Determination method, determination program, and information processing device

Abstract

In the determination method according to the present invention, a computer executes processing of: calculating, when identification information identifying a person detected from an image is acquired, a certainty value indicating certainty that the identification information corresponds to the person; referring to, when location information relating to the person is acquired, a storage unit which stores a certainty threshold value and location information in association with each other for each of a plurality of services; and determining whether or not the calculated certainty value is greater than or equal to a certainty threshold value of a service corresponding to the location information relating to the person among the plurality of services.　

Description

Determination method, determination program, and information processing device

[0001] The present invention relates to a determination method, a determination program, and an information processing device.

[0002] A method for determining a threshold value of an authentication score for permitting use of a service for each individual service type has been disclosed (see, for example, Patent Document 1).

[0003] Japanese Patent Application Laid-Open No. 2003-248661

[0004] According to this method, the authentication processing device recognizes the service type selected by the individual, determines an authentication score for permitting use of the service, and acquires biometric information. The authentication processing device then determines whether the individual is permitted to use the service type selected by the individual. However, this method imposes a large processing load on the authentication process.

[0005] In one aspect, the present invention aims to provide a determination method, a determination program, and an information processing device that can reduce the processing load required for authentication processing.

[0006] In one aspect, the determination method involves a computer executing a process in which, when it acquires identification information that identifies a person detected from an image, it calculates a probability value indicating the likelihood that the identification information corresponds to the person, and when it acquires location information of the person, it references a memory unit that stores a probability threshold and location information associated with each of a plurality of services, and determines whether the calculated probability value is greater than or equal to the probability threshold of the service among the plurality of services that corresponds to the location information of the person.

[0007] The processing load required for authentication processing can be reduced.

[0008] FIG. 1 is a diagram illustrating an authentication space. (a) is a block diagram illustrating an overall configuration of a biometric authentication system according to a first embodiment, and (b) is a functional block diagram illustrating each function of an information processing device. (b) is a flowchart illustrating an ID linking process. (c) is a diagram illustrating an ID table stored in a storage unit. (d) is a diagram illustrating a position information table stored in a storage unit. (e) is a flowchart illustrating a service provision process. (f) is a diagram illustrating a case where feature information of a specific person is extracted from a plurality of images. (f) is a block diagram illustrating a hardware configuration of an information processing device.

[0009] Technologies that use biometric authentication to provide services such as payment are known. By using biometric authentication technology for payment, customers can shop smoothly without having to present cash or payment codes.

[0010] Furthermore, for example, by displaying products for individual customers identified through biometric authentication and the location information of the shelves where those products are located on electronic billboards within the store, customers will be able to shop more smoothly.

[0011] Here, for example, biometric authentication used for payment requires a high certainty value indicating the certainty of identity, so a strict certainty threshold may be set to reduce the false acceptance rate. On the other hand, biometric authentication used for electronic billboards may be acceptable even if it is lower than when used for payment, so a looser certainty threshold may be set to reduce the false rejection rate in order to reduce the load of authentication processing, etc.

[0012] In this way, the required accuracy value may differ depending on the service provided. For example, when providing multiple services at an airport, a high accuracy value is required for whether or not the customer can pass through the exit gate, whether or not the customer can check in their baggage, whether or not the customer can go through security, whether or not the customer can purchase goods, etc. On the other hand, a lower accuracy value may be required for providing flight information, whether or not the customer can enter a lounge, providing campaign information, providing free goods, etc.

[0013] However, when providing multiple services, even if the accuracy value required for each service is different, the user is required to perform an authentication operation for biometric authentication each time they use each service. In this case, the processing load required for the authentication process is large. Therefore, in the following embodiments, a determination method, a determination program, and an information processing device that can reduce the load required for the authentication process will be described.

[0014] An overview of the first embodiment will be described. Fig. 1 is a diagram illustrating an authentication space. A person who enters the authentication space in Fig. 1 registers feature information that can be acquired by a camera and an ID. The authentication space is a space that provides multiple services.

[0015] For example, multiple service providing terminals 130a to 130e are installed in an authentication space. The accuracy value required for each of the service providing terminals 130a to 130e is set individually, and may be different or the same. The accuracy value refers to the degree of certainty that the person detected by the camera is the person in question. If the accuracy value is high, there is a high probability that the person detected by the camera is the person in question.

[0016] In the authentication space, multiple tracking cameras 120 are installed at different positions. A person can be detected from the images captured by each tracking camera 120. The person can be estimated to be the person whose ID has the highest similarity between the feature information extracted from the image and the registered feature information.

