JP6839313B1 - Face recognition methods, programs, recording media and face recognition systems - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a face authentication method, a program, a recording medium and a face authentication system capable of easily performing face authentication with a plurality of authentication engines. A face authentication method performed by a face authentication system 10 is a step of receiving a user's face image data and a user's identification information, and a feature amount of the user's face image based on the received user's face image data. The process of extracting each of the plurality of authentication engines 30A, 30B, 30C and the information related to the feature amount of the user's facial image for each of the plurality of extracted authentication engines 30A, 30B, 30C are associated with the identification information of the received user. It has a process of storing it in a memory. [Selection diagram] Fig. 1

Description

The present invention relates to face recognition methods, programs, recording media and face recognition systems.

Conventionally, as a face recognition system that authenticates based on face data, a face recognition system that authenticates a person by collating the input input face data with the registered face data registered in advance is known (for example,). See Patent Documents 1 and 2).

In the face recognition system described in Patent Document 1, the target person is authenticated using the image of the face portion in the captured image captured by the camera. Further, in the face recognition system described in Patent Document 2, for example, when four face images are registered, two are used as a face pattern for guaranteeing accuracy, one is used as a face pattern for absorbing disturbance components, and one is used. Is the face pattern for the update target. Then, when newly registering the face pattern, among the four face patterns registered in advance, the face pattern for the update target having the second lowest degree of similarity to the newly registered face pattern is deleted. That is, by leaving the face pattern for absorbing disturbance components having the lowest degree of similarity to the newly registered face pattern, face recognition is performed in accordance with the variation in the environment.

JP-A-2019-197426 Japanese Unexamined Patent Publication No. 2006-72540

It is generally widespread to provide an authentication engine using face recognition, but in the face recognition systems of Patent Documents 1 and 2, face recognition cannot be performed by an authentication engine in which face data is not registered. Therefore, in order to authenticate with each of the plurality of authentication engines, the user needs to register the face data with each of the plurality of authentication engines, which is troublesome and stressful.

The present invention has been made in consideration of such a point, and an object of the present invention is to provide a face recognition method, a program, a recording medium, and a face recognition system capable of easily performing face recognition by a plurality of authentication engines. And.

The face authentication method of the present invention is a face authentication method performed by a face authentication system, and a step of accepting a user's face image data and a feature amount of the user's face image based on the received user's face image data are determined. It is provided with a step of extracting for each of a plurality of authentication engines and a step of associating the extracted information related to the feature amount of the user's facial image for each authentication engine with the user's identification information and storing it in the memory. It is a feature.

The program of the present invention is a program executed by a processor for performing a face recognition method by a face recognition system, and the processor executes the program to receive and accept user face image data. The feature amount of the user's face image is extracted for each of a plurality of authentication engines based on the face image data of the user, and the information related to the feature amount of the user's face image for each of the extracted multiple authentication engines is used as the user's identification information. It is characterized in that it is associated and stored in a memory.

The recording medium of the present invention is a recording medium on which a program for performing a face recognition method by a face recognition system, which is executed by a processor, is recorded, and the user's face image is obtained by the processor executing the program. Data is received, the feature amount of the user's face image is extracted for each of the plurality of authentication engines based on the received user's face image data, and the feature amount of the user's face image for each of the extracted plurality of authentication engines is related. The feature is that the information is associated with the user's identification information and stored in the memory.

The face recognition system of the present invention is a face recognition system including a processor, and the processor receives and accepts user's face image data by executing a program and the processor executing the program. The feature amount of the user's face image is extracted for each of a plurality of authentication engines based on the user's face image data, and the information related to the feature amount of the user's face image for each of the extracted plurality of authentication engines is identified by the user. It is characterized in that it is associated with information and stored in a memory.

According to the face recognition method, program, recording medium and face recognition system of the present invention, face recognition can be easily performed by a plurality of authentication engines.

It is a figure which shows typically an example of the face recognition system by embodiment of this invention, and the authentication engine arranged in each service organization. It is a flowchart which shows the processing content which is executed when the face image of a user is registered by the face recognition system shown in FIG. It is a flowchart which shows an example of the processing content which is executed when the user is authenticated by the authentication engine and the face recognition system shown in FIG. It is a figure which shows typically another example of the face recognition system by embodiment of this invention, and the authentication engine arranged in each service organization. It is a flowchart which shows the processing content which is executed when the face image of a user is registered by the face recognition system shown in FIG. It is a flowchart which shows an example of the processing content which is executed when the user is authenticated by the authentication engine and the face recognition system shown in FIG. It is a figure which shows the still other example of the face recognition system by embodiment of this invention and the authentication engine arranged in each service organization schematicly. It is a flowchart which shows the processing content which is executed when the face image of a user is registered by the face recognition system shown in FIG. 7. It is a flowchart which shows an example of the processing content which is executed when the user is authenticated by the authentication engine and the face recognition system shown in FIG. 7. It is a figure which shows the initial screen displayed on a user terminal. It is a figure which shows the registration screen displayed on the user terminal. It is a figure which shows the registration screen displayed on the user terminal. It is a figure which shows the service addition screen displayed on the user terminal. It is a figure which shows the screen of the authentication history displayed on the user terminal. It is a figure which shows the detailed screen of the authentication history displayed on a user terminal. It is a figure which shows the still other example of the face recognition system by embodiment of this invention and the authentication engine arranged in each company schematicly.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. 1 to 15 are diagrams showing a face recognition system according to the present embodiment and an authentication engine arranged in each service organization.

As shown in FIG. 1, authentication engines 30A, 30B, and 30C for authenticating users are arranged in each service organization such as restaurants, hotels, transportation facilities, office buildings, apartment buildings, and convenience stores. Although three authentication engines 30A, 30B, and 30C are shown in FIG. 1, two or four or more authentication engines may be used. Further, a face recognition system 10 installed in a company different from each service organization is connected to each of the authentication engines 30A, 30B, and 30C so as to be able to communicate with each other via a network such as an Internet line. Further, the user terminal 20 such as a smartphone owned by the user is communicably connected to the face authentication system 10 and each authentication engine 30A, 30B, 30C via a network such as an Internet line. Hereinafter, the details of the face recognition system 10 and the authentication engines 30A, 30B, and 30C will be described.

The face recognition system 10 is composed of, for example, a computer or the like, and the face recognition system 10 includes a processor 12 such as a CPU, a memory 16, and a communication unit 18. The processor 12 is extracted by the feature amount extraction means 14 that extracts the feature amount of the user's face image corresponding to each authentication engine 30A, 30B, 30C as a hash value based on the face image data received from the user terminal 20. It has a registration means 13 for registering a feature amount of a user's face image. Then, the processor 12 executes the program stored in the memory 16 to extract the feature amount of the user's face image corresponding to each authentication engine 30A, 30B, 30C as a hash value in the feature amount extracting means 14. It has become like. Specifically, the feature amount extracting means 14 extracts a hash value obtained by a predetermined hash function from the received user's face image data as a feature amount of the user's face image. Even if the user's face image data is the same, the feature amount of the user's face image may differ depending on the types of the authentication engines 30A, 30B, and 30C. This is because the specifications of the authentication engines 30A, 30B, and 30C used in these service organizations are different when the service organizations such as business companies are different. Therefore, the feature amount extracting means 14 extracts the feature amount of the user's face image for each of the authentication engines 30A, 30B, and 30C. That is, the feature amounts of the face images of a plurality of users corresponding to the authentication engines 30A, 30B, and 30C are extracted from one face image data. Further, the processor 12 executes the program stored in the memory 16 to execute the feature amount of the user's face image for each of the authentication engines 30A, 30B, and 30C extracted by the feature amount extracting means 14 (specifically, the feature amount of the user's face image). , Hash value) is stored in the memory 16 in association with the user's identification information and the identification information of the authentication engines 30A, 30B, 30C by the registration means 13. The program executed by the processor 12 is not limited to the one stored in the memory 16. The feature amount extraction means 14 and the registration means are executed by the processor 12 executing a program transmitted from an external device to the face recognition system 10 or a program stored in a recording medium detachably attached to the face recognition system 10. Processing may be performed in each of the thirteenth.

As described above, in the memory 16, the feature amount (specifically, the hash value) of the user's face image extracted by the feature amount extracting means 14 is the user's identification information and the authentication engines 30A, 30B, 30C. It is associated with the identification information and stored. As described above, the feature amount of the user's face image extracted by the feature amount extracting means 14 may differ depending on the types of the authentication engines 30A, 30B, and 30C. Therefore, the feature amount extracting means 14 extracts the feature amount of the user's face image for each of the authentication engines 30A, 30B, and 30C. As a result, it is necessary to store the feature amount of the user's face image extracted by the feature amount extracting means 14 in the memory 16 in association with each of the authentication engines 30A, 30B, and 30C. Further, the memory 16 stores the user's identification information in association with the service institution selected by the user. Further, as described above, the memory 16 stores a program for causing the processor 12 to perform various processes. Further, the processor 12 transmits and receives signals to and from the authentication engines 30A, 30B, 30C or the user terminal 20 arranged in each service organization via a network such as an Internet line by the communication unit 18.

