Determination method, determination program, and information processing device
The method and device improve person identification accuracy in continuous authentication systems by dynamically switching between open and closed ReID based on user population changes, addressing the accuracy issues in existing methods.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- FUJITSU LTD
- Filing Date
- 2023-05-17
- Publication Date
- 2026-07-23
AI Technical Summary
Existing continuous authentication methods, such as open ReID and closed ReID, face challenges in maintaining accurate person identification when there are changes in the number of individuals within a continuous authentication space, leading to decreased accuracy.
A method and device that dynamically switch between open ReID and closed ReID based on changes in the number of registered users within a predetermined time, using stringent criteria when changes occur and less stringent criteria when no changes are detected, thereby improving identification accuracy.
Enhances the accuracy of person identification by adaptively adjusting authentication methods based on user population changes, reducing false identifications and maintaining high accuracy in varying environments.
Smart Images
Figure 0007894036000001 
Figure 0007894036000002 
Figure 0007894036000003
Abstract
Description
Technical Field
[0001] This case relates to a determination method, a determination program, and an information processing apparatus.
Background Art
[0002] Techniques for extracting feature information using a camera and performing authentication have been disclosed (see, for example, Patent Document 1 or Non-Patent Document 1).
Prior Art Documents
Patent Documents
[0003]
Patent Document 1
Non-Patent Documents
[0004]
Non-Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0005] In recent years, continuous authentication technologies have been developed that do not require repeated authentication actions and allow services to be enjoyed with the authentication status being continued through a single authentication. In continuous authentication, for example, open ReID, closed ReID, etc. are executed.
[0006] Open ReID is a continuous authentication method that accepts a person detected from an image captured by a camera as a new person if they are not identified as a pre-registered user. With Open ReID, the system determines whether the detected person is a registered user each time a person is detected from an image, which may lead to false authentication.
[0007] Closed ReID is a continuous authentication method that does not treat a detected person as a new person even if it is not identified as a registered user. Instead of determining whether a person is a registered user each time a person is detected from an image, it associates them with one of the registered users. Therefore, compared to Open ReID, improved matching accuracy can be expected. However, if new people enter or leave the continuous authentication space, the accuracy of person identification will decrease.
[0008] In one aspect, this invention aims to provide a method for determining a person, a program for determining a person, and an information processing device that can improve the accuracy of person identification. [Means for solving the problem]
[0009] In one embodiment, the determination method involves a computer performing a process that includes referring to a storage unit that stores information about registered users, determining whether the number of registered users stored in the storage unit has changed within a predetermined time, determining whether the detected user is a registered user based on predetermined criteria if a person is detected from an image obtained by photographing a predetermined space, determining whether the detected user is a registered user based on predetermined criteria if it is determined that the detected user is not a registered user and the number of registered users has changed within a predetermined time, and determining whether the detected user is a registered user based on criteria less stringent than the predetermined criteria if it is determined that the detected user is not a registered user and the number of registered users has not changed within a predetermined time. [Effects of the Invention]
[0010] This can improve the accuracy of person identification. [Brief explanation of the drawing]
[0011] [Figure 1] (a) and (b) are diagrams illustrating the entry and exit of people into the authentication space. [Figure 2] (a) is a block diagram illustrating the overall configuration of the biometric authentication system according to Example 1, and (b) is a functional block diagram representing each function of the information processing device. [Figure 3] This is a flowchart illustrating the registration process. [Figure 4] This diagram illustrates a person table stored in the person registration unit. [Figure 5] This flowchart illustrates an example of a population change observation process performed by an information processing device. [Figure 6] This flowchart shows an example of the open ReID process in step S11. [Figure 7] This diagram illustrates a table of users in a specific area, corresponding to a particular area. [Figure 8] This diagram illustrates the difference between the time a particular person disappears from a camera's view and the time they reappear in another camera's view. [Figure 9] This diagram illustrates the number of people table recorded by the number of people change recording unit. [Figure 10] This flowchart illustrates an example of a switching process performed by an information processing unit. [Figure 11] This flowchart illustrates an example of a closed ReID process. [Figure 12] This is a diagram illustrating an overview of an example. [Figure 13] This is a functional block diagram showing the various functions of the information processing device according to Example 2. [Figure 14] This flowchart illustrates the user update process performed by an information processing device. [Figure 15] This is a block diagram illustrating the hardware configuration of an information processing device.
