JP2023132476A - Management apparatus, management method, and management program - Google Patents

Abstract

To provide a management apparatus, a management method, and a management program, configured to accurately register a face image with less burden on a user.SOLUTION: A management apparatus according to the present disclosure includes: a receiving unit which receives, from a user, a registration image which is a basis of ground truth data to be used for collation by each of multiple face authentication engines, for executing authentication processing using the face authentication engines; a quality determination unit which determines whether the registration image has quality equivalent to the ground truth data for authentication processing through the face authentication engines, for each of the face authentication engines; and a registration unit which registers, on the basis of results determined by the quality determination unit, the registration image when a determination is made that the registration image has quality equivalent to the ground truth data for authentication processing, for all the face authentication engines.SELECTED DRAWING: Figure 1

Description

The present disclosure relates to a management device, a management method, and a management program that manage authentication processing.

Facial recognition processing is known as a means for verifying an individual's identity. In the face authentication process, the user's face image is registered in advance, and the user's identity is verified by comparing the face image acquired during authentication with the registered image.

The registration of a face image in advance in the face authentication process is performed by the user himself/herself using a terminal such as a smartphone, for example. As a technology related to facial photo registration, there is a known technology that supports efficient registration by determining the suitability of the facial image and then processing the facial photo to match the medium to be registered (for example, Patent Document 1). Furthermore, in face authentication processing, a technique is known that improves authentication accuracy by using a plurality of face authentication engines (for example, Patent Document 2).

JP2016-81141A Patent No. 6839313

According to the technique disclosed in Patent Document 1, a face image to be used in a face image-attached medium can be efficiently registered.

However, the technique disclosed in Patent Document 1 may not be suitable for registering facial images used in face authentication processing. In face recognition processing, authentication is performed based on the features extracted from the face image, so when registering the face image, check whether the image is such that features can be extracted (for example, the face is located in an appropriate range in the entire image). It is necessary to determine whether the image is included in the image, whether the brightness of the image is appropriate, etc. For this reason, if you perform processing such as trimming to suit the medium after determining the suitability of a face image, depending on the degree of processing, the processed image may become an image that cannot be used for face recognition processing, and registration may not be possible. There is a possibility of failure. Retrying registration due to registration failure can increase the load on the server and burden the user.

Therefore, the present disclosure proposes a management device, a management method, and a management program that can accurately register facial images with less burden on the user.

In order to solve the above problems, a management device according to one embodiment of the present disclosure collects correct data that each of the plurality of face recognition engines uses for verification in order to perform authentication processing using the plurality of face recognition engines. a reception unit that receives the original registration image from the user; and a reception unit that determines whether the registration image has a quality that can be used as correct data for authentication processing of the plurality of face recognition engines, and each of the plurality of face recognition engines and a quality determination unit that determines the quality of the registration image, and a quality determination unit that determines that the registration image has a quality that can be used as correct data for authentication processing for all of the plurality of face recognition engines, based on the determination result by the quality determination unit, A registration unit that registers the registration image.

FIG. 3 is a diagram schematically showing the flow of registration processing according to the embodiment. FIG. 3 is a diagram schematically showing the flow of face authentication processing according to the embodiment. FIG. 2 is a diagram (1) for explaining registration processing according to the embodiment. FIG. 2 is a diagram (2) for explaining registration processing according to the embodiment. FIG. 3 is a sequence diagram showing the procedure of registration processing according to the embodiment. FIG. 3 is a sequence diagram showing the procedure of face authentication processing according to the embodiment. 2 is a flowchart illustrating a determination procedure of face authentication processing according to an embodiment. 1 is a diagram illustrating a configuration example of a management device according to an embodiment. FIG. 3 is a diagram illustrating an example of a user information storage unit according to the embodiment. It is a figure showing an example of a quality check item storage part concerning an embodiment. FIG. 3 is a diagram showing an example of a face photo storage unit according to the embodiment. It is a figure showing an example of an external service information storage part concerning an embodiment. FIG. 2 is a hardware configuration diagram showing an example of a computer that implements the functions of a management device.

Embodiments of the present disclosure will be described in detail below based on the drawings. In the following embodiments, the same parts are given the same reference numerals and redundant explanations will be omitted.

(1. Embodiment)
(1-1. An example of management processing according to the embodiment)
FIGS. 1 and 2 schematically show the flow of management processing according to the embodiment. The management process according to the embodiment is executed by the management device 100 included in the management system 1 shown in FIG. As shown in FIG. 1, the management system 1 includes a management device 100, a user terminal 20, and a face authentication provider 50.

The management device 100 is an information processing device that executes management processing according to the embodiment. For example, the management device 100 is a cloud server or the like that can communicate with the user terminal 20 and the face authentication provider 50 via a network.

The user terminal 20 is an information processing terminal used by the user 10. For example, the user terminal 20 is a smartphone, a tablet terminal, a PC (Personal Computer), or the like. It is assumed that the user terminal 20 according to the embodiment is equipped with a camera and can take a facial image of the user 10.

The face recognition provider 50 is a general term for businesses that provide face recognition engines for performing face recognition processing. The face recognition provider 50 includes, for example, a first provider that provides the first face recognition engine 60, a second provider that provides the second face recognition engine 70, a third provider that provides the third face recognition engine 80, etc. include.

The first provider, the second provider, and the third provider are businesses that provide different facial recognition engines. It is assumed that the face recognition engines provided by each provider have different characteristics in various respects, such as methods for detecting a face range in an image, extracting facial features, and matching features. That is, different facial recognition engines have different image characteristics (characteristics such as image brightness, proportion of the area occupied by the face in the image, and skin color and depth of the user to be authenticated). Note that in the following description, the first provider, the second provider, and the third provider may be collectively referred to as the face authentication provider 50 when there is no need to distinguish them.

The management device 100 according to the embodiment improves the accuracy of authentication and improves the reliability of authentication by registering and checking with these multiple facial recognition engines when the user 10 registers a facial image or performs authentication. Improve your sexuality. That is, the management device 100 plays a role as a platform for executing authentication processing using a multi-face authentication engine.

Specifically, the management device 100 receives from the user 10 a facial image that is the source of correct data that each of the plurality of face recognition engines uses for verification in order to perform authentication processing by the plurality of face recognition engines. Furthermore, the management device 100 provides processing means for the user 10 to process the received facial image. Then, the management device 100 determines whether the processed image, which is the processed face image, has a quality that can be used as correct answer data for authentication processing by a plurality of face authentication engines.

In this way, the management device 100 prevents the user 10 from having to go through the trouble of redoing the image by determining whether the image processed by the user 10 can be used with each facial recognition engine. do. In addition, the management device 100 verifies the quality of the facial image for each of the plurality of facial recognition engines, and registers the facial image only when the quality is verified to be usable for all of the facial recognition engines used for authentication. It's okay. Thereby, the management device 100 can smoothly proceed with the registration of facial images, which tends to be complicated, while performing highly accurate authentication using a plurality of facial recognition engines.

Regarding such management processing, first, using FIG. 1, the registration processing of facial images with respect to a plurality of facial recognition engines will be explained along the flow. FIG. 1 is a diagram schematically showing the flow of registration processing according to the embodiment.

In the example shown in FIG. 1, the user 10 accesses the management device 100 using the user terminal 20 and requests registration of a face image to be used in face authentication (step S1). For example, the user 10 requests registration of a facial image via a website provided by the management device 100.