Each of the authentication engines 30A, 30B, 30C arranged in each service organization is composed of, for example, a computer or the like, and each of the authentication engines 30A, 30B, 30C has processors 32A, 32B, 32C such as a CPU, and memory 36A, It has 36B, 36C, imaging units 38A, 38B, 38C, processing units 40A, 40B, 40C, and communication units 42A, 42B, 42C. The processors 32A, 32B, 32C and the feature amount extraction means 34A, 34B, 34C that extract the feature amount of the user's face image as a hash value based on the user's face image data imaged by the imaging units 38A, 38B, 38C. , 35A, 35B, 35C and the authentication means. Then, the processors 32A, 32B, 32C extract the feature amount of the user's face image as a hash value by the feature amount extracting means 34A, 34B, 34C by executing the program stored in the memories 36A, 36B, 36C. It is designed to do. Specifically, the feature amount extraction means 34A, 34B, 34C extracts a hash value obtained from the received user's face image data by a predetermined hash function as a feature amount of the user's face image. Further, the processors 32A, 32B, 32C authenticate the user in the authentication means 35A, 35B, 35C by executing the program stored in the memories 36A, 36B, 36C. Details of such processing contents will be described later. The programs executed by the processors 32A, 32B, 32C are not limited to those stored in the memories 36A, 36B, 36C. The processors 32A, 32B, 32C execute a program transmitted from an external device to the authentication engines 30A, 30B, 30C and a program stored in a recording medium detachably attached to the authentication engines 30A, 30B, 30C. Therefore, various processes may be performed in each of the feature amount extraction means 34A, 34B, 34C and the authentication means 35A, 35B, 35C. Further, the authentication engines 30A, 30B, and 30C are not limited to those corresponding to one service organization, respectively. For example, the authentication engine 30A may correspond to a plurality of service organizations.

Further, the memories 36A, 36B, and 36C are adapted to store the user's identification information registered in advance in association with the feature amount of the face image of the user. Further, as described above, the memories 36A, 36B, 36C are adapted to store programs for performing various processes in the processors 32A, 32B, 32C. Further, the imaging units 38A, 38B, and 38C have, for example, a camera, and acquire the face image data of the user by imaging the user.

In addition, the processing units 40A, 40B, and 40C are adapted to perform various processes on the authenticated user. For example, when the authentication engines 30A, 30B, 30C are arranged in an office building, an apartment building, or the like, the processing units 40A, 40B, 40C may use these office buildings or the office buildings when the user is authenticated. Doors placed at the entrances and exits of apartment buildings will be unlocked. In addition, when the authentication engines 30A, 30B, and 30C are installed in restaurants, hotels, transportation facilities, convenience stores, etc., cashless payment is possible when paying fees at these service institutions. .. In this case, the payment information is sent from the authentication engines 30A, 30B, and 30C to the server of the financial institution or credit card company, so that the payment amount is automatically deducted from the user's bank account or the credit card usage statement. Will be added to. Further, the processors 32A, 32B, and 32C are adapted to transmit and receive signals to and from the face recognition system 10 or the user terminal 20 via a network such as an Internet line by the communication units 42A, 42B, and 42C.

Further, in each authentication engine 30A, 30B, 30C, the user's face image is imaged by the imaging units 38A, 38B, 38C, and the feature amount of the user's face image is registered in each authentication engine 30A, 30B, 30C. You can do it. Specifically, when the feature amount of the user's face image is registered by each authentication engine 30A, 30B, 30C, when the user's face image is imaged by the imaging units 38A, 38B, 38C, the processors 32A, 32B, The 32C extracts the feature amount of the user's face image as a hash value by the feature amount extracting means 34A, 34B, 34C based on the user's face image data captured by the imaging units 38A, 38B, 38C. Then, the feature amount (specifically, the hash value) of the extracted user's face image is stored in the memories 36A, 36B, and 36C. In this way, the feature amount of the user's face image is registered in each of the authentication engines 30A, 30B, and 30C.

The face recognition application can be installed on the user terminal 20 from an online store or the like. When such a face recognition application is installed, the user can register face image data, register a service institution that uses the service, and the like on the user terminal 20. The processing content of such a face recognition application will be described later. In addition, such a face recognition application may be provided by the face recognition system 10, or may be provided by a system different from the face recognition system 10.

In the present embodiment, the face recognition system 10 and the authentication engines 30A, 30B, and 30C arranged in each service organization are API-linked (application programming interface). As a result, the systems of the authentication engines 30A, 30B, and 30C can be easily configured as compared with the systems of the authentication engines 30A, 30B, and 30C that are independently configured.

Next, the processing contents when the user is authenticated by the face authentication system 10 and the authentication engines 30A, 30B, and 30C will be described with reference to FIGS. 2, 3, and 10 to 12.

First, a process in which the user registers the face image data in the face authentication system 10 using the user terminal 20 will be described. First, the user installs the face recognition application on the user terminal 20. On the initial screen of such a face recognition application, the screen shown in FIG. 10 is displayed. The user first registers the user on the user terminal 20 with such a face recognition application. Specifically, when the account button is pressed on the screen shown in FIG. 10, the user registration screen as shown in FIG. 11 is displayed. On such a user registration screen, enter registration information such as name, date of birth, telephone number, e-mail address, password, etc., check the box to agree to the terms of use, and then press the registration button, as shown in FIG. It becomes an imaging screen of such a face image. When the user captures a face image with the user terminal 20 on such an imaging screen, various information input on the user registration screen and the user's face image data are transmitted from the user terminal 20 to the face recognition system 10. In this way, the processor 12 of the face recognition system 10 receives the face image data from the user terminal 20 (STEP 1). Further, the identification information of the user terminal 20 and the registration information such as the user's name, date of birth, telephone number, e-mail address, password, etc. input to the user terminal 20 can also be obtained from the user terminal 20 by the processor 12 of the face recognition system 10. receive. Further, the processor 12 issues a user ID as user identification information, and stores the issued user ID in the memory 16.

Next, the processor 12 of the face authentication system 10 uses the feature amount of the face image corresponding to each authentication engine 30A, 30B, 30C as a hash value based on the face image data received from the user terminal 20, and the feature amount extraction means 14 Extracted by (STEP2). At this time, the feature amount of the user's face image extracted by the feature amount extracting means 14 may differ depending on the types of the authentication engines 30A, 30B, and 30C. Therefore, the feature amount extracting means 14 extracts the feature amount of the user's face image for each of the authentication engines 30A, 30B, and 30C. Further, the feature amount extracting means 14 extracts only the feature amount of the user's face image corresponding to the authentication engines 30A, 30B, 30C of the service organization selected in advance by the user. For example, the service institution where the authentication engine 30A is installed allows the user to use the face image data, but the service institution where the authentication engine 30B is installed does not allow the user to use the face image data. In this case, the feature amount extracting means 14 extracts only the feature amount of the user's face image corresponding to the authentication engine 30A. As described above, the memory 16 stores the user's identification information in association with the service institution selected by the user. Then, the processor 12 registers the feature amounts of the face images corresponding to the authentication engines 30A, 30B, and 30C extracted by the feature amount extraction means 14 into the memory 16 of the face recognition system 10 by the registration means 13 (user ID (user's). Identification information) and storage in association with the identification information of the authentication engines 30A, 30B, 30C (STEP3). In this way, the registration of the user's face image captured by the user terminal 20 is completed. Further, the information related to the feature amount of the user's face image stored in the memory 16 is transmitted from the face recognition system 10 to the authentication engines 30A, 30B, 30C of the service organization selected by the user (STEP 4). At this time, the feature amount of the user's face image corresponding to the authentication engines 30A, 30B, 30C is transmitted to the authentication engines 30A, 30B, 30C. The authentication engines 30A, 30B, and 30C store information related to the feature amount of the user's face image transmitted from the face recognition system 10 in the memories 36A, 36B, and 36C in association with the user's identification information.

Next, the process of authenticating the user in each of the authentication engines 30A, 30B, and 30C will be described. When user authentication is required at the service organization where each authentication engine 30A, 30B, 30C is installed, first, the user is imaged by the image pickup units 38A, 38B, 38C of the authentication engines 30A, 30B, 30C. Acquires the user's face image data by (STEP 11). Further, in the authentication engines 30A, 30B, 30C, the processors 32A, 32B, 32C extract the feature amount of the user's face image as a hash value by the feature amount extracting means 34A, 34B, 34C based on the acquired face image data. (STEP12). Then, the processors 32A, 32B, 32C are stored in the feature amounts (specifically, hash values) of the user's face image extracted by the feature amount extraction means 34A, 34B, 34C and the memories 36A, 36B, 36C. The user is authenticated by comparing the feature amount (specifically, the hash value) of the user's face image (STEP 13). More specifically, the matching rate between the feature amount of the user's face image extracted by the feature amount extraction means 34A, 34B, 34C and the feature amount of the user's face image stored in the memories 36A, 36B, 36C When exceeds a predetermined threshold value (for example, 80%), the authentication means 35A, 35B, and 35C authenticate the user. As described above, the memories 36A, 36B, and 36C store the user's identification information registered in advance and the feature amount of the user's face image in association with each other.