Best Mode for Carrying Out the Invention
[0012] Biometric authentication is a technology for verifying the identity of a person using biometric features such as fingerprints, faces, veins, etc. In biometric authentication, in a scenario where identity verification is required, the biometric feature data for verification obtained by a biometric sensor is compared (matched) with the registered biometric feature data registered in advance, and it is determined whether the similarity is equal to or greater than the identity determination threshold value, thereby performing identity verification. Biometric authentication is used in various fields such as bank ATMs and entry / exit management, and particularly in recent years, it has also begun to be used for cashless payment at supermarkets and convenience stores.
[0013] These biometric authentications are "point" authentications performed at specific authentication spots such as in front of an authentication device. However, in "point" authentication, once the user leaves the authentication spot, the authentication state is interrupted, and in cases where the service is received again or at a place where authentication is performed multiple times, authentication has to be performed each time. Therefore, there is a need for a continuous authentication technology that does not require the authentication action to be performed multiple times and can maintain the authentication state with a single authentication and enjoy the service.
[0014] Here, the outline of the continuous authentication technology will be described. In continuous authentication, mainly the following three types of authentications are performed.
[0015] First, it is the authentication at the time of check-in. At a gate to the continuous authentication space or the like, the user performs highly accurate authentication such as palm vein authentication, and when the authentication is successful, the appearance of the user is photographed by a camera, and the feature information obtained by the camera is linked as registration information to the user's ID.
[0016] <00Next is re-authentication. For example, if a user enters a blind spot of the camera due to another person or a pillar, and the authentication state is interrupted, re-authentication is performed. By performing authentication in this manner, continuous authentication can be achieved.
[0018] In continuous authentication, methods such as open ReID and closed ReID are implemented. Open ReID is a type of continuous authentication where, if a person detected from an image acquired by a camera is not identified as a pre-registered user, it is accepted as a new person. With open ReID, each time a person is detected from an image, it is determined whether or not the detected person is a registered user, which can lead to false authentication. Closed ReID is a type of continuous authentication where, even if a detected person is not identified as a registered user, it is not treated as a new person. Instead of determining whether or not a detected person is a registered user each time a person is detected from an image, it is associated with one of the registered users. Therefore, compared to open ReID, improved matching accuracy can be expected.
[0019] For example, as illustrated in Figure 1(a), it is preferable to use open ReID when the entry and exit of individuals into the continuous authentication space is not controlled. However, as mentioned above, if false authentication occurs, the accuracy of person identification will decrease. As illustrated in Figure 1(b), it is preferable to use closed ReID when there is no entry or exit of individuals into the continuous authentication space. However, if new individuals enter or exit the continuous authentication space, the accuracy of person identification will decrease.
[0020] Therefore, the following embodiments will describe a determination method, a determination program, and an information processing device that can improve the accuracy of person identification. [Examples]
[0021] Figure 2(a) is a block diagram illustrating the overall configuration of the biometric authentication system 200 according to Embodiment 1. As illustrated in Figure 2(a), the biometric authentication system 200 includes an information processing device 100, a linking camera 110, a tracking camera 120, a biosensor 130, and the like. Each of these devices is connected via a telecommunications line.
[0022] The linking camera 110 is a camera installed at a gate or similar location in the continuous authentication space, and is positioned to easily acquire characteristic information of a person. There may be one linking camera 110 or multiple cameras. The biosensor 130 is installed at the gate or similar location and is positioned near the linking camera 110.
[0023] The tracking camera 120 is a camera used to track people in the continuous authentication space and is installed on the ceiling or elsewhere to facilitate tracking. There may be one tracking camera 120 or multiple tracking cameras 120. If multiple tracking cameras 120 are installed, at least a portion of the fields of view of two or more tracking cameras 120 may overlap. Alternatively, the fields of view of each tracking camera 120 may not overlap.
[0024] Figure 2(b) is a functional block diagram representing the various functions of the information processing device 100. As illustrated in Figure 2(b), the information processing device 100 functions as an acquisition unit 11, a person detection unit 12, a feature extraction unit 13, a ReID processing unit 14, a person registration unit 15, a user registration unit 16, an authentication processing unit 17, a person observation unit 18, a person change recording unit 19, a switching unit 20, an output unit 21, and so on.
[0025] Next, we will explain each process performed by the information processing device 100.