Upon receiving the request, the management device 100 provides the user terminal 20 with a web page or the like for facial image registration. The user 10 photographs himself or herself using a camera included in the user terminal 20 and transmits the facial image to be registered to the management device 100. Note that details of facial image registration using a camera will be described later using FIG. 3.

The management device 100 stores the facial image acquired from the user 10 as facial photo data 30 in the storage unit. At this time, the management device 100 determines whether the facial image satisfies the quality required by each facial recognition engine, and stores the facial image as facial photo data 30 only when determining that the facial image satisfies the quality. It's okay. Further, if the management device 100 determines that the facial image does not meet the quality requirements, the management device 100 may request the user 10 to reshoot the image or make adjustments by processing the facial image.

If the facial image satisfies the quality of each facial recognition engine and can be registered, the management device 100 notifies the user 10 to that effect (step S2).

Upon acquiring the face photo data 30, the management device 100 registers the face photo data 30 in each face authentication engine. For example, the management device 100 transmits the face photo data 30 to a first provider having the first face recognition engine 60, and requests registration to enable the first face recognition engine 60 to be used for facial recognition of the user 10. (Step S3).

The first provider extracts the feature amount from the acquired facial photo data 30 by the feature amount extraction process executed by the first face authentication engine 60, and stores it in the feature amount storage unit 62. For example, the first face recognition engine 60 extracts facial features of the user 10 as parameters by hashing the face image included in the face photo data 30, and uses the extracted values as a feature quantity in the feature quantity storage unit 62. Store in. After this, the first provider may discard the face photo data 30 itself. As a result, the first provider does not hold the personal information such as the face photo data 30, but only the feature amount, and thus can reduce the resources and labor involved in data management.

Similarly, the management device 100 transmits the facial photo data 30 to the second provider and requests registration to enable the second facial authentication engine 70 to be used for facial authentication of the user 10 (step S4). Similar to the first provider, the second provider extracts feature quantities from the acquired face photo data 30 and stores them in the feature quantity storage section 72.

Similarly, the management device 100 transmits the facial photo data 30 to the third provider and requests registration to enable the third facial authentication engine 80 to be used for facial authentication of the user 10 (step S5). The third provider, like the first provider and the second provider, extracts the feature amount from the acquired face photo data 30 and stores it in the feature amount storage section 82.

The management device 100 registers the face photo data 30 of the users 10 as many times as there are face authentication engines to be used. When the registration is completed, the management device 100 associates information about the user 10, the registered facial photo data 30, the registered facial recognition engine, etc., and stores the information in the storage unit.

Next, the flow of confirming the identity of the user 10 using face authentication (verification process) will be described using FIG. 2. FIG. 2 is a diagram schematically showing the flow of face authentication processing according to the embodiment. The management system 1 shown in FIG. 2 further includes an authentication terminal 40 and a service provider 200.

The authentication terminal 40 is an information processing terminal installed at a place where identity verification is required. The authentication terminal 40 is capable of communicating with the management device 100 via a network, and, for example, transmits to the management device 100 an image taken with a camera included in the authentication terminal 40 . Then, the authentication terminal 40 acquires the authentication result from the management device 100 and notifies the person who requested authentication whether the identity has been confirmed (whether the person has been matched with the registered face image). Furthermore, the authentication terminal 40 determines whether or not to provide a predetermined service based on the verification result.

The service provider 200 is a business that provides various services. For example, the service provider 200 manages the authentication terminal 40 and provides various services to users whose identity has been confirmed. Note that in this disclosure, a service is not limited to commercial services, but is a general term for any mechanism or mechanism that the user 10 attempts to use, such as a factory entrance/exit control gate. For example, the services include permission to enter and exit a specific facility, attendance management, financial product settlement, and the like. The service provider 200 has user information 212 that is information about users who use the service, business data 214 that stores the type and content of the service, and the like.

In the example of FIG. 2, when the user 10 attempts to use a service provided by the service provider 200, the user 10 causes the authentication terminal 40 to take a picture of himself or herself and attempts facial authentication. For example, when the user 10 enters a facility whose admission is controlled, the user 10 requests the authentication terminal 40 to perform an authentication process in order to confirm his/her identity.

After photographing the user 10, the authentication terminal 40 transmits the facial image for authentication obtained by photographing to the management device 100 (step S11). The management device 100 transmits the acquired authentication facial image to each facial authentication provider.

For example, the management device 100 transmits the authentication face image to the first provider, and causes the first face authentication engine 60 to perform face authentication based on the authentication face image (step S12). Specifically, the first face authentication engine 60 compares the registered feature amount of the user 10 with the feature amount extracted from the authentication face image. Then, the first face authentication engine 60 transmits to the management device 100 a result (matching result) indicating a comparison result between the feature amount of the user 10 and the feature amount extracted from the authentication face image. When the degree of similarity between the feature amounts acquired from the first face authentication engine 60 is higher than a predetermined value, the management device 100 connects the previously registered user 10 to the authentication face image. It is determined that the matching with the relevant person (user 10 in this example) has been successful. On the other hand, if the degree of similarity is lower than the predetermined value, the management device 100 cannot match the pre-registered user 10 with the person associated with the authentication face image, and the person's identity cannot be confirmed. judge.

When the first face authentication engine 60 completes face authentication, the management device 100 stores the verification result in the authentication log 32. Note that the first face recognition engine 60 may perform the verification after recognizing that the person to be verified is the user 10 using an ID card or the like, or may perform the verification after recognizing that the person to be verified is the user 10 using an ID card or the like, or may perform the verification after recognizing that the person to be verified is the user 10 using an ID card or the like. A comparison with an authentication face image may also be performed. In the latter case, it is assumed that the first face recognition engine 60 extracts as a matching target a user 10 whose similarity is determined to be higher than other targets as a result of the matching.

Subsequently, the management device 100 transmits the authentication face image to the second provider and causes the second face authentication engine 70 to perform face authentication based on the authentication face image (step S13). When the face authentication is completed, the second face authentication engine 70 transmits the matching result to the management device 100. The management device 100 determines whether or not the verification of the user 10 was successful based on the degree of similarity between the feature amounts obtained as a verification result by the second face recognition engine 70, and stores the result in the authentication log 32. . Similarly, the management device 100 transmits the authentication face image to the third provider, and causes the third face authentication engine 80 to perform face authentication based on the authentication face image (step S14). When the face authentication is completed, the third face authentication engine 80 transmits the matching result to the management device 100. The management device 100 determines whether or not the verification of the user 10 was successful based on the degree of similarity between the feature amounts obtained as a verification result by the third face recognition engine 80, and stores the result in the authentication log 32. .

Then, the management device 100 transmits the authentication result for the user 10 to the authentication terminal 40 based on the matching results by the plurality of face authentication engines (step S15). In other words, the management device 100 transmits the result of determining whether the service requested by the user 10 may be provided to the user 10 to the authentication terminal 40 .

At this time, the management device 100 may refer to the external service partnership information 34 and determine the authentication result for the user 10. For example, it is assumed that the external service cooperation information 34 includes the level of security required by the external service. That is, the management device 100 compares the level of security required by the service provider 200 with the verification results from the plurality of facial recognition engines, and then transmits the authentication result for the user 10 to the authentication terminal 40 (step S15).