Then, when the user is authenticated by the authentication means 35A, 35B, 35C (“YES” in STEP 14), the processors 32A, 32B, 32C are connected to the authentication engines 30A, 30B, 30C by the processing units 40A, 40B, 40C. The corresponding service can be implemented (STEP 15). Specifically, as described above, when the authentication engines 30A, 30B, 30C are arranged in an office building, an apartment building, or the like, the processing units 40A, 40B, 40C have authenticated the user. In some cases, unlock the doors located at the entrances and exits of these office buildings and housing complexes. In addition, when the authentication engines 30A, 30B, and 30C are installed in restaurants, hotels, transportation facilities, convenience stores, etc., cashless payment is possible when paying fees at these service institutions. .. When cashless payment is performed, two-step verification may be performed. After that, the processors 32A, 32B, 32C of the authentication engines 30A, 30B, 30C transmit the information related to the usage status of the service to the face recognition system 10 (STEP 17). As a result, in the face recognition system 10, information relating to the usage status of the service in each service institution is stored in the memory 16 for each user.

On the other hand, the feature amount of the user's face image extracted by the feature amount extraction means 34A, 34B, 34C does not substantially match the feature amount of the user's face image stored in the memories 36A, 36B, 36C, and the authentication means. If the user cannot be authenticated by 35A, 35B, 35C (“NO” in STEP 14), the processors 32A, 32B, 32C are subjected to the authentication engines 30A, 30B, by the processing units 40A, 40B, 40C, respectively. The service corresponding to 30C cannot be implemented (STEP 16). In this case as well, the processors 32A, 32B, and 32C of the authentication engines 30A, 30B, and 30C use the information related to the service usage status (specifically, the user tries to authenticate with the authentication engines 30A, 30B, and 30C, but the authentication is performed. Information that the service could not be used without being used) is transmitted to the face authentication system 10 (STEP 17).

According to such an authentication method, the user's facial image features are registered in the authentication engine of the service organization (for example, the authentication engine 30A), but the authentication engine of another service organization (for example, the authentication engine 30B). ), The feature amount of this user's face image is stored in the memory 16 of the face authentication system 10 and the memories 36A, 36B, 36C of each authentication engine 30A, 30B, 30C of the authentication engines 30A, 30B, 30C. By storing it in association with the identification information, even a service institution in which the feature amount of the user's face image is not registered can authenticate the user by capturing the user's face image by the imaging unit. In this case, since it is not necessary for the user to register the face image data in all of the plurality of authentication engines, the user's trouble can be saved.

The face recognition system 10 and the face recognition method according to the present embodiment are not limited to those shown in FIGS. 1 to 3. Other examples of the face recognition system 10 and the face recognition method according to the present embodiment will be described with reference to FIGS. 4 to 6. The face recognition system 10 and the authentication engines 30A, 30B and 30C shown in FIG. 4 are designated by the same reference numerals for the same components as the face recognition system 10 and the authentication engines 30A, 30B and 30C shown in FIG. The description will be omitted.

As shown in FIG. 4, the processor 12 of the face authentication system 10 has a hash value of the feature amount of the user's face image corresponding to each authentication engine 30A, 30B, 30C based on the face image data received from the user terminal 20. The feature amount extraction means 14 to be extracted as the above, the registration means 13 to register the feature amount of the extracted user's face image, and the user's face image captured by the imaging units 38A to 38C of each authentication engine 30A, 30B, 30C. It has an authentication means 15 that authenticates a user based on the feature amount (specifically, a hash value) of the above. By executing the program stored in the memory 16, the processor 12 obtains the feature amount (specifically, the hash value) of the user's face image extracted by the feature amount extracting means 14 in the registration means 13 from the user. The identification information of the above and the identification information of the authentication engines 30A, 30B, and 30C are associated with each other and stored in the memory 16. Further, the processor 12 executes a program stored in the memory 16 to perform a feature amount of the user's face image captured by the image pickup units 38A to 38C of the authentication engines 30A, 30B, and 30C in the authentication means 15. (Specifically, the hash value) is compared with the feature amount (specifically, the hash value) of the user's face image corresponding to each of the authentication engines 30A, 30B, and 30C stored in the memory 16. Therefore, the user is authenticated. The program executed by the processor 12 is not limited to the one stored in the memory 16. The feature amount extraction means 14 and the registration means are executed by the processor 12 executing a program transmitted from an external device to the face recognition system 10 or a program stored in a recording medium detachably attached to the face recognition system 10. Processing may be performed in each of the 13 and the authentication means 15.

Each of the authentication engines 30A, 30B, 30C arranged in each service organization is composed of, for example, a computer or the like, and each of the authentication engines 30A, 30B, 30C has processors 32A, 32B, 32C such as a CPU, and memory 36A, It has 36B, 36C, imaging units 38A, 38B, 38C, processing units 40A, 40B, 40C, and communication units 42A, 42B, 42C. The processors 32A, 32B, 32C use feature amount extraction means 34A, 34B, 34C for extracting the feature amount of the user's face image as a hash value based on the user's face image data imaged by the imaging units 38A, 38B, 38C. Have. In the example shown in FIG. 4, the processors 32A, 32B, 32C do not have the authentication means 35A, 35B, 35C as shown in FIG. The processors 32A, 32B, 32C execute the program stored in the memories 36A, 36B, 36C so that the feature amount extracting means 34A, 34B, 34C extracts the feature amount of the user's face image as a hash value. It has become. Specifically, the feature amount extraction means 34A, 34B, 34C extracts a hash value obtained from the received user's face image data by a predetermined hash function as a feature amount of the user's face image. The programs executed by the processors 32A, 32B, 32C are not limited to those stored in the memories 36A, 36B, 36C. The processors 32A, 32B, 32C execute the program transmitted from the external device to the authentication engines 30A, 30B, 30C and the program stored in the recording medium detachably attached to the authentication engines 30A, 30B, 30C. Therefore, various treatments may be performed in the feature amount extraction means 34A, 34B, 34C. Further, in the example shown in FIG. 4, the features of the user's face image are not stored in the memories 36A, 36B, and 36C.

In the example shown in FIG. 4, the face recognition system 10 and the authentication engines 30A, 30B, and 30C arranged in each service organization are API-linked (application programming interface). As a result, the systems of the authentication engines 30A, 30B, and 30C can be easily configured as compared with the systems of the authentication engines 30A, 30B, and 30C that are independently configured.

Next, the processing contents when the user is authenticated by the face authentication system 10 and the authentication engines 30A, 30B, and 30C as shown in FIG. 4 will be described with reference to FIGS. 5 and 6.

First, a process in which the user registers the face image data in the face authentication system 10 using the user terminal 20 will be described. Since the specific method for the user to capture the face image by the user terminal 20 has already been described, the specific method will be omitted here. When the user first registers the user with such a face recognition application on the user terminal 20, various information input on the user registration screen and the user's face image data are transmitted from the user terminal 20 to the face recognition system 10. .. In this way, the processor 12 of the face recognition system 10 receives the face image data from the user terminal 20 (STEP 21). Further, the identification information of the user terminal 20 and the registration information such as the user's name, date of birth, telephone number, e-mail address, password, etc. input to the user terminal 20 can also be obtained from the user terminal 20 by the processor 12 of the face recognition system 10. receive. Further, the processor 12 issues a user ID as user identification information, and stores the issued user ID in the memory 16.

Next, the processor 12 of the face authentication system 10 uses the feature amount of the face image corresponding to each authentication engine 30A, 30B, 30C as a hash value based on the face image data received from the user terminal 20, and the feature amount extraction means 14 (STEP22). At this time, the feature amount of the user's face image extracted by the feature amount extracting means 14 may differ depending on the type of the authentication engines 30A, 30B, 30C, and the feature amount extracting means 14 uses the authentication engines 30A, 30B, The feature amount of the user's face image is extracted every 30C. Further, the feature amount extracting means 14 extracts only the feature amount of the user's face image corresponding to the authentication engines 30A, 30B, 30C of the service organization selected in advance by the user. Then, the processor 12 registers the feature amounts of the face images corresponding to the authentication engines 30A, 30B, and 30C extracted by the feature amount extraction means 14 into the memory 16 of the face recognition system 10 by the registration means 13 (user ID (user's). Identification information) and the identification information of the authentication engines 30A, 30B, and 30C are stored in association with the identification information (STEP23). In this way, the registration of the user's face image captured by the user terminal 20 is completed. In the example shown in FIG. 4, the information related to the feature amount of the user's face image stored in the memory 16 is transmitted from the face recognition system 10 to the authentication engines 30A, 30B, and 30C of the service organization selected by the user. Not done.