[0026] (Registration process) Figure 3 is a flowchart representing the registration process. The acquisition unit 11 acquires biometric data for verification from the biosensor 130 when each user enters the continuous authentication space (step S1). The biometric data in this case is not particularly limited, but could be, for example, fingerprints, veins, or irises, and is biometric data with high authentication accuracy.
[0027] Next, the authentication processing unit 17 determines whether the similarity between the matching biometric data obtained in step S1 and the pre-registered registered biometric data is above a threshold (step S2). If "No" is determined in step S2, the execution of the flowchart ends. In this case, for example, the user restarts from the operation in step S1.
[0028] If "Yes" is determined in step S2, the user's ID is identified. In this case, an image of the user is obtained from the linked camera 110 (step S3). The user will assume a posture that makes it easy to obtain their characteristic information in order to register their own characteristic information. For example, the user will face the linked camera 110 directly.
[0029] Next, the person detection unit 12 detects the person region from the image acquired in step S3 (step S4). Methods for detecting the person region include detecting the person's region by background subtraction, and detecting the person's features from the input image after pre-learning the person's features.
[0030] Next, the feature extraction unit 13 extracts feature information, which is identification information for identifying a person, from the person area acquired in step S4, and registers it as registration information in the person table stored by the person registration unit 15, linked to the identified ID (step S5). Here, the registration information includes clothing colors, physique, facial features, behavioral features, etc. After that, the execution of the flowchart ends. This registration information is feature information for continuous authentication.
[0031] Figure 4 is an example of a person table stored by the person registration unit 15. As illustrated in Figure 4, in the person table, each ID is linked to a name, registration information, etc.
[0032] (Processing to observe changes in the number of people) Figure 5 is a flowchart showing an example of the population change observation process performed by the information processing device 100. The population change observation process is performed at predetermined intervals. An example of the population change observation process will be explained with reference to Figure 5.
[0033] First, the ReID processing unit 14 performs an open ReID process for a specific area (step S11). Here, the specific area may be part or all of the continuous authentication space described above. For example, the continuous authentication space may be divided into multiple specific areas. Since the position and shooting range of the tracking camera 120 are predetermined, the correspondence between each specific area and each tracking camera 120 may be predetermined. In the following explanation, we will focus on one specific area.
[0034] Figure 6 is a flowchart showing an example of the open ReID process in step S11. First, the acquisition unit 11 acquires images from each tracking camera 120 (step S21).
[0035] Next, the person detection unit 12 detects the person region from each image acquired in step S21 (step S22). Methods for detecting the person region include detecting the region of a moving object by background subtraction, and detecting person features from the input image after pre-learning person features.
[0036] Next, the feature extraction unit 13 extracts feature information from the image for each person region detected in step S22 and stores the feature information corresponding to the person region (step S23). Here, the feature information from the image includes the color of clothing, body type, facial features, behavioral features, etc.
[0037] Next, the ReID processing unit 14 compares the characteristic information of a person detected simultaneously between different tracking cameras 120, or compares the characteristic information of a person detected by the same tracking camera 120 at a predetermined time difference, and determines that the person is the same person if the similarity between the characteristic information is above a threshold (step S24). By executing step S24, the target of the ReID processing can be narrowed down.
[0038] Next, the ReID processing unit 14 identifies which ID user in the person table of the person registration unit 15 corresponds to each person narrowed down in step S24 (step S25). For example, the ReID processing unit 14 identifies the person to be the user of the registration information ID that has the highest similarity between the feature information extracted in step S22 and the registration information registered in the person table.
[0039] Next, the ReID processing unit 14 registers the users identified from the results of step S25 into the specific area user table and stores them in the user registration unit 16 (step S26). Figure 7 is an example of a specific area user table corresponding to a specific area. In the example in Figure 7, people A, B, C, and D are identified in the specific area. Note that registration information corresponding to each person may be associated in the specific area user table. This registration information can be obtained from the person table. After that, the execution of the flowchart ends. This process makes it possible to count the number of people in the specific area. In addition, people detected from the image but not registered in the specific area user table can be registered as new people in the specific area user table.
[0040] Refer to Figure 5 again. After step S11 is executed, the people observation unit 18 counts the number of people in the specific area (step S12). The people observation unit 18 counts the number of people in the specific area by referring to the specific area user table stored in the user registration unit 16.
[0041] Next, the people observation unit 18 observes the passage of time (step S13). For example, it can count the time elapsed since the first time step S13 is performed.