Various security levels can be set for a service in terms of its operation. For example, when using the authentication process according to the embodiment for daily attendance management, etc., the authentication strength required is too high and authentication errors (a situation in which it is difficult to confirm the identity of the user attempting authentication) occur frequently. Even if safety is high, usability may be low. On the other hand, when the authentication process according to the embodiment is used for payment processing at a financial institution, etc., it is desirable that the authentication strength be set extremely high to prevent unauthorized use.

Therefore, the management device 100 manages information such as how much authentication strength is required for each service. For example, if the service has low authentication strength, the management device 100 may permit the user 10 to provide the service if at least one of the plurality of facial recognition engines succeeds in authentication. Alternatively, if the service has a high authentication strength, the management device 100 may not permit the user 10 to provide the service unless all of the plurality of facial recognition engines succeed in authentication.

The management device 100 acquires information regarding authentication strength and the like as described above from the service provider 200. Then, the management device 100 transmits a determination result as to whether the service is available or not to the service provider 200 based on the comparison result between the authentication face image and the user 10 (step S16). The service provider 200 provides a predetermined service to the user 10 whose identity has been confirmed based on the determination result by the management device 100 (step S17). As an example, the service provider 200 notifies the user 10 that the identity of the user 10 has been confirmed via the authentication terminal 40, and operates the gate opening/closing device to permit the user 10 to enter. Note that provision of such a service may be performed by the management device 100 entrusted by the service provider 200.

As shown in FIG. 2, the management device 100 determines whether to permit the request of the user 10 based on the authentication results of each of the plurality of facial authentication engines that authenticated the facial image. That is, by using a plurality of face authentication engines, the management device 100 can improve the accuracy of authenticating the identity of the user and also make flexible settings according to the strength of authentication regarding the provision of services that involve authentication. Note that in the above example, the management device 100 acquires the degrees of similarity determined by each face recognition engine, and determines whether or not the authentication of the user 10 is successful by determining the degrees of similarity determined by the respective facial recognition engines. Indicated. However, the determination of success or failure of face authentication may be executed by each face authentication engine instead of by the management device 100. In this case, each face recognition engine holds a reference value of similarity. Then, each face recognition engine compares the registered feature values of the user 10 and the feature values extracted from the authentication face image, and if the similarity between the compared feature values is higher than a predetermined value, , determines that the verification of the user 10 is successful, and transmits the determination result to the management device 100. That is, the entity that executes the success/failure determination of face authentication may be either the management device 100 or each face authentication engine.

Next, facial image registration according to the embodiment will be described in detail using FIGS. 3 and 4. FIG. 3 is a diagram (1) for explaining the registration process according to the embodiment.

The management device 100 provides a predetermined registration page to the user 10 who wishes to register a facial image, and accepts a registration request. A screen 300 shown in FIG. 3 shows a dedicated page for facial image registration displayed on the user terminal 20.

On screen 300, the user 10 sets an ID and password for registration, and proceeds with the facial image registration process. When the user 10 inputs the ID and password, the screen 300 changes to a screen 302 (step S21).

The screen 302 includes a preview image 304 to be photographed with a camera provided in the user terminal 20. The management device 100 displays a guide 306 in the preview image 304. The guide 306 indicates a desirable range for the face to occupy in the photographed image. Since face recognition processing includes a step in which a face recognition engine detects a face from an image, it is desirable that all parts of the face be included in the image. For this reason, the management device 100 displays a guide 306 such as an approximate range occupied by the face and a line that serves as a guide for the user 10 to bring the face to the center of the image, and prompts the user 10 to take an appropriate photograph. .

The user 10 takes a photo of himself with a camera provided inside the user terminal 20 (on the screen side), adjusts his face to fit within the guide 306 while looking at the preview image 304, and then presses a photo button 308. When the user 10 performs photography according to such an operation, the management device 100 can acquire a facial image for registration via the user terminal 20.

When the photographing is finished, the screen 302 changes to the screen 310 (step S22). At this stage, the management device 100 may provide the user 10 with processing means for processing the captured image. Specifically, the management device 100 provides a trimming tool and the like for removing unnecessary parts of a captured image.

For example, when the user 10 uses a photographed image as an ID photo with a specified size, the user 10 can crop the photographed image to a specified size. Furthermore, with the trimming tool, the user 10 can also move the captured image. Therefore, if the user 10 does not include his or her own face within the circle of the guide 306, the user 10 can shift the entire image so that the face is included within the circle of the guide 306. Further, the user 10 may enlarge or reduce the image by pinching in or pinching out, and adjust the size so that the face is included within the circle of the guide 306 to an appropriate size. Note that the management device 100 may automatically perform the trimming process. For example, the management device 100 detects the face of the user 10 through face detection processing, and automatically adjusts the position of the face so that the coordinates of the detected face fall within the circle of the guide 306. Furthermore, the management device 100 automatically adjusts the size of the detected face so that it fits within the circle of the guide 306. Furthermore, the management device 100 may remove extra parts (background, etc.) other than the detected face. In this way, the management device 100 automatically performs trimming, so that the user 10 can automatically obtain an image suitable for registration by simply photographing an image that includes a face. Thereby, the management device 100 can reduce the burden on the user 10 related to face registration.

Furthermore, the management device 100 may provide other processing means. This point will be explained using FIG. 4. FIG. 4 is a diagram (2) for explaining the registration process according to the embodiment.

When the user 10 finishes trimming, the management device 100 transitions the screen 310 in FIG. 3 to the screen 320 shown in FIG. 4. On the screen 320, the user 10 can adjust the brightness of the entire image. For example, the user 10 can adjust the overall brightness to be brighter or darker by operating the brightness adjustment bar 322. Further, when the user 10 presses the automatic brightness adjustment button 324, the management device 100 refers to the current image brightness and the brightness standard value required for the face recognition engine, and adjusts the brightness to the face recognition engine. The brightness may be automatically adjusted so that it can be registered. Thereby, the user 10 can adjust the image to a brightness that can be registered without any complicated effort.

When the user 10 finishes adjusting the brightness, the management device 100 changes the screen 320 to the screen 326 (step S31). At screen 326, user 10 can select a background for the image. For example, if the user 10 wants to delete the background because the captured image includes an unnecessary background, the user 10 can press the background selection button 330 to change the captured image to a processed image 328 in which the background is changed to solid white or the like. I can do it. Alternatively, when the user 10 uses the photographed image as an ID photo with a certain background defined, the user 10 can change the background to a defined one (such as light blue).

When the user 10 finishes adjusting the background, the management device 100 transitions the screen 326 to the screen 332 (step S32). At this stage, the management device 100 determines the quality of the processed image, which is the photographed image after the processing has been completed.

Although details will be described later, the management device 100 determines for each face recognition engine whether the processed image satisfies the criteria required of each face recognition engine for registered images. When the management device 100 determines that the processed image satisfies the registration request criteria of all face authentication engines, it displays a determination result 336 indicating that the processing image satisfies the determination criteria. After confirming the determination result 336, the user 10 presses the save button 334 if the user 10 desires to register this processed image. When the user 10 presses the save button 334, the management device 100 saves this processed image in the storage unit as a registration image used for registration.

Note that if the processed image does not meet the criteria, the management device 100 causes the screen 326 to transition to a screen 338. On the screen 338, the management device 100 displays a determination result 340 indicating that the processed image does not meet the registration request criteria of any of the face authentication engines and cannot be registered. Note that, as shown in FIG. 4, the management device 100 may display, in the determination result 340, the details of the criteria that the processed image does not satisfy, based on the quality determination result. Thereby, the user 10 can obtain information on what points to pay attention to when creating a newly processed image when redoing photographing or processing.

