as encoding conversion means, division means, transmission means, reception means, integration means, decoding conversion means, and authentication means,
The encoding conversion means acquires a character string by performing encoding conversion on the face image received from the user terminal by the receiving means,
The dividing means divides the acquired character string into a plurality of character string parts according to a division rule,
issuing user identification information when the receiving means receives face image information from the user terminal;
The transmission means links each of the divided character string parts to the issued user identification information and transmits the character string parts to a plurality of data stores;
The receiving means receives each character string part associated with the user identification information from each data store,
The integration means acquires a character string by integrating each character string part of the same user identification information according to the division rule;
The decoding conversion means restores the face image by decoding the acquired character string,
When the receiving means receives face image information from the authentication engine, it requests each data store to transmit the character string portion, thereby obtaining the character string portion linked to each of the user identification information from each of the data stores. , the authentication means receives from the authentication engine by comparing the face image received from the authentication engine and the face image linked to each of the user identification information restored by the decoding conversion means; A program for identifying the user identification information corresponding to the face image. The authentication means converts the face image received from the authentication engine and the face image linked to each of the user identification information restored by the decode conversion means into homomorphic encryption, and converts the converted homomorphic encryption. 2. The program according to claim 1, wherein the user identification information corresponding to the face image received from the authentication engine is specified by comparing . The authentication means compares the homomorphic encryption in which the face image received from the authentication engine is encrypted with the homomorphic encryption in which the face images of the plurality of users respectively encrypted by the decode conversion means are encrypted. In this case, the Euglid distance is calculated for each of the user identification information, and the identification information of the user whose Euglid distance is the smallest and smaller than a predetermined threshold is used as the user identification corresponding to the face image received from the authentication engine. 3. The program of claim 2, identifying information. 4. A homomorphic encryption is obtained by converting a face image of a user into 512-dimensional vector data before being encrypted by said authentication means, and performing an encryption process on the converted 512-dimensional vector data. program as described. In the authentication means, homomorphic encryption in which the user's facial image transmitted from the authentication engine is encrypted, and homomorphic encryption for each user identification information corresponding to the user's facial image restored by the decoding conversion means. 4. The program according to claim 3, wherein user authentication is not performed if the Eugrid distance between is greater than a predetermined threshold for all of the user identification information. The transmission means transmits the division rule in association with the user identification information to a data store different from the plurality of data stores to which each character string part was transmitted,
When the integration means integrates each character string portion of the same user identification information to obtain a character string, the reception means receives the division rule linked to the user identification information, and the received division rule 2. The program according to claim 1, wherein the program uses A program that causes a computer to function as division means, encoding conversion means, transmission means, reception means, integration means, decoding conversion means, and authentication means,
The dividing means divides the face image received from the user terminal by the receiving means into a plurality of face image portions according to a division rule,
The encoding conversion means acquires a plurality of character string portions by performing encoding conversion on each divided face image portion,
issuing user identification information when the receiving means receives face image information from the user terminal;
The transmitting means links each acquired character string part to the issued user identification information and transmits the acquired character string part to a plurality of data stores;
The receiving means receives each character string part associated with the user identification information from each data store,
The decoding conversion means acquires a plurality of face image portions by performing decoding conversion on each received character string portion,
The integrating means integrates a plurality of face image portions of the same user identification information according to the division rule to restore the face image;
When the receiving means receives face image information from the authentication engine, it requests each data store to transmit the character string portion, thereby obtaining the character string portion linked to each of the user identification information from each of the data stores. , the authentication means compares the face image received from the authentication engine with the face image linked to each of the user identification information restored by the integration means, thereby identifying the face received from the authentication engine; A program for identifying the user identification information corresponding to an image. The authentication means converts the face image received from the authentication engine and the face image data linked to each of the user identification information restored by the integration means into homomorphic encryption, and converts the converted homomorphic encryption. 8. The program according to claim 7, wherein the user identification information corresponding to the face image received from the authentication engine is specified by comparing . When the authentication means compares the homomorphic encryption in which the face images received from the authentication engine are encrypted and the homomorphic encryption in which the face images of the plurality of users respectively restored by the integration means are encrypted, Then, the Euglid distance is calculated for each of the user identification information, and the identification information of the user whose Euglid distance is the smallest and smaller than a predetermined threshold is the user identification information corresponding to the face image received from the authentication engine. 9. The program according to claim 8, which specifies: 10. The user's facial image is converted into 512-dimensional vector data before being encrypted by said authentication means, and homomorphic encryption is obtained by performing encryption processing of the converted 512-dimensional vector data. program as described. In the authentication means, homomorphic encryption in which the user's face image transmitted from the authentication engine is encrypted, and homomorphic encryption for each user identification information corresponding to the user's face image restored by the integration means. 10. The program according to claim 9, wherein if the Euclidean distance between is greater than a predetermined threshold for all of the user identities, user authentication is not performed. The transmission means transmits the division rule in association with the user identification information to a data store different from the plurality of data stores to which each character string part was transmitted,