[0042] Next, the people observation unit 18 determines whether a certain period T1 has elapsed (step S14). If "No" is determined in step S14, the process is repeated from step S11. For example, the unit observes the period during which the same person is detected by different tracking cameras 120 within a specific area, and sets the longest period as the certain period T1. Alternatively, the time it takes to travel the greatest distance between the installed tracking cameras 120 is set as the certain period T1. For example, the difference between the time when a specific person is no longer detected by one camera and the time when that specific person begins to be detected by other cameras is set as the certain period T1 (e.g., 10 seconds). For example, as illustrated in Figure 8, for the specific person, the difference between the time when they disappear from view of camera 002 (exit time) and the time when they appear in view of camera 003 (appearance time) is set as the certain period T1.
[0043] If "Yes" is determined in step S14, the number of users observation unit 18 stores the number of users (step S15).
[0044] Next, the people observation unit 18 records the change of users (step S16). For example, the people observation unit 18 refers to the specific area user table of the user registration unit 16 and records whether or not there has been a change in the ID registered in the specific area user table in the people change recording unit 19's people table. Figure 9 is an example of the people table recorded by the people change recording unit 19.
[0045] (Switching process) Figure 10 is a flowchart illustrating an example of a switching process performed by the information processing device 100. The switching process is executed at predetermined intervals. Furthermore, the switching process is executed in parallel with the person change observation process. An example of the switching process will be explained with reference to Figure 10.
[0046] First, the switching unit 20 refers to the number of people change recording unit 19 to confirm the change in users (step S31).
[0047] Next, the switching unit 20 determines whether or not the user has changed (step S32). If "Yes" is determined in step S32, the process is repeated from step S31.
[0048] If "No" is determined in step S32, the switching unit 20 compares the number of people (step S33).
[0049] Next, the switching unit 20 determines whether the number of people has changed within the specified period T2 (step S34). For example, it determines whether the number of people has increased or decreased.
[0050] If "Yes" is determined in step S34, the switching unit 20 specifies open ReID processing (step S35). As a result, the ReID processing unit 14 executes the open ReID processing described in Figure 6. If "No" is determined in step S34, the switching unit 20 specifies closed ReID processing (step S36). As a result, the ReID processing unit 14 executes closed ReID processing. For example, the output unit 21 outputs information to an external device, etc., regarding whether open ReID processing was specified or closed ReID processing was executed. This allows, for example, an administrator to know which type of ReID processing is being performed.
[0051] Figure 11 is a flowchart illustrating an example of closed ReID processing. The flowchart in Figure 11 is repeated at predetermined intervals. First, the acquisition unit 11 acquires images from each tracking camera 120 (step S41).
[0052] Next, the person detection unit 12 detects the person region from each image acquired in step S41 (step S42). Methods for detecting the person region include detecting the region of a moving object by background subtraction, and detecting person features from the input image after pre-learning person features.
[0053] Next, the feature extraction unit 13 extracts feature information from the image for each person region detected in step S42 and stores the feature information corresponding to the person region (step S43). Here, the feature information from the image includes the color of clothing, body type, facial features, behavioral features, etc.
[0054] Next, the ReID processing unit 14 compares the characteristic information of individuals detected at predetermined time differences between different tracking cameras 120 or by the same tracking camera 120. If the similarity between the characteristic information is above a threshold, it determines that the individuals are the same person (step S44). Furthermore, the ID and name of each individual can be identified by inheriting the ID and name of each individual identified in the open ReID processing shown in Figure 6. This allows each individual to be tracked.
[0055] Next, the ReID processing unit 14 determines whether or not there are any individuals who were not determined to be the same person (step S45). If "No" is determined in step S45, the execution of the flowchart ends.
[0056] If "Yes" is determined in step S45, the ReID processing unit 14 associates individuals stored in the specific area user table stored by the user registration unit 16 with individuals who were not determined to be the same person (step S46). After that, the execution of the flowchart ends.
[0057] According to this embodiment, as illustrated in Figure 12, if a change in the number of people occurs in a specific area, open ReID processing is performed. This makes it possible to identify people entering and leaving the specific area. On the other hand, if there is no change in the number of people in the specific area, it is treated as a pseudo-closed room with no entry or exit, or an area where entry and exit are controlled, and closed ReID processing is performed. This improves the accuracy of the matching process. As described above, this embodiment can improve the accuracy of person identification. Furthermore, in closed ReID, since the matching is performed on a limited number of people who have been authenticated once and counted within the specific area, high accuracy and reduced processing time can be expected.