[0017] The position of each person can be detected by storing position information of each tracking camera 120 and detecting which tracking camera 120 the person is captured in. By detecting the position of each person, it can be detected which service providing terminal each person is in a state where they can receive service from.

[0018] 1, for example, when person A appears in an image acquired by a specific tracking camera 120, it is detected that person A is located near the service providing terminal 130a and is in a state where he or she can receive services from the service providing terminal 130a. In this case, it is determined whether or not the certainty value of person A is equal to or greater than the certainty value required by the service providing terminal 130a. If the certainty value of person A is equal to or greater than the certainty threshold required by the service providing terminal 130a, person A can receive services from the service providing terminal 130a.

[0019] For example, as illustrated in FIG. 1 , person B is detected as being near multiple service providing terminals 130b-130d and in a state where he or she can receive services from the multiple service providing terminals 130b-130d because he or she appears in an image acquired by a specific tracking camera 120. In this case, the system determines whether person B's accuracy value is equal to or greater than the highest accuracy threshold among the accuracy thresholds requested by the service providing terminals 130b-130d. If person B's accuracy value is less than the highest accuracy threshold, the service providing terminals 130b-130d display a message to improve person B's accuracy value. For example, a message to pass through a less crowded area or a message to move to a location with a better view may be displayed. Passing through a less crowded area makes person B more likely to be captured in its entirety by the tracking camera 120, allowing feature information to be acquired with high accuracy, thereby increasing the accuracy value. Alternatively, by moving person B to a position where he or she can directly face the tracking camera 120, feature information can be acquired with high accuracy, thereby increasing the accuracy value.

[0020] For example, as illustrated in FIG. 1 , person C is detected as being near the service providing terminal 130e and in a state where he or she can receive services from the service providing terminal 130e because he or she appears in an image acquired by a specific tracking camera 120. In this case, it is determined whether or not the certainty value of person C is equal to or greater than the certainty threshold required by the service providing terminal 130e. If the area around person C is crowded, the certainty value of person C may become low. Therefore, the certainty value of person C may become less than the certainty threshold. In this case, a message to improve the certainty value of person C is displayed on the service providing terminal 130e.

[0021] Details will be explained below. Fig. 2(a) is a block diagram illustrating an example of the overall configuration of a biometric authentication system 200 according to Example 1. As illustrated in Fig. 2(a), the biometric authentication system 200 includes an information processing device 100, a linking camera 110, a tracking camera 120, a service providing terminal 130, and the like. These devices are connected via telecommunications lines.

[0022] The linking camera 110 is a camera installed at the gate of the authentication space or the like, and is installed in a position where it is easy to obtain characteristic information of a person. There may be one or more linking cameras 110. The tracking camera 120 is a camera for tracking a person in the authentication space, and is installed on the ceiling or the like so that it is easy to track the person. There may be one or more tracking cameras 120. The service providing terminal 130 is a terminal that provides services to users in the authentication space. There may be multiple service providing terminals 130.

[0023] Fig. 2(b) is a functional block diagram showing each function of the information processing device 100. As illustrated in Fig. 2(b), the information processing device 100 functions as an acquisition unit 10, a feature extraction unit 20, a storage unit 30, a probability calculation unit 40, a position detection unit 50, a determination unit 60, an output unit 70, etc.

[0024] Next, each process executed by the information processing device 100 will be described.

[0025] (ID Linking Process) Fig. 3 is a flowchart showing the ID linking process. When each user enters the authentication space, the acquisition unit 10 acquires an image from the linking camera 110 (step S1). In order to register their own feature information, each user assumes a posture that makes it easy to acquire the feature information. For example, each user faces the linking camera 110 directly.

[0026] Next, the feature extraction unit 20 extracts feature information x_(i, 0), which is identification information for identifying a person, from the image acquired in step S1, and stores the extracted feature information in the storage unit 30 in association with the ID (step S2). Note that "i" is a user number (i=1, 2, ..., I) assigned to each user. Thereafter, execution of the flowchart ends.

[0027] 4 is a diagram illustrating an example of an ID table stored in the storage unit 30. As illustrated in FIG. 4, in the ID table, feature information is linked to each ID. Note that, since each user assumes a posture that makes it easy to obtain feature information, the feature information stored in the storage unit 30 is highly accurate.

[0028] Fig. 5 is a diagram illustrating a location information table stored in the storage unit 30. As illustrated in Fig. 5, in the location information table, the identification information (camera ID) of each tracking camera 120 is linked to the identification information (terminal ID) of each service providing terminal 130. By doing so, it is possible to detect which tracking camera 120 a person appears in and thereby detect which service providing terminal the person is able to receive service from.