Next, the process of authenticating the user in each of the authentication engines 30A, 30B, and 30C will be described. When user authentication is required at the service organization where each authentication engine 30A, 30B, 30C is installed, first, the user is imaged by the image pickup units 38A, 38B, 38C of the authentication engines 30A, 30B, 30C. Acquires the user's face image data by (STEP31). Further, in the authentication engines 30A, 30B, 30C, the processors 32A, 32B, 32C extract the feature amount of the user's face image as a hash value by the feature amount extracting means 34A, 34B, 34C based on the acquired face image data. (STEP32). Then, the processors 32A, 32B, 32C perform face recognition of the feature amount (specifically, the hash value) of the user's face image extracted by the feature amount extraction means 34A, 34B, 34C by the communication units 42A, 42B, 42C. It is transmitted to the processor 12 of the system 10 (STEP 33). When the processor 12 of the face recognition system 10 receives information related to the feature amount of the user's face image from the authentication engines 30A, 30B, 30C, the authentication of the service organization selected by the user, which is stored in the memory 16. It accepts information related to the feature amount of the user's facial image transmitted only from the engines 30A, 30B, and 30C. Then, in the face recognition system 10, the processor 12 receives the feature amount (specifically, the hash value) of the user's face image received from the authentication engines 30A, 30B, and 30C, and the user's face stored in the memory 16. The user is authenticated by comparing the feature amount (specifically, the hash value) of the image with the authentication means 15 (STEP34). More specifically, the matching rate between the feature amount of the user's face image received from the authentication engines 30A, 30B, 30C and the feature amount of the user's face image stored in the memory 16 is a predetermined threshold value (for example, If it exceeds 80%), the authentication means 15 authenticates the user. As described above, the memory 16 stores the user's identification information and the authentication engines 30A, 30B, and 30C registered in advance in association with the feature amount of the user's face image.

Then, when the user is authenticated by the authentication means 15 (“YES” in STEP 35), the information related to the authentication result is transmitted from the processor 12 of the face recognition system 10 to the authentication engines 30A, 30B, and 30C by the communication unit 18. (STEP36). Then, the processors 32A, 32B, 32C of the authentication engines 30A, 30B, 30C that have received the information related to the authentication result can provide the services corresponding to the authentication engines 30A, 30B, 30C by the respective processing units 40A, 40B, 40C. (STEP37). After that, the processors 32A, 32B, 32C of the authentication engines 30A, 30B, 30C transmit the information related to the usage status of the service to the face recognition system 10. As a result, in the face recognition system 10, information relating to the usage status of the service in each service institution is stored in the memory 16 for each user.

On the other hand, the feature amount of the user's face image received from the authentication engines 30A, 30B, and 30C does not substantially match the feature amount of the user's face image stored in the memory 16, and the user is authenticated by the authentication means 15. Even if this is not possible (“NO” in STEP 35), the information related to the authentication result is transmitted from the processor 12 of the face recognition system 10 to the authentication engines 30A, 30B, and 30C by the communication unit 18 (STEP 38). Then, the processors 32A, 32B, and 32C of the authentication engines 30A, 30B, and 30C that have received the information related to the authentication result cannot provide the services corresponding to the authentication engines 30A, 30B, and 30C by the processing units 40A, 40B, and 40C. (STEP39). In this case as well, the processors 32A, 32B, and 32C of the authentication engines 30A, 30B, and 30C use the information related to the service usage status (specifically, the user tries to authenticate with the authentication engines 30A, 30B, and 30C, but the authentication is performed. Information that the service could not be used without being used) is transmitted to the face authentication system 10.

According to such an authentication method, the user's facial image features are registered in the authentication engine of the service organization (for example, the authentication engine 30A), but the authentication engine of another service organization (for example, the authentication engine 30B). ), The feature amount of the user's face image is stored in the memory 16 of the face recognition system 10 in association with the identification information of the authentication engines 30A, 30B, and 30C, so that the user's face image is stored. Even in a service institution in which the feature amount of the above is not registered, the user can be authenticated by capturing the user's face image by the imaging unit. In this case, since it is not necessary for the user to register the face image data in all of the plurality of authentication engines, the user's trouble can be saved.

Also. Further examples of the face recognition system 10 and the face recognition method according to the present embodiment will be described with reference to FIGS. 7 to 9. Regarding the face recognition system 10 and the authentication engines 30A, 30B and 30C shown in FIG. 7, the same reference reference numerals are provided for the same components as the face recognition system 10 and the authentication engines 30A, 30B and 30C shown in FIGS. 1 and 4. Is added to omit the description.

As shown in FIG. 7, the processor 12 of the face authentication system 10 has a hash value of the feature amount of the user's face image corresponding to each authentication engine 30A, 30B, 30C based on the face image data received from the user terminal 20. The feature amount extraction means 14 to be extracted as the above, the registration means 13 to register the feature amount of the extracted user's face image, and the user's face image captured by the imaging units 38A to 38C of each authentication engine 30A, 30B, 30C. It has an authentication means 15 that authenticates a user based on the feature amount (specifically, a hash value) of the above. By executing the program stored in the memory 16, the processor 12 obtains the feature amount (specifically, the hash value) of the user's face image extracted by the feature amount extracting means 14 in the registration means 13 from the user. The identification information of the above and the identification information of the authentication engines 30A, 30B, and 30C are associated with each other and stored in the memory 16. Further, the processor 12 executes a program stored in the memory 16 to perform a feature amount of the user's face image captured by the image pickup units 38A to 38C of the authentication engines 30A, 30B, and 30C in the authentication means 15. (Specifically, the hash value) is compared with the feature amount (specifically, the hash value) of the user's face image corresponding to each of the authentication engines 30A, 30B, and 30C stored in the memory 16. Therefore, the user is authenticated. The program executed by the processor 12 is not limited to the one stored in the memory 16. The feature amount extraction means 14 and the registration means are executed by the processor 12 executing a program transmitted from an external device to the face recognition system 10 or a program stored in a recording medium detachably attached to the face recognition system 10. Processing may be performed in each of the 13 and the authentication means 15.

Each of the authentication engines 30A, 30B, 30C arranged in each service organization is composed of, for example, a computer or the like, and each of the authentication engines 30A, 30B, 30C has processors 32A, 32B, 32C such as a CPU, and memory 36A, It has 36B, 36C, imaging units 38A, 38B, 38C, processing units 40A, 40B, 40C, and communication units 42A, 42B, 42C. The processors 32A, 32B, 32C use feature amount extraction means 34A, 34B, 34C for extracting the feature amount of the user's face image as a hash value based on the user's face image data imaged by the imaging units 38A, 38B, 38C. Have. In the example shown in FIG. 7, the processors 32A, 32B, 32C do not have the feature amount extraction means 34A, 34B, 34C and the authentication means 35A, 35B, 35C as shown in FIG. Further, in the example shown in FIG. 7, the features of the user's face image are not stored in the memories 36A, 36B, and 36C.

In the example shown in FIG. 7, the face recognition system 10 and the authentication engines 30A, 30B, and 30C arranged in each service organization are API-linked (application programming interface). As a result, the systems of the authentication engines 30A, 30B, and 30C can be easily configured as compared with the systems of the authentication engines 30A, 30B, and 30C that are independently configured.

Next, the processing contents when the user is authenticated by the face authentication system 10 and the authentication engines 30A, 30B, and 30C as shown in FIG. 7 will be described with reference to FIGS. 8 and 9.

First, a process in which the user registers the face image data in the face authentication system 10 using the user terminal 20 will be described. Since the specific method for the user to capture the face image by the user terminal 20 has already been described, the specific method will be omitted here. When the user first registers the user with such a face recognition application on the user terminal 20, various information input on the user registration screen and the user's face image data are transmitted from the user terminal 20 to the face recognition system 10. .. In this way, the processor 12 of the face recognition system 10 receives the face image data from the user terminal 20 (STEP 41). Further, the identification information of the user terminal 20 and the registration information such as the user's name, date of birth, telephone number, e-mail address, password, etc. input to the user terminal 20 can also be obtained from the user terminal 20 by the processor 12 of the face recognition system 10. receive. Further, the processor 12 issues a user ID as user identification information, and stores the issued user ID in the memory 16.

Next, the processor 12 of the face authentication system 10 uses the feature amount of the face image corresponding to each authentication engine 30A, 30B, 30C as a hash value based on the face image data received from the user terminal 20, and the feature amount extraction means 14 (STEP 42). At this time, the feature amount of the user's face image extracted by the feature amount extracting means 14 may differ depending on the type of the authentication engines 30A, 30B, 30C, and the feature amount extracting means 14 uses the authentication engines 30A, 30B, The feature amount of the user's face image is extracted every 30C. Further, the feature amount extracting means 14 extracts only the feature amount of the user's face image corresponding to the authentication engines 30A, 30B, 30C of the service organization selected in advance by the user. Then, the processor 12 registers the feature amounts of the face images corresponding to the authentication engines 30A, 30B, and 30C extracted by the feature amount extraction means 14 into the memory 16 of the face recognition system 10 by the registration means 13 (user ID (user's). Identification information) and storage in association with the identification information of the authentication engines 30A, 30B, 30C (STEP43). In this way, the registration of the user's face image captured by the user terminal 20 is completed. In the example shown in FIG. 7, the information related to the feature amount of the user's face image stored in the memory 16 is transmitted from the face recognition system 10 to the authentication engines 30A, 30B, and 30C of the service organization selected by the user. Not done.