When the management device 100 saves the processed image in the storage unit as a registration image to be used for registration, the management device 100 transitions the screen 332 to a screen 342, displays a facial image registration completion display 344, and notifies the user 10 that registration is complete. Notify (step S33).

In this way, the management device 100 determines the quality regarding whether or not the registration process of the face recognition engine is possible based on the processed image, so that the user 10 may be cropped after the quality determination and the determination result changes. It is possible to prevent troublesome situations from occurring. Furthermore, according to the management device 100, processing such as trimming and quality determination processing can be performed all at once on the server-side platform, making it possible to obtain high-quality facial images and improving the accuracy of feature extraction in the subsequent stage. can be connected to.

(1-2. Procedures for management processing according to embodiment)
Next, the procedure of the management process according to the embodiment will be explained using FIGS. 5 to 7. First, the flow of registration processing will be explained using FIG. 5. FIG. 5 is a sequence diagram showing the procedure of registration processing according to the embodiment.

As shown in FIG. 5, the user 10 applies for facial image registration via a website etc. provided by the management device 100 (step S40). Subsequently, the user 10 photographs the facial image to be applied for. Note that, if the user 10 has an image taken in advance, the user 10 may select the image as the image to be used for registration (step S41).

When the user 10 photographs or selects an image, the facial image is transmitted to the management device 100 (step S42). The management device 100 receives the facial image sent from the user 10 (step S43).

Thereafter, the user 10 uses the processing means provided by the management device 100 to perform a series of processing on the facial image. For example, the user 10 performs a trimming operation to cut out a face part, etc. (step S44). The management device 100 performs a trimming process on the face image according to the user's operation (step S45).

The user 10 also adjusts the brightness of the face image using the processing means provided by the management device 100 (step S46). The management device 100 performs brightness adjustment such as brightening or darkening the face image according to the user's operation (step S47).

The user 10 also changes the background of the face image using the processing means provided by the management device 100 (step S48). The management device 100 changes the background of the face image according to the user's operation (step S49).

After the user 10 finishes processing the face image and generates the processed image, the user 10 requests the management device 100 to check the quality of the processed image (step S50). When the management device 100 receives a quality check request from the user 10, it transmits the processed image to the face authentication provider 50 (step S51).

The face authentication provider 50 performs a quality check to determine whether the processed image can be used for authentication processing in the face authentication engine (step S52). Such processing is performed for each facial recognition engine of each provider. The face recognition provider 50 transmits the quality check results to the management device 100 (step S53).

The management device 100 refers to the quality check results from the plurality of face recognition engines and determines whether there are any problems with all the results (step S54). If it is determined that there is a quality problem in any of the face recognition engines (step S54; No), the management device 100 informs that the registration is NG and the reason, as shown in the screen 338 of FIG. The user 10 is notified (step S55).

On the other hand, if it is determined that there is no quality problem in all face recognition engines (step S54; Yes), the management device 100 informs the user that the registration determination is OK, as shown in the screen 332 of FIG. (Step S56).

After confirming the notified result, the user 10 presses the save button 334 displayed on the screen 332 to request the management device 100 to register the facial image for authentication (step S57).

Upon receiving the registration request from the user 10, the management device 100 registers the quality-determined facial image as an authentication facial image (step S58). Then, the management device 100 transmits the registered facial images for authentication to each facial authentication provider 50 (step S59). The face authentication provider 50 causes the face authentication engine to read the facial image for authentication, extracts the feature amount, and then registers the feature amount and the user ID in association with each other (step S60). Thereafter, the face authentication provider 50 may discard the acquired raw data of the registration face image itself.

Next, the procedure of authentication processing according to the embodiment will be described using FIG. 6. FIG. 6 is a sequence diagram showing the procedure of face authentication processing according to the embodiment.

As shown in FIG. 6, the authentication terminal 40 photographs the user 10 requesting face authentication and obtains a face image (step S70). The authentication terminal 40 transmits the acquired facial image to the management device 100 (step S71).

After acquiring the facial image, the management device 100 transmits the acquired facial image to the first face authentication engine 60 (step S72). The first face recognition engine 60 compares the acquired face image with the registered face image and executes face recognition processing (step S73). The management device 100 acquires the verification result (first verification result) by the first face authentication engine 60 (step S74).

Subsequently, the management device 100 transmits the face image acquired in step S72 to the second face authentication engine 70 (step S75). The second face recognition engine 70 compares the acquired face image with the registered face image and executes face recognition processing (step S76). The management device 100 acquires the verification result (second verification result) by the second face authentication engine 70 (step S77).

The management device 100 similarly transmits the facial images to all the facial recognition engines in which the facial images for authentication are registered, and acquires all the matching results (step S80). That is, the management device 100 similarly transmits the face image to the third face authentication engine 80 (not shown) and other registered face authentication engines, and obtains matching results.

Then, the management device 100 determines the verification result (step S81). As described above, the success or failure of authentication by each face recognition engine is determined based on whether the degree of similarity between the feature amounts transmitted from the face recognition engine exceeds a predetermined degree of similarity. Note that the details of the determination process for determining whether or not to allow the user 10 to use the service based on a plurality of authentication determinations will be described later using FIG. 7 . After this determination, the management device 100 transmits the determination result to the authentication terminal 40 (step S82). The authentication terminal 40 displays the results sent from the management device 100 on a screen or the like (step S83).

Next, the determination procedure of the face authentication process according to the embodiment will be explained using FIG. FIG. 7 is a flowchart showing a determination procedure of face authentication processing according to the embodiment.

As shown in FIG. 7, the management device 100 identifies the service that the user intends to use (step S101). For example, the management device 100 identifies the service that the user intends to use based on external service partnership information linked to the authentication terminal 40 and information on the service provider.

The management device 100 then refers to the criteria set for the service that the user intends to use. For example, the management device 100 may set a standard based on the authentication strength of the service, such as not allowing the service to be used unless the matching results of all face recognition engines are OK, or if the matching results of any one face recognition engine are OK. For example, the standard that the service can be used is referred to (step S102).

Then, the management device 100 determines whether or not the comparison results obtained from the plurality of face recognition engines satisfy the service determination criteria (step S103). If the service determination criteria are met (step S103; Yes), the management device 100 determines that face authentication is successful (step S104). In other words, the management device 100 determines to permit the authenticated user to use the service. On the other hand, if the service determination criteria are not satisfied (step S103; No), the management device 100 determines that face authentication has failed (step S105). In other words, the management device 100 determines not to allow the authenticated user to use the service.

(1-3. Configuration of management device according to embodiment)
Next, the configuration of the management device 100 that executes management processing according to the embodiment will be described. FIG. 8 is a diagram showing a configuration example of the management device 100 according to the embodiment.

As shown in FIG. 8, the management device 100 includes a communication section 110, a storage section 120, and a control section 130. Note that the management device 100 includes an input unit (for example, a keyboard, a mouse, etc.) that receives various operations from an administrator who manages the management device 100, and a display unit (for example, a liquid crystal display, etc.) for displaying various information. May have.