[0029] (Service Provision Processing) Fig. 6 is a flowchart showing the service provision processing when providing a service to each user who has registered an ID. As shown in Fig. 6, the acquisition unit 10 acquires images from each tracking camera 120 at a predetermined interval (step S11).

[0030] Next, the feature extraction unit 20 extracts feature information x_(t, j, k) of the kth person appearing in the image acquired by the jth tracking camera 120 at time t (step S12). By executing step S12, feature information of all people appearing in the images acquired by all tracking cameras 120 is extracted.

[0031] Next, the accuracy calculation unit 40 calculates the similarity D_(i, t, j, k) between each piece of feature information x_(i, 0) stored in the storage unit 30 and each piece of feature information x_(t, j, k) extracted in step S12 (step S13). For example, cosine similarity can be used as the similarity. By executing step S13, the similarity between all people appearing in the images acquired by all tracking cameras 120 and the feature information of each ID stored in the storage unit 30 is calculated.

[0032] Next, the accuracy calculation unit 40 calculates an estimated ID value ID_(t, j, k) = argmax_i(D_(i, t, j, k)) of the ID of the kth person captured by the jth tracking camera 120 at time t, and an accuracy value Pr_(t, j, k) = max_i(D_(i, t, j, k)) (step S14). Here, the accuracy value is an index indicating the likelihood that the feature information corresponds to the target person. The accuracy value Pr_(t, j, k) = max_i(D_(i, t, j, k)) means the maximum value of the similarity between the feature information of the IDs stored in the storage unit 30 and the kth person. Therefore, by calculating the accuracy value Pr_(t, j, k), it is estimated which of the IDs stored in the storage unit 30 the kth person corresponds to. The estimated value ID_(t,j,k)=argmax_i(D_(i,t,j,k)) is the estimated ID.

[0033] Next, the position detection unit 50 refers to the position table in the storage unit 30 and detects the service providing terminal 130 linked to the tracking camera 120 that captured the image showing the jth person (step S15).

[0034] Next, the determination unit 60 determines whether the number of service providing terminals 130 detected in step S15 is one (step S16). By executing step S16, it can be determined whether the number of service providing terminals 130 from which the j-th person can receive a service is one or multiple.

[0035] If the result of step S16 is "Yes," the judgment unit 60 judges whether the probability value Pr_(t, j, k) of the jth person is greater than or equal to the probability threshold set for the service providing terminal 130 detected in step S15 (step S17).

[0036] If the determination in step S17 is "Yes," the determination unit 60 grants authorization for the j-th person to use the service providing terminal 130 detected in step S15 (step S18). As a result, the k-th person becomes able to receive the provision of the service authorized by the determination unit 60 using the corresponding service providing terminal 130. Thereafter, execution of the flowchart ends.

[0037] If step S17 returns "No," the output unit 70 causes the service providing terminal 130 detected in step S15 to display a message to improve the accuracy value (step S19). For example, a message to pass through a less crowded area or to move to a location with a good image may be displayed. Alternatively, a message to encourage a change in posture to adopt a more accurate posture may be displayed. Alternatively, a change in posture may be encouraged by varying the size of the displayed characters to guide the person toward or away from the display screen. Execution of the flowchart then ends.

[0038] If the result of step S16 is "No," the judgment unit 60 judges whether the probability value Pr_(t, j, k) of the jth person is greater than or equal to the highest probability threshold among the probability thresholds of the multiple candidate service providing terminals 130 detected in step S15 (step S20).

[0039] If the determination in step S20 is "Yes," step S18 is executed. As a result, the kth person can receive the provision of the service approved by the determination unit 60 using the corresponding service providing terminal 130. If the determination in step S20 is "No," step S19 is executed.

[0040] Note that feature information of a specific person may be extracted from a plurality of images acquired by a plurality of tracking cameras 120. Fig. 7 is a diagram illustrating an example of extracting feature information of a specific person from the plurality of images. As illustrated in Fig. 7, when an ID and feature information are linked and stored in the storage unit 30, the feature information obtained from each camera (Cam1, Cam2, Cam3) may be stored individually.

[0041] When a person whose ID is not yet determined is captured on camera, the person may be compared with the feature information of each ID. For example, as shown in the lower part of Figure 7, for an ID for which there is no feature information of Cam2, the person may be compared with the average feature information of the other cameras (the feature information of Cam1 and Cam2).