When the integration means integrates each character string portion of the same user identification information to obtain a character string, the reception means receives the division rule linked to the user identification information, and the received division rule 8. The program according to claim 7, which uses By executing the program, it functions as encoding conversion means, division means, transmission means, reception means, integration means, decoding conversion means, and authentication means,
The encoding conversion means acquires a character string by performing encoding conversion on the face image received from the user terminal by the receiving means,
The dividing means divides the acquired character string into a plurality of character string parts according to a division rule,
issuing user identification information when the receiving means receives face image information from the user terminal;
The transmission means links each of the divided character string parts to the issued user identification information and transmits the character string parts to a plurality of data stores;
The receiving means receives each character string part associated with the user identification information from each data store,
The integration means acquires a character string by integrating each character string part of the same user identification information according to the division rule;
The decoding conversion means restores the face image by decoding the acquired character string,
When the receiving means receives face image information from the authentication engine, it requests each data store to transmit the character string portion, thereby obtaining the character string portion linked to each of the user identification information from each of the data stores. , the authentication means receives from the authentication engine by comparing the face image received from the authentication engine and the face image linked to each of the user identification information restored by the decoding conversion means; A computer that identifies the user identification information corresponding to the facial image. By executing the program, it functions as dividing means, encoding conversion means, transmission means, reception means, integration means, decoding conversion means, and authentication means,
The dividing means divides the face image received from the user terminal by the receiving means into a plurality of face image portions according to a division rule,
The encoding conversion means acquires a plurality of character string portions by performing encoding conversion on each divided face image portion,
issuing user identification information when the receiving means receives face image information from the user terminal;
The transmitting means links each acquired character string part to the issued user identification information and transmits the acquired character string part to a plurality of data stores;
The receiving means receives each character string part associated with the user identification information from each data store,
The decoding conversion means acquires a plurality of face image portions by performing decoding conversion on each received character string portion,
The integrating means integrates a plurality of face image portions of the same user identification information according to the division rule to restore the face image;
When the receiving means receives face image information from the authentication engine, it requests each data store to transmit the character string portion, thereby obtaining the character string portion linked to each of the user identification information from each of the data stores. , the authentication means compares the face image received from the authentication engine with the face image linked to each of the user identification information restored by the integration means, thereby identifying the face received from the authentication engine; A computer that identifies the user identification information corresponding to an image. A computer-implemented information processing method comprising:
obtaining a character string by performing encoding conversion on a face image received from a user terminal;
splitting the obtained string into a plurality of string parts according to a splitting rule;
a step of issuing user identification information when face image information is received from the user terminal;
a step of linking each of the divided character string portions to the issued user identification information and transmitting the string to a plurality of data stores;
receiving each character string part associated with the user identification information from each data store;
obtaining a character string by integrating each character string part of the same user identification information according to the division rule;
a step of restoring a face image by decoding the acquired character string;
with
When face image information is received from the authentication engine, each data store is requested to transmit the character string portion, thereby obtaining the character string portion linked to each of the user identification information from each of the data stores, and the authentication engine. Information that identifies the user identification information corresponding to the face image received from the authentication engine by comparing the face image received from and the restored face image linked to each of the user identification information Processing method. A computer-implemented information processing method comprising:
dividing a face image received from a user terminal into a plurality of face image parts according to a division rule;
obtaining a plurality of character string portions by performing encoding conversion on each divided face image portion;
a step of issuing user identification information when face image information is received from the user terminal;
a step of linking each of the obtained character string parts to the issued user identification information and transmitting them to a plurality of data stores;
receiving each character string part associated with the user identification information from each data store;
a step of obtaining a plurality of face image parts by performing decoding conversion on each received character string part;
a step of restoring a face image by integrating a plurality of face image portions of the same user identification information according to the division rule;
with
When face image information is received from the authentication engine, each data store is requested to transmit the character string portion, thereby obtaining the character string portion linked to each of the user identification information from each of the data stores, and the authentication engine. Information that identifies the user identification information corresponding to the face image received from the authentication engine by comparing the face image received from and the restored face image linked to each of the user identification information Processing method.

Images