The communication unit 110 is realized by, for example, a NIC (Network Interface Card), a network interface controller, or the like. The communication unit 110 is connected to a network N (for example, the Internet) by wire or wirelessly, and transmits and receives information to and from the user terminal 20, the authentication terminal 40, the face authentication provider 50, etc. via the network N. For example, the communication unit 110 transmits and receives information using any communication standard or communication technology such as Wi-Fi (registered trademark), Bluetooth, SIM (Subscriber Identity Module), or LPWA (Low Power Wide Area). Good too.

The storage unit 120 is realized by, for example, a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disk. The storage unit 120 includes a user information storage unit 121 , a quality check item storage unit 122 , a face photo storage unit 123 , an authentication log storage unit 124 , and an external service information storage unit 125 .

Hereinafter, each storage unit will be explained in order using FIGS. 9 to 12. Note that in the examples shown in FIGS. 9 to 12, the information stored in the storage unit 120 may be conceptually shown as “A01”, but in reality, each piece of information to be described later is stored in the storage unit 120. shall be memorized.

The user information storage unit 121 stores information on users who use face authentication. FIG. 9 shows an example of information stored in the user information storage unit 121. FIG. 9 is a diagram illustrating an example of the user information storage unit 121 according to the embodiment. In the example shown in FIG. 9, the user information storage unit 121 includes items such as "user ID", "login information", "face photo registration", and "external service information".

"User ID" indicates identification information that identifies a user. "Login information" indicates login information required for facial photo registration, such as a password associated with a user ID. "Face photo registration" indicates whether or not the user has registered his or her face photo. "External service information" indicates information on an external service that the user is using.

The quality check item storage unit 122 stores information regarding quality check items for each face recognition engine. FIG. 10 is a diagram illustrating an example of the quality check item storage unit 122 according to the embodiment. In the example shown in FIG. 10, the quality check item storage unit 122 includes items such as "check item", "first face recognition engine", "second face recognition engine", and "third face recognition engine".

The quality check items shown in FIG. 10 indicate on what items each face recognition engine checks the quality of a face image and outputs a result of registration OK or NG. In the example shown in FIG. 10, the first face recognition engine determines horizontal, vertical, and tilt blurring in the angle of the face in the image, and determines whether these are within predetermined standards. It shows. In other words, the first face recognition engine makes it a requirement for registration that the face in the image is accurately facing the front, and uses the angle of the face as a quality check item. On the other hand, the example in FIG. 10 shows that the third face recognition engine only checks the tilt blur, regardless of the horizontal or vertical direction of the face angle in the image. In this way, each face recognition engine has different face angles and characteristics that it is good at recognizing, and this also affects the difference in authentication results. Note that when the management device 100 receives the determination of registration NG from the first face authentication engine, it acquires what check items were checked. The management device 100 can list the acquired information in the determination result 340 shown in FIG. 4, for example, and notify the user of which items the processed image to be registered is inappropriate.

Note that the quality check items shown in FIG. 10 are merely examples, and the management device 100 may store information related to various quality check items in addition to the example shown in FIG. 10.

The face photo storage unit 123 stores face photo data including a face image used in face authentication. FIG. 11 is a diagram illustrating an example of the facial photo storage unit 123 according to the embodiment. In the example shown in FIG. 11, the face photo storage unit 123 has items such as "user ID", "registered face photo data", "for authentication", and "for ID photo".

"User ID" indicates identification information that identifies a user. "Registered facial photo data" indicates facial photo data registered by the user. "For authentication" indicates face photo data used for face authentication. "For ID photo" indicates face photo data that has been cropped or the background has been changed for ID photo use. That is, the management device 100 can separately hold the user's face photo data for authentication and ID photo data.

The authentication log storage unit 124 stores a log of face authentication performed by the user. For example, the authentication log storage unit 124 stores the date and time when the user attempted facial authentication, the results of facial authentication, and the like. Note that the authentication log storage unit 124 may store a log of the user's use of the service. For example, the authentication log storage unit 124 may store a user's attendance record and the like.

The external service information storage unit 125 stores information regarding external services that can be used by the user after identity verification such as face authentication. FIG. 12 is a diagram illustrating an example of the external service information storage unit 125 according to the embodiment. In the example shown in FIG. 12, the external service information storage unit 125 has items such as "service ID", "type", and "judgment criteria".

“Service ID” indicates identification information that identifies a service. “Type” indicates the type of service. The "judgment criterion" indicates a criterion for whether or not to allow a user to use a service based on the authentication strength regarding face authentication. For example, when the determination criterion is "OR determination", if the matching result of any one of the plurality of face recognition engines is OK, the user can use the service. Further, when the determination criterion is "AND determination", the user can use the service only when the matching results are OK in all of the plurality of face recognition engines. In addition, if the judgment criterion is "OK judgment with 2 or more", if the matching result is OK with any two or more of the multiple facial recognition engines, the user will be able to confirm that the service is OK. Becomes available. Note that the determination criteria shown in FIG. 12 are only an example, and the external service may register various determination criteria in the management device 100 based on the desired authentication strength.

Returning to FIG. 8, the explanation will be continued. The control unit 130 allows programs stored in the management device 100 to use RAM (Random Access Memory) or the like as a work area using, for example, a CPU (Central Processing Unit), an MPU (Micro Processing Unit), a GPU (Graphics Processing Unit), or the like. This is achieved by executing as . Further, the control unit 130 is a controller, and is realized by, for example, an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

As shown in FIG. 8, the control unit 130 includes a registration processing unit 130A and an authentication processing unit 130B. The registration processing unit 130A controls processing related to facial image registration. The registration processing unit 130A includes a reception unit 131, a provision unit 132, a quality determination unit 133, and a registration unit 134. The authentication processing unit 130B controls processing related to authentication processing. The authentication processing section 130B includes an acquisition section 135, a transmission section 136, and a result determination section 137.

The receiving unit 131 receives from a user a facial image that is the source of correct data used by each of the plurality of face recognition engines for verification in order to perform authentication processing by the plurality of face recognition engines. Note that the correct data used by the face recognition engine for verification is not the face image itself but the feature amount extracted from the face image.

For example, the reception unit 131 provides the user with a page (such as a website) for registering a facial image, and receives the facial image from the user via this page. Such a face image is used as a registration image to be registered in the face recognition engine. Note that when the face image is processed by a processing means provided by the providing unit 132 described later, the receiving unit 131 receives the processed image as the image for registration.

The providing unit 132 provides processing means for the user to process the facial image received by the receiving unit 131.

For example, the providing unit 132 provides processing means for converting at least one of the size of the face image, the range occupied by the face in the face image, the brightness of the face image, and the background of the face image. Specifically, as shown in FIGS. 3 and 4, the providing unit 132 provides editing that allows the user to perform trimming, brightness adjustment, background change, etc. on the page for registering a face image. Provide tools. Note that the types of processing are not limited to those described above, and the providing unit 132 may provide a tool that can perform teeth whitening, complexion whitening processing, etc. on a face image, for example. These processes may be performed by the user himself or herself by operating tools to adjust the teeth and facial color to a desired color, or by the provider 132 to automatically adjust the color to a suitable color.

Further, the providing unit 132 may provide reference information for processing the image into a processed image suitable for the service used by the user through a plurality of face authentication engines or authentication processing. The reference information is, for example, a guide 306 as shown in FIG. 3 that is displayed on the screen of the user terminal 20 in order for the user to process the image into a suitable processed image. The guide 306 is displayed based on, for example, the size ratio and position of the face included in the face image. Further, the reference information may include a horizontal line for checking the parallelism of the face and a vertical line for checking the center of the image.