[0042] In the above example, the camera ID of each tracking camera 120 is linked with the terminal ID of each service providing terminal 130 to identify the service providing terminal 130 that is ready to receive each person service, but this is not limiting. For example, the relationship between each position in the authentication space and the service providing terminals 130 within a predetermined range (for example, a radius of 10 m) from that position may be stored in the storage unit 30, and the service providing terminal 130 corresponding to the position n of each person detected from the image acquired by each tracking camera 120 may be identified.

[0043] FIG. 8 is a block diagram illustrating an example of the hardware configuration of the information processing device 100. As illustrated in FIG. 8, the information processing device 100 includes a CPU 101, a RAM 102, a storage device 103, and the like. The CPU (Central Processing Unit) 101 is a central processing unit. The RAM (Random Access Memory) 102 is a volatile memory that temporarily stores programs executed by the CPU 101, data processed by the CPU 101, and the like. The storage device 103 is a non-volatile storage device. For example, the storage device 103 may be a read-only memory (ROM), a solid-state drive (SSD) such as a flash memory, or a hard disk driven by a hard disk drive. The CPU 101 executes a determination program stored in the storage device 103 to realize the functions of each unit of the information processing device 100. Note that the functions of each unit of the information processing device 100 may be configured using dedicated circuits or the like.

[0044] In each of the above examples, the accuracy calculation unit 40 is an example of an accuracy calculation unit that calculates an accuracy value indicating the likelihood that identification information that identifies a person detected from an image corresponds to the person. The determination unit 60 is an example of a determination unit that, when acquiring the person's location information, refers to a storage unit that stores an accuracy threshold and location information for each of a plurality of services in association with each other, and determines whether the calculated accuracy value is equal to or greater than the accuracy threshold of a service among the plurality of services that corresponds to the person's location information. The output unit 70 is an example of an output unit that, when it is determined that the accuracy value is not equal to or greater than the accuracy threshold, outputs a message to the person to improve the accuracy value.

[0045] Although the embodiments of the present invention have been described in detail above, the present invention is not limited to such specific embodiments, and various modifications and changes are possible within the scope of the gist of the present invention as described in the claims.

[0046] REFERENCE SIGNS LIST 10 Acquisition unit 20 Feature extraction unit 30 Storage unit 40 Accuracy calculation unit 50 Position detection unit 60 Determination unit 70 Output unit 100 Information processing device 110 Linking camera 120 Tracking camera 130 Service providing terminal 200 Biometric authentication system

Claims

1. The computer When identification information for identifying a person detected from the image is acquired, a probability value indicating the likelihood that the identification information corresponds to the person is calculated; A determination method characterized by executing a process in which, when the person's location information is obtained, a memory unit that stores a probability threshold and location information for each of a plurality of services in association with each other, and determines whether the calculated probability value is greater than or equal to the probability threshold of the service among the plurality of services that corresponds to the person's location information.

2. The computer The method according to claim 1, further comprising the step of: if it is determined that the certainty value is not equal to or greater than the certainty threshold, outputting a message to the person to improve the certainty value.

3. The computer The determination method described in claim 1, characterized in that if there are multiple services among the multiple services that correspond to the person's location information, a process is executed to determine whether the accuracy value is greater than or equal to the highest accuracy threshold among the corresponding services.

4. The computer The determination method according to claim 3, characterized in that, if the certainty value is not determined to be equal to or greater than the highest certainty value, a process is executed in which a message to the person to improve the certainty value is output.

5. The computer The determination method according to claim 1, characterized in that a process is performed in which a probability value indicating the likelihood that the identification information corresponds to the person is calculated using a plurality of different images in which the person is detected.

6. On the computer, When identification information for identifying a person detected from the image is acquired, a probability value indicating the likelihood that the identification information corresponds to the person is calculated; A determination program characterized by executing a process in which, when the person's location information is obtained, the program refers to a memory unit that stores a probability threshold and location information for each of a plurality of services in association with each other, and determines whether the calculated probability value is greater than or equal to the probability threshold of the service among the plurality of services that corresponds to the person's location information.

7. a probability calculation unit that calculates a probability value indicating the likelihood that identification information for identifying a person detected from an image corresponds to the person; An information processing device characterized by comprising: when the person's location information is acquired, a determination unit that refers to a memory unit that stores a probability threshold and location information for each of a plurality of services in association with each other, and determines whether the calculated probability value is greater than or equal to the probability threshold of the service among the plurality of services that corresponds to the person's location information.