[0058] For example, in large spaces such as airports, authentication areas may be set up by dividing the space into designated areas. Furthermore, if entry and exit to the authentication area are unrestricted, even a person staying within the authentication area may be misidentified as a new person due to overlap or angle effects. Applying this embodiment in such cases can suppress the decrease in authentication accuracy for individuals staying within the authentication area.
[0059] In this embodiment, the open ReID process identifies individuals using strict criteria by comparing the registration information registered in the specific area user table with the characteristic information of individuals detected from the image. The closed ReID process uses looser criteria for identifying individuals than the open ReID process, as it does not involve matching the registration information registered in the specific area user table with the characteristic information of individuals detected from the image. However, the embodiment is not limited to this.
[0060] For example, in closed ReID processing, measures such as lowering the similarity threshold when matching with registered information may be taken. By doing so, it is possible to avoid identifying a new person and improve the accuracy of the matching process. [Examples]
[0061] Example 2 describes the updating of users who utilize a specific area. The differences from Example 1 will be explained.
[0062] Figure 13 is a functional block diagram showing the functions of the information processing device 100a according to Embodiment 2. As illustrated in Figure 13, the information processing device 100a further includes a user update unit 22.
[0063] (User update process) Figure 14 is a flowchart illustrating the user update process performed by the information processing device 100a. The user update process is executed in parallel with the user change observation process and switching process described above. An example of the user update process will be explained with reference to Figure 14.
[0064] The user update unit 22 compares the number of users counted in step S12 of Figure 5 with the number of users registered in the specific area user table (step S51).
[0065] Next, the user update unit 22 determines in step S51 whether the number of registered users is greater than the number of people counted (step S52). If "No" is determined in step S52, the process is repeated from step S51.
[0066] If "Yes" is determined in step S52, check the exit time in the specific area user table (step S53).
[0067] Next, the user update unit 22 determines whether the specified period T2 has exceeded the time of departure (step S54). If "No" is determined in step S54, the process is repeated from step S51.
[0068] If "Yes" is determined in step S54, the user update unit 22 updates the specific area user table (step S55). For example, for individuals registered in the specific area user table who were not identified in the specific area, they are either deleted or their current time is registered as their exit time. The process is then executed again from step S51.
[0069] Figure 15 is a block diagram illustrating the hardware configuration of the information processing devices 100 and 100a. As illustrated in Figure 15, the information processing devices 100 and 100a include a CPU 101, RAM 102, and a storage device 103. 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, etc. The storage device 103 is a non-volatile storage device. As the storage device 103, for example, a ROM (Read Only Memory), a solid-state drive (SSD) such as flash memory, or a hard disk driven by a hard disk drive can be used. The functions of each part of the information processing devices 100 and 100a are realized by the CPU 101 executing the judgment program stored in the storage device 103. Note that the functions of each part of the information processing devices 100 and 100a may be configured by dedicated circuits, etc.
[0070] In each of the above examples, the user registration unit 16 is an example of a storage unit that stores information about registered users. The number of people observation unit 18 is an example of a determination unit that refers to the storage unit and determines whether the number of registered users stored in the storage unit has changed within a predetermined time. The switching unit 20 and the ReID processing unit 14 are an example of a processing unit that, when a person is detected from an image obtained by photographing a predetermined space, determines whether the detected person is a registered user based on predetermined criteria, and if it is determined that the detected person is not a registered user and the number of registered users has changed within a predetermined time, stores the detected person in the storage unit as a new registered user. If it is determined that the detected person is not a registered user and the number of registered users has not changed within a predetermined time, determines whether the detected person is a registered user based on criteria less stringent than the predetermined criteria.
[0071] Although embodiments of the present invention have been described in detail above, the present invention is not limited to these specific embodiments, and various modifications and changes are possible within the scope of the gist of the present invention as described in the claims. [Explanation of symbols]
[0072] 11 Acquisition Department 12 Person Detection Unit 13 Feature Extraction Unit 14 ReID Processing Unit 15. Person Registration Department 16. User Registration Section 17 Authentication Processing Unit 18. Personnel Observation Department 19. Personnel Change Recording Department 20 Switching section 21 Output section 22 User Update Department 100 Information Processing Devices 110 Camera with strap 120 Tracking Cameras 130 Biosensors 200 Biometric Authentication Systems
Claims
1. The system refers to a memory unit that stores information about registered users and determines whether the number of registered users stored in the memory unit has changed within a predetermined time period. If a person is detected in an image obtained by photographing a predetermined space, it is determined whether the detected person is the registered person or not based on predetermined criteria. If it is determined that the detected person is not a registered person, and it is determined that the number of registered persons has changed within a predetermined time, the detected person is stored in the storage unit as a new registered person. If it is determined that the detected person is not a registered person, and it is determined that the number of registered persons has not changed within a predetermined time period, then it is determined whether or not the detected person is a registered person based on criteria less stringent than the predetermined criteria. A determination method characterized by the processing being performed by a computer.