Next, the process of authenticating the user in each of the authentication engines 30A, 30B, and 30C will be described. When user authentication is required at the service organization where each authentication engine 30A, 30B, 30C is installed, first, the user is imaged by the image pickup units 38A, 38B, 38C of the authentication engines 30A, 30B, 30C. Acquires the user's face image data by (STEP51). The processors 32A, 32B, 32C of the authentication engines 30A, 30B, 30C transmit the user's face image data imaged by the image pickup units 38A, 38B, 38C to the processor 12 of the face recognition system 10 by the communication units 42A, 42B, 42C. (STEP52). When the processor 12 of the face recognition system 10 receives the user's face image data from the authentication engines 30A, 30B, 30C, the authentication engines 30A, 30B, of the service organization selected by the user, stored in the memory 16. The user's face image data transmitted only from 30C is accepted. Then, in the face recognition system 10, the processor 12 extracts the feature amount of the user's face image as a hash value by the feature amount extracting means 14 based on the face image data received from the authentication engines 30A, 30B, and 30C (STEP53). .. Then, the processor 12 has a feature amount (specifically, a hash value) of the user's face image extracted by the feature amount extracting means 14 and a feature amount (specifically, a hash value) of the user's face image stored in the memory 16. The user is authenticated by comparing the hash value) with the authentication means 15 (STEP 54). More specifically, the matching rate between the feature amount of the user's face image extracted by the feature amount extracting means 14 and the feature amount of the user's face image stored in the memory 16 is a predetermined threshold value (for example, 80). %), The authentication means 15 authenticates the user. As described above, the memory 16 stores the user's identification information and the authentication engines 30A, 30B, and 30C registered in advance in association with the feature amount of the user's face image.

Then, when the user is authenticated by the authentication means 15 (“YES” in STEP 55), the information related to the authentication result is transmitted from the processor 12 of the face recognition system 10 to the authentication engines 30A, 30B, and 30C by the communication unit 18. (STEP56). Then, the processors 32A, 32B, 32C of the authentication engines 30A, 30B, 30C that have received the information related to the authentication result can provide the services corresponding to the authentication engines 30A, 30B, 30C by the respective processing units 40A, 40B, 40C. (STEP57). After that, the processors 32A, 32B, 32C of the authentication engines 30A, 30B, 30C transmit the information related to the usage status of the service to the face recognition system 10. As a result, in the face recognition system 10, information relating to the usage status of the service in each service institution is stored in the memory 16 for each user.

On the other hand, the feature amount of the user's face image received from the authentication engines 30A, 30B, and 30C does not substantially match the feature amount of the user's face image stored in the memory 16, and the user is authenticated by the authentication means 15. Even if this is not possible (“NO” in STEP 55), the information related to the authentication result is transmitted from the processor 12 of the face recognition system 10 to the authentication engines 30A, 30B, and 30C by the communication unit 18 (STEP 58). Then, the processors 32A, 32B, and 32C of the authentication engines 30A, 30B, and 30C that have received the information related to the authentication result cannot provide the services corresponding to the authentication engines 30A, 30B, and 30C by the processing units 40A, 40B, and 40C. (STEP 59). In this case as well, the processors 32A, 32B, and 32C of the authentication engines 30A, 30B, and 30C use the information related to the service usage status (specifically, the user tries to authenticate with the authentication engines 30A, 30B, and 30C, but the authentication is performed. Information that the service could not be used without being used) is transmitted to the face authentication system 10.

According to such an authentication method, the user's facial image features are registered in the authentication engine of the service organization (for example, the authentication engine 30A), but the authentication engine of another service organization (for example, the authentication engine 30B). ), The feature amount of the user's face image is stored in the memory 16 of the face recognition system 10 in association with the identification information of the authentication engines 30A, 30B, and 30C, so that the user's face image is stored. Even in a service institution in which the feature amount of the above is not registered, the user can be authenticated by capturing the user's face image by the imaging unit. In this case, since it is not necessary for the user to register the face image data in all of the plurality of authentication engines, the user's trouble can be saved.

Next, in the face authentication system 10 as shown in FIGS. 1 to 9, a process when the user wants to increase the number of service institutions to which the above-mentioned face authentication service is applied will be described. After the user has performed the above-mentioned user registration, when the user presses the "Add Service" button on the initial screen of the user terminal 20 as shown in FIG. 10, a list of service institutions as shown in FIG. 13 is displayed. Will be done. The list of such service organizations is registered in the face recognition system 10 in advance. On the screen as shown in FIG. 13, the icon of the unregistered service and the icon of the registered service are displayed in different colors, the icon of the unregistered service is displayed lightly, and the icon of the unregistered service is displayed. It is displayed separately from the registered service icon. Then, when the user presses the icon of the unregistered service with a finger, the terms of the service to be added are displayed. When the user inputs an instruction to agree to the terms of service, the unregistered service becomes a registered service. At this time, in the face recognition system 10, the service institution related to this registered service is associated with the user ID (user identification information) and stored in the memory 16. After that, the user terminal 20 displays an imaging screen of a face image as shown in FIG. When the user captures a face image with the user terminal 20 on such an imaging screen, the user's face image data is transmitted from the user terminal 20 to the face recognition system 10. In this way, the processor 12 of the face recognition system 10 receives the face image data from the user terminal 20. Then, the processor 12 of the face recognition system 10 extracts the feature amount of the face image corresponding to the authentication engine of the new service organization as a hash value by the feature amount extracting means 14 based on the face image data received from the user terminal 20. To do. The processor 12 displays the feature amount of the face image corresponding to the new authentication engine extracted by the feature amount extracting means 14 into the memory 16 of the face recognition system 10 by the registration means 13 as a user ID (user identification information) and a new one. It is stored in association with the identification information of the authentication engine of the service organization. In this way, the registration of the new service institution is completed. When the user's face image data is stored in the memory 16 as described later, when the user registers a new service institution on the user terminal 20, the user terminal 20 captures the user's face image. Even if it is not, the feature amount of the face image corresponding to the authentication engine of the new service organization may be extracted by the feature amount extracting means 14 as a hash value based on the face image data of the user stored in the memory 16. .. Further, the user may be able to delete the registered service on the screen as shown in FIG. In this case, the registered service becomes an unregistered service. In addition, the information related to the feature amount of the user's face image corresponding to the authentication engine of the service organization related to the deleted registered service stored in the memory 16 is deleted.

Further, after the user performs the above-mentioned user registration, when the user presses the "authentication history" button on the initial screen of the user terminal 20 as shown in FIG. 10 with a finger, the past authentication history as shown in FIG. 14 A list of information (in other words, a list of information related to the usage status of the service organization by the user) is displayed on the user terminal 20. More specifically, the memory 16 of the face recognition system 10 stores the user's past authentication history information in association with the user's identification information. Then, when the "authentication history" button is pressed by the user on the initial screen of the user terminal 20 as shown in FIG. 10, the user terminal 20 displays the user's past authentication history information from the processor 12 of the face recognition system 10. The desired signal is transmitted. When the processor 12 of the face recognition system 10 receives a signal for requesting the user's past authentication history information from the user terminal 20, the processor 12 of the face recognition system 10 receives the past authentication history information corresponding to the user's identification information stored in the memory 16 in the communication unit. It is transmitted to the user terminal 20 by 18. As a result, the user terminal 20 will display a list of past authentication history information of this user. Further, in the list of past authentication history information as shown in FIG. 14, when the user presses the display of a certain authentication history with a finger, the details of the authentication history are displayed on the user terminal 20 as shown in FIG. Will be.

Further, in the present embodiment, instead of the user registering the feature amount of the user's face recognition in the face recognition system 10 using the user terminal 20, the user uses some authentication engines 30A, 30B, and 30C for face recognition. The feature amount of the user's face recognition in the system 10 may be registered. Specifically, when the user inputs registration information in a certain authentication engine 30A, 30B, 30C and images the user's face image by the imaging units 38A, 38B, 38C, the input registration information and the user's face image data are obtained. It is transmitted from the authentication engines 30A, 30B, and 30C to the face recognition system 10. In this way, the processor 12 of the face recognition system 10 receives the face image data from the authentication engines 30A, 30B, and 30C. Further, the processor 12 issues a user ID as user identification information based on the registration information received from the authentication engines 30A, 30B, and 30C, and stores the issued user ID in the memory 16.

Next, the processor 12 of the face recognition system 10 features the feature amount of the face image corresponding to each authentication engine 30A, 30B, 30C as a hash value based on the face image data received from the authentication engines 30A, 30B, 30C. It is extracted by the quantity extraction means 14. At this time, the feature amount extracting means 14 extracts the feature amount of the user's face image for each of the authentication engines 30A, 30B, and 30C. Further, the feature amount extracting means 14 extracts only the feature amount of the user's face image corresponding to the authentication engines 30A, 30B, 30C of the service organization selected in advance by the user. For example, the service institution where the authentication engine 30A is installed allows the user to use the face image data, but the service institution where the authentication engine 30B is installed does not allow the user to use the face image data. In this case, the feature amount extracting means 14 extracts only the feature amount of the user's face image corresponding to the authentication engine 30A. As described above, the memory 16 stores the user's identification information in association with the service institution selected by the user. Then, the processor 12 registers the feature amounts of the face images corresponding to the authentication engines 30A, 30B, and 30C extracted by the feature amount extraction means 14 into the memory 16 of the face recognition system 10 by the registration means 13 (user ID (user's). Identification information) and the identification information of the authentication engines 30A, 30B, and 30C are stored in association with the identification information. In this way, the registration of the user's face image captured by the imaging units 38A, 38B, 38C of the authentication engines 30A, 30B, 30C is completed. In such an embodiment, the user can register the feature amount of the user's face authentication in the face authentication system 10 without using the user terminal 20.