Note that the providing unit 132 does not have to provide all the processing means shown in FIGS. 3 and 4. For example, if it is clear that the user will not use the processed image for an ID photo, the providing unit 132 may skip the background change page that is mainly used for creating an ID photo and proceed with the registration process. .

Furthermore, if it is difficult for the user to select the brightness, size, etc. of the image, the providing unit 132 may perform a process of automatically adjusting the brightness and size of the image. That is, the providing unit 132 converts at least one of the size of the face image, the range occupied by the face in the face image, the brightness of the face image, and the background of the face image according to the user's request, and converts at least one of the following: A processed image that can be used as correct data for authentication processing may be automatically generated. In this case, the providing unit 132 refers to the quality check item storage unit 122, for example, and obtains information such as the brightness required for the face image, the size of the image, and the position where the face should be placed in the image. Then, the providing unit 132 automatically processes the facial image and generates a processed image in accordance with those requests. At this time, the providing unit 132 may use known techniques such as processing to detect a face included in an image. Thereby, the providing unit 132 can generate a processed image that ensures the required quality without requiring any effort on the part of the user.

Further, when the quality determining unit 133 (described later) determines that a predetermined quality check item regarding the processed image does not satisfy the quality, the providing unit 132 transmits the information in the processed image related to the quality check item to the quality check item. You may convert it to meet the requirements. For example, if the quality determining unit 133 determines that the processed image is insufficiently bright, the providing unit 132 automatically converts the processed image to brighten it to meet the quality, and converts the processed image into a converted image. may be generated. In this case, the providing unit 132 may present the converted image to the user once, or may automatically replace the original processed image with the converted image and proceed with the registration process.

The quality determination unit 133 can use registration images such as face images accepted by the reception unit 131 and processed images that are face images processed by processing means as correct answer data for authentication processing by a plurality of face authentication engines. Determine whether the quality is good or not.

Specifically, the quality determination unit 133 determines for each of the plurality of face recognition engines whether the registration image has a quality that can be used as correct data for the authentication processing of the plurality of face recognition engines.

More specifically, the quality determination unit 133 determines whether the registration image satisfies standards regarding predetermined quality check items set for each of the plurality of face recognition engines, and based on the determination result, It is determined whether the registration image has a quality that can be used as correct data for authentication processing by a plurality of face authentication engines.

The registration unit 134 registers the registration image when it is determined based on the determination result by the quality determination unit 133 that the registration image has a quality that can be used as correct answer data for authentication processing for all of the plurality of face recognition engines. register.

If the quality determination unit 133 determines that any of the predetermined quality check items regarding the registration image does not meet the quality, the registration unit 134 notifies the user of which quality check item does not meet the standards. You may. For example, the registration unit 134 displays a screen 338 and a determination result 340 shown in FIG. 4 on the user terminal 20 to notify the user which quality check item does not meet the criteria.

Note that, if the registration unit 134 determines that the registration image does not have a quality that can be used as correct data for authentication processing for any of the plurality of face recognition engines, based on the determination result by the quality determination unit 133, the registration unit 134 does not accept the registration image. Notify the user that the face image cannot be registered as an authentication face image. That is, when there are multiple face recognition engines defined in a service, etc., the registration unit 134 registers only face images that can be registered in all the face recognition engines. Thereby, the registration unit 134 can ensure safety in face authentication processing.

Further, when registering an image for registration, the registration unit 134 defines the image for registration as having at least one of the following: the size of the face image, the range occupied by the face in the face image, the brightness of the face image, and the background of the face image. A different second face image may be stored in association with the registration image. Specifically, when registering a user, the registration unit 134 may store the ID photo face image together with the authentication face image. Thereby, the management device 100 can provide the face image for ID photo to the service side as needed, such as when creating an ID card for the service.

The acquisition unit 135 acquires facial images from the user for causing a plurality of facial recognition engines to perform authentication processing. For example, the acquisition unit 135 acquires a facial image for performing authentication processing via the authentication terminal 40 that photographed the user.

The transmitter 136 transmits the face image acquired by the acquirer 135 to each of the plurality of face recognition engines.

Note that the transmitting unit 136 does not necessarily need to transmit the facial image to all the facial recognition engines depending on the characteristics of the acquired facial image or the request for the service. Depending on the facial recognition engine, some engines may be more accurate when matching bright images (e.g., images taken during the day) or dark images (e.g., images taken at night), or may be more accurate when matching specific races. There may be engines that have been trained to do so. For this reason, when the acquired face image has a brightness lower or higher than a predetermined reference value, or when a face image of a specific race is photographed, the transmitting unit 136 performs facial recognition with high accuracy for such a face image. The facial image may be sent to the engine preferentially. A service that a user attempts to use after authentication may specify a specific face recognition engine (for example, a face recognition engine that is highly reliable for the service). In such a case, the transmitter 136 may select a facial recognition engine to transmit the facial image depending on the type of user's request (that is, the service the user intends to use).

That is, the transmitting unit 136 may select which facial recognition engine to transmit the facial image to based on information regarding the facial image or a user's request. For example, the transmitting unit 136 determines which facial recognition engine to send the facial image to based on the brightness of the facial image or the user's characteristic information (skin color, race, etc.) included in the facial image. You may choose.

The result determination unit 137 determines whether or not to permit the user's request, depending on the number of face recognition engines that have successfully matched the face image with the correct data among the plurality of face recognition engines.

For example, the result determination unit 137 determines whether the number of face recognition engines that have successfully matched the face image with the correct data among the plurality of face recognition engines is based on the determination criteria set by the service used by the user after the authentication process. Based on whether or not the request is satisfied, it is determined whether or not to permit the user's request. Specifically, the result determination unit 137 refers to the external service information storage unit 125 and determines that the user's identity has been authenticated if a verification result that satisfies the criteria required by the service is obtained. , permit you to use the Service. Note that if the service side does not specify the determination criteria, the result determination unit 137 may approve the user's request based on successful verification using one of the facial recognition engines (OR determination), or may approve the user's request based on successful verification using any of the facial recognition engines (OR determination). The user's request may be permitted if the face recognition engine succeeds in matching (AND determination).

Furthermore, when the sending unit 136 sends the face image only to a specific face recognition engine, the result determination unit 137 determines whether to permit the user's request based on the authentication result of the face recognition engine to which the face image was sent. It may be determined whether or not.

(2. Modification of embodiment)
(2-1. Equipment configuration)
In the embodiment described above, an example was shown in which, in the management system 1, the user terminal 20 is used when registering a face image, and the authentication terminal 40 is used during authentication processing. However, the device configuration and roles within the management system 1 can be changed as appropriate.

For example, when a user attempts to use a service on a website displayed on the user terminal 20, the user terminal 20 may play the role of the authentication terminal 40. That is, the user terminal 20 may transmit a photographed face image of the user to the management device 100 and have the management device 100 execute the user's face authentication process. For example, if the user terminal 20 receives a result from the management device 100 that the face authentication is OK, the user terminal 20 may perform operations after the user is permitted to use the service, such as displaying the service that the user attempted to access. may be executed.