2. The determination method according to claim 1, characterized in that if it is determined that the detected person is not a registered person, and it is determined that the number of registered persons has not changed within a predetermined time, the computer performs a process of associating the detected person with one of the registered persons stored in the storage unit.
3. The determination method according to claim 1, characterized in that the computer performs a process of comparing the number of people counted from an image obtained by photographing the predetermined space with the number of registered persons stored in the storage unit, and if the number of registered persons stored in the storage unit is greater than the number of people counted, reducing the number of registered persons stored in the storage unit.
4. The determination method according to claim 1, characterized in that the case in which the number of registered persons changes within a predetermined time period is a case in which the number of registered persons increases within a predetermined time period.
5. The determination method according to claim 1, characterized in that the case in which the number of registered persons changes within a predetermined time period is a case in which the number of registered persons decreases within a predetermined time period.
6. The system refers to a memory unit that stores information about registered users and determines whether the number of registered users stored in the memory unit has changed within a predetermined time period. If a person is detected in an image obtained by photographing a predetermined space, it is determined whether the detected person is the registered person or not based on predetermined criteria. If it is determined that the detected person is not a registered person, and it is determined that the number of registered persons has changed within a predetermined time, the detected person is stored in the storage unit as a new registered person. If it is determined that the detected person is not a registered person, and it is determined that the number of registered persons has not changed within a predetermined time period, then it is determined whether or not the detected person is a registered person based on criteria less stringent than the predetermined criteria. A determination program characterized by having a computer perform the processing.
7. The determination program according to claim 6, characterized in that, if it is determined that the detected person is not a registered person, and it is determined that the number of registered persons has not changed within a predetermined time, the computer is instructed to perform a process of associating the detected person with one of the registered persons stored in the storage unit.
8. The determination program according to claim 6, characterized in that it causes the computer to perform a process of comparing the number of people counted from an image obtained by photographing the predetermined space with the number of registered persons stored in the storage unit, and if the number of registered persons stored in the storage unit is greater than the number of people counted, it reduces the number of registered persons stored in the storage unit.
9. The determination program according to claim 6, characterized in that the case in which the number of registered persons changes within a predetermined time period is a case in which the number of registered persons increases within a predetermined time period.
10. The determination program according to claim 6, characterized in that the case in which the number of registered persons changes within a predetermined time period is a case in which the number of registered persons decreases within a predetermined time period.
11. A determination unit that refers to a storage unit that stores information about registered users and determines whether the number of registered users stored in the storage unit has changed within a predetermined time period, An information processing device comprising: a processing unit that, when a person is detected from an image obtained by photographing a predetermined space, determines whether the detected person is a registered person based on predetermined criteria; if it is determined that the detected person is not a registered person and it is determined that the number of registered people has changed within a predetermined time, stores the detected person as a new registered person in the storage unit; and if it is determined that the detected person is not a registered person and it is determined that the number of registered people has not changed within a predetermined time, determines whether the detected person is a registered person based on criteria less stringent than the predetermined criteria.
12. The information processing apparatus according to claim 11, characterized in that, if the processing unit determines that the detected person is not a registered person and that the number of registered persons has not changed within a predetermined time, it associates the detected person with one of the registered persons stored in the storage unit.
13. The information processing apparatus according to claim 11, further comprising an update unit that compares the number of people counted from an image obtained by photographing the predetermined space with the number of registered persons stored in the storage unit, and reduces the number of registered persons stored in the storage unit if the number of registered persons stored in the storage unit is greater than the number of people counted.
14. The information processing apparatus according to claim 11, characterized in that the case in which the number of registered persons changes within a predetermined time period is a case in which the number of registered persons increases within a predetermined time period.
15. The information processing apparatus according to claim 11, characterized in that the case in which the number of registered users changes within a predetermined time period is a case in which the number of registered users decreases within a predetermined time period.