According to the face authentication method performed by the face authentication system 10 according to the present embodiment having the above configuration, the user's face image data and the user's identification information are received, and the received user's face image data is used. Based on this, the feature amount of the user's facial image is extracted for each of the plurality of authentication engines 30A, 30B, and 30C. Then, the information related to the feature amount of the user's face image for each of the plurality of extracted authentication engines 30A, 30B, and 30C is stored in the memory 16 in association with the received user's identification information. According to such a face recognition method, face recognition can be easily performed by a plurality of authentication engines 30A, 30B, and 30C.

Specifically, it is generally widespread to provide an authentication engine using face recognition even in the past, but in the face recognition system of the prior art, face recognition is performed by an authentication engine that does not register face data. I couldn't. Therefore, in order to authenticate with each of the plurality of authentication engines, the user needs to register the face data with each of the plurality of authentication engines, which is troublesome and stressful. On the other hand, in the present embodiment, the feature amount of the user's face image is registered by the authentication engine of the service organization (for example, the authentication engine 30A), but the authentication engine of another service organization (for example, the authentication engine). Even if it is not registered in 30B), the feature amount of the user's face image is stored in the memory 16 of the face authentication system 10 for each authentication engine 30A, 30B, 30C. Even in the authentication engines 30A, 30B, 30C of the service organization in which the user is not registered, the user can be authenticated by capturing the user's face image with the image pickup units 38A, 38B, 38C. In this case, since it is not necessary for the user to register the face image data in all of the plurality of authentication engines 30A, 30B, and 30C, the user's labor can be saved. As a result, it is troublesome to register the face image data, and it is possible to encourage users who have not used various services using the authentication engine that does not register the face image data to feel free to use the various services. ..

Further, when the feature amount of the user's face image such as a hash value is extracted and stored in the memory 16, the user's face image data itself is stored in the memory 16 as compared with the case where the user's face image data itself is stored in the memory 16. The amount of data can be significantly reduced. This is because the amount of data of the feature amount of the user's face image is smaller than the amount of data of the user's face image data itself. Further, in this case, since the user's face image data itself is not stored in the memory 16, the privacy of the user can be protected more reliably. Further, when information related to the feature amount of the user's face image is sent between the face recognition system 10 and each authentication engine 30A, 30B, 30C, the user's face image data is sent to the face recognition system 10 and each authentication engine 30A. The amount of data communication can be significantly reduced as compared with the case of sending between the 30B and 30C.

Further, in the face authentication method of the present embodiment, as described above, when extracting the feature amount of the user's face image based on the received user's face image data, a plurality of feature amounts of the user's face image are used. Is extracted for each authentication engine 30A, 30B, 30C, and the feature amount of the user's face image for each of the extracted plurality of authentication engines 30A, 30B, 30C is stored in the memory in association with the user's identification information by the registration means 13. Let me. At this time, it is possible to deal with the case where the feature amount of the user's face image extracted by the feature amount extracting means 14 differs depending on the types of the authentication engines 30A, 30B, and 30C. When the same program is used by a plurality of authentication engines 30A, 30B, 30C and the feature amount of the user's face image extracted by the feature amount extracting means 14 is common to the authentication engines 30A, 30B, 30C, It is not necessary to extract the feature amount of the user's face image for each of a plurality of authentication engines 30A, 30B, and 30C.

Further, in the face authentication method of the present embodiment, as described above, the memory 16 stores information related to the service institution selected by the user in association with the user's identification information, and the received user's face authentication method is stored. Based on the face image data, the feature amount of the user's face image is selected by the user stored in the memory 16 in the step of extracting the feature amount of the user's face image for each of the plurality of authentication engines 30A, 30B, and 30C by the feature amount extracting means 14. The feature amount extracting means 14 extracts the feature amount of the user's face image corresponding only to the authentication engine of the service organization. In this case, since the feature amount of the user's face image corresponding to the authentication engines 30A, 30B, 30C of the service organization not selected by the user is not stored in the memory 16, the security can be improved and the user can be used. You can rest assured. In addition, companies and the like participating in the group of authentication engines managed by the face recognition system 10 can be authenticated by a large number of authentication engines simply by registering face image data on the user terminal 20 or the like. It will be possible to appeal the improvement of sex to customers.

Further, in the face authentication method of the present embodiment, as described above, the feature amount extracting means 14 obtains a hash value obtained from the received user's face image data by a predetermined hash function as the feature of the user's face image. The amount is extracted for each of a plurality of authentication engines 30A, 30B, and 30C. In this case, since the hash value is stored in the memory 16 as the feature amount of the user's face image, the amount of data stored in the memory 16 is reduced as compared with the case where the face image data itself is stored in the memory 16. be able to. Further, it is difficult to restore or infer the face image data from the hash value, and since only the hash value is stored in the face recognition system 10, the privacy and security of the user can be enhanced.

In the present embodiment, the feature amount extracting means 14 extracts the feature amount of the user's face image as a hash value for each of a plurality of authentication engines 30A, 30B, and 30C based on the received user's face image data. It is not limited to that. The feature amount extracting means 14 extracts the feature amount of the user's face image for each of a plurality of authentication engines 30A, 30B, and 30C with a value different from the hash value based on the received user's face image data. It may be. For example, as a feature amount of the user's face image, the relative position, size, shape, etc. of each part of the face (eyes, nose, ears, etc.) may be extracted as a feature. Even in this case, if the amount of data of a value other than the hash value is smaller than the amount of data of the face image itself, the amount of data stored in the memory 16 can be reduced.

Further, in the face authentication method of the present embodiment, as described above, based on the received user's face image data, the feature amount of the user's face image is set for each of the plurality of authentication engines 30A, 30B, and 30C. In the step of extracting by the extraction means 14, the feature amount of the user's face image is characterized for each of the plurality of authentication engines 30A, 30B, and 30C based on the user's face image data transmitted from the user terminal 20 possessed by the user. It is extracted by the quantity extraction means 14. Alternatively, in the step of extracting the feature amount of the user's face image by the feature amount extracting means 14 for each of the plurality of authentication engines 30A, 30B, 30C based on the received user's face image data, the plurality of authentication engines 30A, Based on the user's face image data transmitted from a certain authentication engine among the 30B and 30C, the feature amount of the user's face image is extracted by the feature amount extracting means 14 for each of the plurality of authentication engines 30A, 30B and 30C. May be good.

Further, in the present embodiment, a program executed by the processor 12 for performing the face recognition method by the face recognition system 10 and a recording medium on which this program is recorded are used. Here, the processor 12 receives the user's face image data and the user's identification information by executing the program, and based on the received user's face image data, the feature amount of the user's face image is calculated by the plurality of authentication engines 30A. , 30B, 30C. Then, the information related to the feature amount of the user's face image for each of the plurality of extracted authentication engines 30A, 30B, and 30C is stored in the memory 16 in association with the received user's identification information. According to such a program and a recording medium, face recognition can be easily performed by a plurality of authentication engines 30A, 30B, and 30C.

Further, in the present embodiment, the face recognition system 10 provided with the processor 12 is used. In such a face recognition system 10, the processor 12 receives the user's face image data and the user's identification information by executing the program, and the feature of the user's face image is based on the received user's face image data. The amount is extracted for each of the plurality of authentication engines 30A, 30B, and 30C. Then, the information related to the feature amount of the user's face image for each of the plurality of extracted authentication engines 30A, 30B, and 30C is stored in the memory 16 in association with the received user's identification information. According to such a face recognition system 10, face recognition can be easily performed by a plurality of authentication engines 30A, 30B, and 30C.

The face recognition method and face recognition system according to the present invention are not limited to the above-described aspects, and various changes can be made.

For example, a face recognition system as shown in FIG. 16 may be used. The face recognition system 50 shown in FIG. 16 has a processor 52, a first server 56, a second server 58, and a third server 60. Here, the servers 56, 58, and 60 correspond to the authentication engines 30A, 30B, and 30C. Specifically, the first server 56 corresponds to the authentication engine 30A, the second server 58 corresponds to the authentication engine 30B, and the third server 60 corresponds to the authentication engine 30C. Further, in FIG. 16, three authentication engines 30A, 30B, and 30C are illustrated, but when two or four or more authentication engines are used, the server is connected to the face recognition system 10 corresponding to each authentication engine. Provided. Further, the processor 52 includes a server management means 52a, a feature amount extraction means 54, a registration means 53, and an authentication means 55. The feature amount extraction means 54, the registration means 53, and the authentication means 55 of the processor 52 are the feature amount extraction means 14, the registration means 13, and the authentication means 15 of the processor 12 of the face recognition system 10 shown in FIGS. 1, 4, and 7. It has almost the same function. The server management means 52a manages the servers 56, 58, and 60, respectively.

In each of the servers 56, 58, and 60, the feature amount of the data of the user's face image corresponding to each of the authentication engines 30A, 30B, and 30C is stored in association with the user's identification information. Further, the servers 56, 58, and 60 are communicably connected to the corresponding authentication engines 30A, 30B, and 30C via a network such as an Internet line. Each of the servers 56, 58, and 60 is configured to be collectively manageable by an API (application programming interface) by the server management means 52a. As a result, the system of each server 56, 58, 60 can be easily configured as compared with the system (program) of each server 56, 58, 60 which is independently configured. Further, the server management means 52a and the service organization in which the authentication engines 30A, 30B, and 30C are installed are linked by API. For example, by opening the platform of the server management means 52a to each service institution and using the same platform as the server management means 52a in each service institution, it is easy to construct a system for providing services in each service institution. can do.