Furthermore, in the embodiment described above, an example has been shown in which each face authentication provider is operated by a different business operator from the management device 100, and each face authentication provider is operated by a different business operator. However, the management device 100 may perform the processing according to the embodiment in an on-premise environment, in which the management device 100 operates multiple facial recognition engines on its own, rather than receiving face recognition engines from a facial recognition provider. good. In this case, the management device 100 may include a plurality of face recognition engines having different characteristics, for example by generating the plurality of face recognition engines using different learning methods. Thereby, the management device 100 can execute management processing using a plurality of face recognition engines without necessarily involving a plurality of businesses. Furthermore, the management device 100 does not necessarily need to be a cloud server, and may function as a platform that manages multiple facial recognition engines in a system built on a local network.

(3. Other embodiments)
The processing according to the embodiment described above may be implemented in various different forms other than the embodiment described above.

For example, among the processes described in the above embodiments, all or part of the processes described as being performed automatically can be performed manually, or the processes described as being performed manually can be performed manually. All or part of the process can also be performed automatically using known methods. In addition, information including the processing procedures, specific names, and various data and parameters shown in the above documents and drawings may be changed arbitrarily, unless otherwise specified. For example, the various information shown in each figure is not limited to the illustrated information.

Furthermore, each component of each device shown in the drawings is functionally conceptual, and does not necessarily need to be physically configured as shown in the drawings. In other words, the specific form of distributing and integrating each device is not limited to what is shown in the diagram, and all or part of the devices can be functionally or physically distributed or integrated in arbitrary units depending on various loads and usage conditions. Can be integrated and configured.

Furthermore, the embodiments and modifications described above can be combined as appropriate within a range that does not conflict with the processing contents.

Furthermore, the effects described in this specification are merely examples and are not limited, and other effects may also be present.

(4. Effects of the management device according to the present disclosure)
As described above, the management device according to the present disclosure (the management device 100 in the embodiment) includes a reception unit (the reception unit 131 in the embodiment), a provision unit (the provision unit 132 in the embodiment), and a quality determination unit ( In the embodiment, the quality determination unit 133) is included. The reception unit receives from a user a facial image that is a source of correct data that each of the plurality of face recognition engines uses for verification in order to perform authentication processing by the plurality of face recognition engines. The providing unit provides processing means for the user to process the facial image received by the receiving unit. The quality determination unit determines whether the processed image, which is a face image processed by the processing means, has a quality that can be used as correct answer data for authentication processing by a plurality of face authentication engines.

In this way, the management device according to the present disclosure provides a means for processing a face image received from a user, and also determines the quality of the processed image, such as whether the face recognition engine can use it for authentication processing. . In this way, the management device 100 can perform processing such as trimming and quality determination processing all at once on the server-side platform, making it possible to obtain high-quality facial images and improving the accuracy of feature extraction in the subsequent stage. can be connected to. In addition, the management device can prevent the occurrence of image capture by checking the quality of the processed image, thereby reducing the burden on the user and accurately registering facial images.

Further, the providing unit provides reference information for processing the processed image into a processed image suitable for the service used by the user through a plurality of face authentication engines or authentication processing.

In this way, the management device can standardize the size and quality of registration images received from users by providing the user with reference information indicating face size, etc., along with processing means, and can smoothly proceed with registration. I can do it.

The providing unit also provides processing means for converting at least one of the size of the face image received by the reception unit, the range occupied by the face in the face image, the brightness of the face image, and the background of the face image. do.

In this way, the management device can provide an environment in which the user can edit images as desired by providing a means for processing images taken by the user.

Further, the providing unit converts at least one of the size of the face image accepted by the reception unit, the range occupied by the face in the face image, the brightness of the face image, and the background of the face image in accordance with the user's request, A processed image is generated in a manner that can be used as correct answer data for authentication processing by a plurality of face recognition engines.

In this way, the management device can automatically process the facial image, so it is possible to generate a processed image that ensures the required quality without any effort on the part of the user.

Furthermore, if the quality determining unit determines that a predetermined quality check item regarding the processed image does not satisfy the quality, the providing unit updates the information in the processed image related to the predetermined quality check item to ensure that the quality is met. Convert to

In this way, the management device can generate a processed image that guarantees the required quality without requiring any effort from the user by automatically converting the image so as to maintain the quality.

Furthermore, the management device may be configured to include an acquisition unit (acquisition unit 135 in the embodiment), a transmission unit (transmission unit 136 in the embodiment), and a result determination unit (result determination unit 137 in the embodiment). good. The acquisition unit acquires facial images from the user for causing the plurality of facial recognition engines to perform authentication processing. The transmitter transmits the facial image acquired by the acquirer to each of the plurality of face recognition engines. The result determination unit determines whether or not to permit the user's request based on the authentication results of each of the plurality of face recognition engines that have authenticated the face image.

In this way, the management device executes authentication processing using a plurality of face authentication engines. Thereby, the management device can improve authentication accuracy and safety, and provide flexible authentication such as setting different criteria for each service.

The result determination unit also determines whether or not to permit the user's request, depending on the number of face recognition engines that have successfully matched the face image with the correct data among the plurality of face recognition engines.

In this way, the management device performs authentication using multiple facial recognition engines, so it is possible to take advantage of the strengths of various facial recognition engines, such as facial recognition engines that are strong in certain environments and facial recognition engines that are strong in certain races. Authentication can be performed.

In addition, the result judgment unit determines that the number of face recognition engines that have successfully matched the face image with the correct data among the plurality of face recognition engines satisfies the determination criteria set by the service used by the user after the authentication process. It is determined whether or not to permit the user's request based on whether or not the request is granted.

In this way, the management device can decide whether or not to approve the user's request based on the service's criteria, so the management device can decide whether or not to approve the user's request based on the service criteria. Rather than simple processing, it is possible to perform flexible authentication processing according to the authentication strength and needs of the service.

Further, the transmitter selects which face authentication engine the facial image is to be transmitted to, based on information regarding the facial image or a user's request. The result determination unit determines whether or not to permit the user's request based on the authentication result of the face recognition engine to which the face image is transmitted.

In this way, the management device may take advantage of the advantage of using a plurality of face recognition engines and selectively use the face recognition engines based on the image characteristics that each engine is good at. This allows the management device to perform authentication with greater safety in mind, such as using only a highly accurate face authentication engine.

Further, the transmitter selects which face recognition engine the face image is to be transmitted to, based on the brightness of the face image or the user's characteristic information included in the face image.

In this way, the management device can selectively transmit facial images that each facial authentication engine is good at by selecting a facial authentication engine according to the characteristics of the image.

Further, the management device may be configured to include a reception section, a quality determination section, and a registration section (registration section 134 in the embodiment). The reception unit receives from a user a registration image that is a source of correct data that each of the plurality of face recognition engines uses for verification in order to perform authentication processing by the plurality of face recognition engines. The quality determination unit determines for each of the plurality of face authentication engines whether the registration image has a quality that can be used as correct answer data for authentication processing of the plurality of face authentication engines. The registration unit registers the registration image when determining that the registration image has a quality that can be used as correct answer data for authentication processing for all of the plurality of face recognition engines based on the determination result by the quality determination unit. .

In this way, the management device registers only facial images that satisfy the quality required by all the facial recognition engines used. Thereby, the management device can smoothly proceed with authentication processing using a plurality of face authentication engines.

In addition, the quality determination unit determines whether the registration image satisfies standards regarding predetermined quality check items set for each of the plurality of face recognition engines, and based on the determination result, the registration image It is determined whether the quality is such that it can be used as correct answer data for authentication processing by a plurality of face recognition engines.