The face recognition system 50 having such a processor 52 and the servers 56, 58, and 60 can perform the same processing as the face recognition system 10 having the processor 12. That is, according to the face recognition method performed by the face recognition system 50 shown in FIG. 16, the processor 52 determines the feature amount of the user's face image based on the received user's face image data, respectively, in the authentication engines 30A and 30B. The feature amount of the user's face image extracted by the feature amount extracting means 54 for each 30C is associated with the user's identification information and the authentication engines 30A, 30B, 30C by the registration means 53, and the servers 56, 58, 60 are associated with each other. To memorize.

Further, the information related to the feature amount of the user's face image stored in each server 56, 58, 60 is transmitted from each server 56, 58, 60 to the corresponding authentication engines 30A, 30B, 30C. As a result, when the user is authenticated by each of the authentication engines 30A, 30B, 30C, the user's face image data is acquired by imaging the user with the imaging units 38A, 38B, 38C, and the acquired face image data is acquired. By extracting the feature amount of the user's face image by the feature amount extracting means 34A, 34B, 34C based on the above, the user can be authenticated by each authentication engine 30A, 30B, 30C. After that, the processors 32A, 32B, 32C of the authentication engines 30A, 30B, 30C transmit the information related to the usage status of the service to the face recognition system 10. As a result, in the face recognition system 50, information related to the service usage status in each service institution is stored in each server 56, 58, 60 for each user.

As another method for authenticating the user, when the user is authenticated by each of the authentication engines 30A, 30B, 30C, the user's face image data is acquired by imaging the user with the imaging units 38A, 38B, 38C. Based on the acquired face image data, the feature amount of the user's face image is extracted by the feature amount extraction means 34A, 34B, 34C. Then, the authentication engines 30A, 30B, and 30C transmit information related to the feature amount of the user's face image to the face recognition system 50. As a result, the user can be authenticated by the authentication means 55. After that, the user authentication result is transmitted from the face authentication system 50 to the original authentication engines 30A, 30B, 30C.

As yet another method for authenticating the user, when the user is authenticated by each of the authentication engines 30A, 30B, 30C, the user's face image data is acquired by imaging the user with the imaging units 38A, 38B, 38C. .. Then, the user's face image data is transmitted from the authentication engines 30A, 30B, and 30C to the face recognition system 50. Then, the processor 52 extracts the feature amount of the user's face image by the feature amount extracting means 54 based on the user's face image data transmitted to each of the servers 56, 58, and 60, and the extracted user's face image. The feature amount of the user's face image stored in each of the servers 56, 58, and 60 is compared with the feature amount of the user's face image by the authentication means 55. As a result, the user can be authenticated by the authentication means 55. After that, the user authentication result is transmitted from the face authentication system 50 to the original authentication engines 30A, 30B, 30C.

According to these face recognition methods, face recognition can be easily performed by a plurality of authentication engines 30A, 30B, and 30C.

Further, in the face recognition system 10 as shown in FIGS. 1, 4, 7, and the like, the processor 12 stores the user's face image data itself transmitted from the user terminal 20 and the authentication engines 30A, 30B, and 30C into the memory 16. You may memorize it.

Further, in the above example, the mode in which the face recognition systems 10 and 50 are installed separately from the authentication engines 30A, 30B, and 30C installed in each service institution is shown, but the authentication installed in each service institution is shown. An authentication engine among the engines 30A, 30B, and 30C may also have the functions of the face recognition systems 10 and 50. That is, the processor of the authentication engine among the authentication engines 30A, 30B, and 30C installed in each service organization is the feature amount extraction means 14 in the processor 12 of the face recognition system 10 shown in FIGS. 1, 4, 7, and the like. It may have a feature amount extraction means, a registration means, and an authentication means having the same functions as the registration means 13 and the authentication means 15.

10 Face authentication system 12 Processor 13 Registration means 14 Feature quantity extraction means 15 Authentication means 16 Memory 18 Communication unit 20 User terminals 30A, 30B, 30C Authentication engine 32A, 32B, 32C Processor 34A, 34B, 34C Feature quantity extraction means 35A, 35B , 35C Authentication means 36A, 36B, 36C Memory 38A, 38B, 38C Imaging unit 40A, 40B, 40C Processing unit 42A, 42B, 42C Communication unit 50 Face authentication system 52 Processor 52a Server management means 53 Registration means 54 Feature extraction means 55 Authentication means 56 1st server 58 2nd server 60 3rd server

Claims (12)