In this way, the management device provides items to be checked in quality determination, and performs quality determination in accordance with the items. Thereby, the management device can reliably register facial images of a quality that can be used by the facial recognition engine.

In addition, if the quality determination unit determines that any of the predetermined quality check items regarding the registration image does not meet the quality, the registration unit notifies the user of which quality check item does not meet the standards. do.

In this way, since the management device notifies the user that the quality is not satisfied for each item, it is possible to clearly tell the user how to modify the registration image.

In addition, if the registration unit determines that the registration image does not have a quality that can be used as correct data for authentication processing for any of the plurality of face recognition engines based on the determination result by the quality determination unit, the registration unit selects the registration image for authentication. Notify the user that it cannot be registered as a face image.

In this manner, the management device can prompt the user to appropriately register with all face recognition engines by notifying the user that registration has not been possible with any of the face recognition engines.

In addition, when registering an image for registration, the registration unit defines at least one of the size of the face image, the range occupied by the face in the face image, the brightness of the face image, and the background of the face image. A different second face image (for example, an ID photo image) is stored in association with the registration image.

In this way, the management device may store an ID photo image separately from the authentication image. Thereby, the management device can respond flexibly as a platform, such as providing a face image for an ID photo to the service side as needed, such as when creating an ID card for the service.

(5. Hardware configuration)
The information devices such as the management device 100 and the user terminal 20 according to the embodiments described above are realized by, for example, a computer 1000 having a configuration as shown in FIG. 13. Hereinafter, the management device 100 according to the embodiment will be described as an example. FIG. 13 is a hardware configuration diagram showing an example of a computer 1000 that implements the functions of the management device 100. Computer 1000 has CPU 1100, RAM 1200, ROM (Read Only Memory) 1300, HDD (Hard Disk Drive) 1400, communication interface 1500, and input/output interface 1600. Each part of computer 1000 is connected by bus 1050.

CPU 1100 operates based on a program stored in ROM 1300 or HDD 1400, and controls each section. For example, the CPU 1100 loads programs stored in the ROM 1300 or HDD 1400 into the RAM 1200, and executes processes corresponding to various programs.

The ROM 1300 stores boot programs such as a BIOS (Basic Input Output System) that are executed by the CPU 1100 when the computer 1000 is started, programs that depend on the hardware of the computer 1000, and the like.

The HDD 1400 is a computer-readable recording medium that non-temporarily records programs executed by the CPU 1100, data used by the programs, and the like. Specifically, HDD 1400 is a recording medium that records a program that executes management processing according to the present disclosure, which is an example of program data 1450.

Communication interface 1500 is an interface for connecting computer 1000 to external network 1550 (eg, the Internet). For example, CPU 1100 receives data from other devices or transmits data generated by CPU 1100 to other devices via communication interface 1500.

Input/output interface 1600 is an interface for connecting input/output device 1650 and computer 1000. For example, the CPU 1100 receives data from an input device such as a keyboard or a mouse via the input/output interface 1600. Further, the CPU 1100 transmits data to an output device such as a display, speaker, or printer via an input/output interface 1600. Furthermore, the input/output interface 1600 may function as a media interface that reads programs and the like recorded on a predetermined recording medium. Media includes, for example, optical recording media such as DVD (Digital Versatile Disc) and PD (Phase change rewritable disk), magneto-optical recording media such as MO (Magneto-Optical disk), tape media, magnetic recording media, semiconductor memory, etc. It is.

For example, when the computer 1000 functions as the management device 100 according to the embodiment, the CPU 1100 of the computer 1000 realizes the functions of the control unit 130 and the like by executing a management processing program loaded onto the RAM 1200. Further, the HDD 1400 stores a program that executes management processing according to the present disclosure and data in the storage unit 120. Note that although the CPU 1100 reads and executes the program data 1450 from the HDD 1400, as another example, these programs may be obtained from another device via the external network 1550.

Although the embodiments of the present application have been described above in detail based on the drawings, these are merely examples, and various modifications and improvements can be made based on the knowledge of those skilled in the art, including the embodiments described in the disclosure section of the invention. It is possible to implement the invention in other forms.

1 Management System 10 User 20 User Terminal 40 Authentication Terminal 50 Face Recognition Provider 60 First Face Recognition Engine 70 Second Face Recognition Engine 80 Third Face Recognition Engine 100 Management Device 110 Communication Unit 120 Storage Unit 121 User Information Storage Unit 122 Quality Check Item storage unit 123 Face photo storage unit 124 Authentication log storage unit 125 External service information storage unit 130 Control unit 131 Reception unit 132 Providing unit 133 Quality determination unit 134 Registration unit 135 Acquisition unit 136 Transmission unit 137 Result determination unit 200 Service provider

Claims (7)

a reception unit that receives from a user a registration image that is a source of correct data used by each of the plurality of face recognition engines for verification in order to perform authentication processing by the plurality of face recognition engines;
a quality determination unit that determines for each of the plurality of face authentication engines whether the registration image has a quality that can be used as correct answer data for authentication processing of the plurality of face authentication engines;
Registration for registering the registration image when it is determined that the registration image has a quality that can be used as correct answer data for authentication processing for all of the plurality of face recognition engines based on the determination result by the quality determination unit. Department and
A management device comprising: The quality determining section includes:
It is determined whether the registration image satisfies standards regarding predetermined quality check items set for each of the plurality of face recognition engines, and based on the determination result, the registration image is checked by the plurality of face recognition engines. Determine whether the quality is such that it can be used as correct answer data for authentication processing.
The management device according to claim 1, characterized in that: The registration department is
If the quality determination unit determines that any of the predetermined quality check items regarding the registration image does not meet the quality, the user is notified of which quality check item does not meet the standards;
The management device according to claim 2, characterized in that: The registration department is
If it is determined based on the determination result by the quality determination unit that the registration image does not have a quality that can be used as correct data for authentication processing for any of the plurality of face recognition engines, the registration image is used as the authentication facial image. informing the user that he or she cannot register as a
The management device according to any one of claims 1 to 3, characterized in that: The registration department is
When registering the registration image, the registration image is a second face that differs in at least one of the following: the size of the face image, the range occupied by the face in the face image, the brightness of the face image, and the background of the face image. Storing the image in association with the registration image,
The management device according to any one of claims 1 to 4, characterized in that: The computer is
In order to perform authentication processing by multiple facial recognition engines, each of the multiple facial recognition engines receives from the user a registration image that is the source of correct data used for verification,
Determining for each of the plurality of face recognition engines whether the registration image has a quality that can be used as correct data for authentication processing of the plurality of face recognition engines;
If it is determined based on the determined result that the registration image has a quality that can be used as correct data for authentication processing for all of the plurality of face recognition engines, registering the registration image;
A management method characterized by comprising: computer,
a reception unit that receives from a user a registration image that is a source of correct data used by each of the plurality of face recognition engines for verification in order to perform authentication processing by the plurality of face recognition engines;
a quality determination unit that determines for each of the plurality of face authentication engines whether the registration image has a quality that can be used as correct answer data for authentication processing of the plurality of face authentication engines;
Registration for registering the registration image when it is determined that the registration image has a quality that can be used as correct answer data for authentication processing for all of the plurality of face recognition engines based on the determination result by the quality determination unit. Department and
A management program that functions as a management device comprising:

Images