It is an information processing method performed by the face recognition system.
The process of accepting the user's face image data and
A process of extracting the feature amount of this user's face image for each of a plurality of authentication engines based on the received user's face image data, and
A process of associating the extracted information related to the feature amount of the user's face image for each authentication engine with the user's identification information and storing it in the memory.
A step of transmitting information related to a feature amount of a user's face image stored in the memory to each of the authentication engines, and
With
In the step of extracting the feature amount of the user's face image for each of the plurality of authentication engines based on the received user's face image data, even if the user's face image data is the same, the user's face image depends on the type of the authentication engine. Extract the feature amount of the user's face image so that the feature amount of the face image is different,
Information related to the feature amount of the user's face image corresponding to the authentication engine is transmitted to each of the authentication engines.
Information related to the service institution selected by the user is stored in the memory in association with the user's identification information.
When extracting the feature amount of this user's face image for each of a plurality of authentication engines based on the received user's face image data, only the authentication engine of the service organization selected by the user stored in the memory is supported. An information processing method that extracts the feature amount of the user's face image. It is a face recognition method performed by the face recognition system.
The process of accepting the user's face image data and
A process of extracting the feature amount of this user's face image for each of a plurality of authentication engines based on the received user's face image data, and
A process of associating the extracted information related to the feature amount of the user's face image for each authentication engine with the user's identification information and storing it in the memory.
The process of receiving information related to the feature amount of the user's face image from the authentication engine, and
A step of authenticating a user by comparing the feature amount of the received user's face image with the feature amount of the user's face image corresponding to the authentication engine stored in the memory.
The process of transmitting the information related to the user's authentication result to the authentication engine to which the information related to the feature amount of the user's face image has been sent, and
With
In the step of extracting the feature amount of the user's face image for each of the plurality of authentication engines based on the received user's face image data, even if the user's face image data is the same, the user's face image depends on the type of the authentication engine. Extract the feature amount of the user's face image so that the feature amount of the face image is different,
Information related to the service institution selected by the user is stored in the memory in association with the user's identification information.
When receiving information related to the feature amount of the user's face image from the authentication engine, it relates to the feature amount of the user's face image transmitted only from the authentication engine of the service organization selected by the user stored in the memory. A face recognition method that accepts information. It is a face recognition method performed by the face recognition system.
The process of accepting the user's face image data and
A process of extracting the feature amount of this user's face image for each of a plurality of authentication engines based on the received user's face image data, and
A process of associating the extracted information related to the feature amount of the user's face image for each authentication engine with the user's identification information and storing it in the memory.
The process of accepting the user's face image data from the authentication engine,
Based on the received user's face image data, the process of extracting the feature amount of the user's face image corresponding to the authentication engine to which the user's face image data has been sent, and
A step of authenticating a user by comparing the extracted feature amount of the user's face image with the feature amount of the user's face image corresponding to the authentication engine stored in the memory.
The process of transmitting information related to the user's authentication result to the authentication engine to which the user's face image data has been sent, and
With
In the step of extracting the feature amount of the user's face image for each of the plurality of authentication engines based on the received user's face image data, even if the user's face image data is the same, the user's face image depends on the type of the authentication engine. Extract the feature amount of the user's face image so that the feature amount of the face image is different,
Information related to the service institution selected by the user is stored in the memory in association with the user's identification information.
A face recognition method that accepts the user's face image data transmitted only from the authentication engine of the service organization selected by the user stored in the memory when receiving the information related to the user's face image data from the authentication engine. .. It is a program executed by a processor to perform information processing methods by a face recognition system.
When the processor executes the program,
Accepts user's face image data,
Based on the received user's face image data, the feature amount of this user's face image is extracted for each of a plurality of authentication engines.
Information related to the feature amount of the user's face image for each of the extracted multiple authentication engines is associated with the user's identification information and stored in the memory.
Information related to the feature amount of the user's face image stored in the memory is transmitted to each authentication engine.
When the feature amount of the user's face image is extracted for each of the plurality of authentication engines based on the received user's face image data, even if the user's face image data is the same, the user's face image depends on the type of the authentication engine. Extract the feature amount of the user's face image so that the feature amount of the face image is different,
Information related to the feature amount of the user's face image corresponding to the authentication engine is transmitted to each of the authentication engines.
Information related to the service institution selected by the user is stored in the memory in association with the user's identification information.
When extracting the feature amount of this user's face image for each of a plurality of authentication engines based on the received user's face image data, only the authentication engine of the service organization selected by the user stored in the memory is supported. A program that extracts the features of the user's face image. A program executed by a processor to perform a face recognition method by a face recognition system.
When the processor executes the program,
Accepts user's face image data,
Based on the received user's face image data, the feature amount of this user's face image is extracted for each of a plurality of authentication engines.
Information related to the feature amount of the user's face image for each of the extracted multiple authentication engines is associated with the user's identification information and stored in the memory.
Receives information related to the feature amount of the user's face image from the authentication engine,
The user is authenticated by comparing the feature amount of the received user's face image with the feature amount of the user's face image corresponding to the authentication engine stored in the memory.
The information related to the user's authentication result is transmitted to the authentication engine to which the information related to the feature amount of the user's face image has been sent.
When the feature amount of the user's face image is extracted for each of the plurality of authentication engines based on the received user's face image data, even if the user's face image data is the same, the user's face image depends on the type of the authentication engine. Extract the feature amount of the user's face image so that the feature amount of the face image is different,
Information related to the service institution selected by the user is stored in the memory in association with the user's identification information.
When receiving information related to the feature amount of the user's face image from the authentication engine, it relates to the feature amount of the user's face image transmitted only from the authentication engine of the service organization selected by the user stored in the memory. A program that accepts information. A program executed by a processor to perform a face recognition method by a face recognition system.
When the processor executes the program,
Accepts user's face image data,
Based on the received user's face image data, the feature amount of this user's face image is extracted for each of a plurality of authentication engines.
Information related to the feature amount of the user's face image for each of the extracted multiple authentication engines is associated with the user's identification information and stored in the memory.
Accepting the user's face image data from the authentication engine,
Based on the received user's face image data, the feature amount of the user's face image corresponding to the authentication engine to which the user's face image data is sent is extracted.
The user is authenticated by comparing the extracted feature amount of the user's face image with the feature amount of the user's face image corresponding to the authentication engine stored in the memory.
Information related to the user's authentication result is transmitted to the authentication engine to which the user's face image data has been sent.
When the feature amount of the user's face image is extracted for each of the plurality of authentication engines based on the received user's face image data, even if the user's face image data is the same, the user's face image depends on the type of the authentication engine. Extract the feature amount of the user's face image so that the feature amount of the face image is different,
Information related to the service institution selected by the user is stored in the memory in association with the user's identification information.
A program that accepts user face image data transmitted only from the authentication engine of a service organization selected by the user stored in the memory when receiving information related to the user's face image data from the authentication engine. A recording medium on which a program executed by a processor for performing an information processing method by a face recognition system is recorded.
When the processor executes the program,
Accepts user's face image data,
Based on the received user's face image data, the feature amount of this user's face image is extracted for each of a plurality of authentication engines.
Information related to the feature amount of the user's face image for each of the extracted multiple authentication engines is associated with the user's identification information and stored in the memory.
Information related to the feature amount of the user's face image stored in the memory is transmitted to each authentication engine.
When the feature amount of the user's face image is extracted for each of the plurality of authentication engines based on the received user's face image data, even if the user's face image data is the same, the user's face image depends on the type of the authentication engine. Extract the feature amount of the user's face image so that the feature amount of the face image is different,
Information related to the feature amount of the user's face image corresponding to the authentication engine is transmitted to each of the authentication engines.
Information related to the service institution selected by the user is stored in the memory in association with the user's identification information.
When extracting the feature amount of this user's face image for each of a plurality of authentication engines based on the received user's face image data, only the authentication engine of the service organization selected by the user stored in the memory is supported. A recording medium that extracts the feature amount of the user's facial image. A recording medium on which a program executed by a processor for performing a face recognition method by a face recognition system is recorded.
When the processor executes the program,
Accepts user's face image data,
Based on the received user's face image data, the feature amount of this user's face image is extracted for each of a plurality of authentication engines.
Information related to the feature amount of the user's face image for each of the extracted multiple authentication engines is associated with the user's identification information and stored in the memory.
Receives information related to the feature amount of the user's face image from the authentication engine,
The user is authenticated by comparing the feature amount of the received user's face image with the feature amount of the user's face image corresponding to the authentication engine stored in the memory.
The information related to the user's authentication result is transmitted to the authentication engine to which the information related to the feature amount of the user's face image has been sent.
When the feature amount of the user's face image is extracted for each of the plurality of authentication engines based on the received user's face image data, even if the user's face image data is the same, the user's face image depends on the type of the authentication engine. Extract the feature amount of the user's face image so that the feature amount of the face image is different,
Information related to the service institution selected by the user is stored in the memory in association with the user's identification information.
When receiving information related to the feature amount of the user's face image from the authentication engine, it relates to the feature amount of the user's face image transmitted only from the authentication engine of the service organization selected by the user stored in the memory. A recording medium that accepts information. A recording medium on which a program executed by a processor for performing a face recognition method by a face recognition system is recorded.
When the processor executes the program,
Accepts user's face image data,
Based on the received user's face image data, the feature amount of this user's face image is extracted for each of a plurality of authentication engines.
Information related to the feature amount of the user's face image for each of the extracted multiple authentication engines is associated with the user's identification information and stored in the memory.
Accepting the user's face image data from the authentication engine,
Based on the received user's face image data, the feature amount of the user's face image corresponding to the authentication engine to which the user's face image data is sent is extracted.
The user is authenticated by comparing the extracted feature amount of the user's face image with the feature amount of the user's face image corresponding to the authentication engine stored in the memory.
Information related to the user's authentication result is transmitted to the authentication engine to which the user's face image data has been sent.
When the feature amount of the user's face image is extracted for each of the plurality of authentication engines based on the received user's face image data, even if the user's face image data is the same, the user's face image depends on the type of the authentication engine. Extract the feature amount of the user's face image so that the feature amount of the face image is different,
Information related to the service institution selected by the user is stored in the memory in association with the user's identification information.
A recording medium that receives user face image data transmitted only from the authentication engine of a service organization selected by the user stored in the memory when receiving information related to the user's face image data from the authentication engine. An information processing system equipped with a processor
By executing the program, the processor
Accepts user's face image data,
Based on the received user's face image data, the feature amount of this user's face image is extracted for each of a plurality of authentication engines.
Information related to the feature amount of the user's face image for each of the extracted multiple authentication engines is associated with the user's identification information and stored in the memory.
Information related to the feature amount of the user's face image stored in the memory is transmitted to each authentication engine.
When the feature amount of the user's face image is extracted for each of the plurality of authentication engines based on the received user's face image data, even if the user's face image data is the same, the user's face image depends on the type of the authentication engine. Extract the feature amount of the user's face image so that the feature amount of the face image is different,
Information related to the feature amount of the user's face image corresponding to the authentication engine is transmitted to each of the authentication engines.
Information related to the service institution selected by the user is stored in the memory in association with the user's identification information.
When extracting the feature amount of this user's face image for each of a plurality of authentication engines based on the received user's face image data, only the authentication engine of the service organization selected by the user stored in the memory is supported. An information processing system that extracts the feature amount of the user's face image. A face recognition system with a processor
By executing the program, the processor
Accepts user's face image data,
Based on the received user's face image data, the feature amount of this user's face image is extracted for each of a plurality of authentication engines.
Information related to the feature amount of the user's face image for each of the extracted multiple authentication engines is associated with the user's identification information and stored in the memory.
Receives information related to the feature amount of the user's face image from the authentication engine,
The user is authenticated by comparing the feature amount of the received user's face image with the feature amount of the user's face image corresponding to the authentication engine stored in the memory.
The information related to the user's authentication result is transmitted to the authentication engine to which the information related to the feature amount of the user's face image has been sent.
When the feature amount of the user's face image is extracted for each of the plurality of authentication engines based on the received user's face image data, even if the user's face image data is the same, the user's face image depends on the type of the authentication engine. Extract the feature amount of the user's face image so that the feature amount of the face image is different,
Information related to the service institution selected by the user is stored in the memory in association with the user's identification information.
When receiving information related to the feature amount of the user's face image from the authentication engine, it relates to the feature amount of the user's face image transmitted only from the authentication engine of the service organization selected by the user stored in the memory. A face recognition system that accepts information. A face recognition system with a processor
By executing the program, the processor
Accepts user's face image data,
Based on the received user's face image data, the feature amount of this user's face image is extracted for each of a plurality of authentication engines.
Information related to the feature amount of the user's face image for each of the extracted multiple authentication engines is associated with the user's identification information and stored in the memory.
Accepting the user's face image data from the authentication engine,
Based on the received user's face image data, the feature amount of the user's face image corresponding to the authentication engine to which the user's face image data is sent is extracted.
The user is authenticated by comparing the extracted feature amount of the user's face image with the feature amount of the user's face image corresponding to the authentication engine stored in the memory.
Information related to the user's authentication result is transmitted to the authentication engine to which the user's face image data has been sent.
When the feature amount of the user's face image is extracted for each of the plurality of authentication engines based on the received user's face image data, even if the user's face image data is the same, the user's face image depends on the type of the authentication engine. Extract the feature amount of the user's face image so that the feature amount of the face image is different,
Information related to the service institution selected by the user is stored in the memory in association with the user's identification information.
A face recognition system that accepts the user's face image data transmitted only from the authentication engine of the service organization selected by the user stored in the memory when receiving the information related to the user's face image data from the authentication engine. ..

Images