Verification device, verification system, and verification method

Abstract

To provide a verification device, a verification system, and a verification method capable of improving a processing speed of verification using facial images of people passing through a specific region.SOLUTION: In a facial recognition system, a verification device for regulating the flow of people includes: a processing unit that performs second facial image verification by using a first candidate facial image narrowed down from a plurality of facial images included in a facial image database and a facial image included in second images obtained by imaging a second region on the basis of the result of first facial image verification using facial images included in first images obtained by imaging a first region and the plurality of facial images included in the facial image database in a route where there is a flow of people moving from the first region to the second region located upstream of a position where the flow of people should be regulated; and a communication unit that outputs the result of the second facial image verification.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present disclosure relates to a collation device, a collation system, and a collation method.

Background Art

[0002] There is known a technique for managing the entry and exit of people passing through gates installed at stations, airports, etc. by face authentication. Patent Document 1 discloses a technique for realizing smooth passage of people through a gate. The technique of Patent Document 1 extracts feature amounts of objects in a captured image obtained by capturing an area in front of the gate before passage, and performs a collation determination based on the registered collation information (information regarding the feature amounts of people, etc.) and the estimated distance from a person approaching the gate to the gate. According to the technique of Patent Document 1, face authentication is performed after confirming whether the estimated distance is a distance suitable for collation.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Since the time for a person to pass through a gate is about several seconds, when collating (or authenticating) a person passing through a gate by a face image, processing in a short time is expected.

[0005] Non-limiting embodiments of the present disclosure contribute to providing a collation device, a collation system, and a collation method that can improve the processing speed of collation (hereinafter sometimes abbreviated as "face image collation" or "face image authentication") using a face image of a person passing through a specific area such as a gate.

Means for Solving the Problems

[0006] The verification device according to an embodiment of the present disclosure is a verification device for regulating the flow of people. In a path where there is a flow of people from a first area to a second area located upstream of the position where the flow of people should be regulated, based on the result of a first face image verification using the face images included in the first image taken of the first area and the plurality of face images included in the face image database, a first candidate face image narrowed down from the plurality of face images included in the face image database, and the face image included in the second image taken of the second area, a second face image verification is performed by a processing unit, and a communication unit outputs the result of the second face image verification.

[0007] The verification system according to an embodiment of the present disclosure is a verification system for regulating the flow of people. It includes a first camera that captures the first area in a flow of people from a first area to a second area located upstream of the position where the flow of people should be regulated, a second camera that captures the second area, a first verification device that performs a first face image verification using the face images included in the first image captured by the first camera and the plurality of face images included in the face image database, a first candidate face image narrowed down from the plurality of face images included in the face image database based on the result of the first face image verification, and a second verification device that performs a second face image verification using the face image included in the second image captured by the second camera.

[0008] The verification method according to an embodiment of the present disclosure is a verification method for regulating the flow of people. In a path where there is a flow of people from a first area to a second area located upstream of the position where the flow of people should be regulated, based on the result of a first face image verification using the face images included in the first image taken of the first area and the plurality of face images included in the face image database, a first candidate face image narrowed down from the plurality of face images included in the face image database, and the face image included in the second image taken of the second area, a second face image verification is performed, and the result of the second face image verification is output.

[0009] Note that these general or specific aspects may be implemented in a system, apparatus, method, integrated circuit, computer program, or recording medium, or may be implemented in any combination of a system, apparatus, method, integrated circuit, computer program, and recording medium.

Advantages of the Invention

[0010] According to an embodiment of the present disclosure, the processing speed of face image verification for a person passing through a specific area can be improved.

[0011] Further advantages and effects in an embodiment of the present disclosure will be clarified from the specification and drawings. Such advantages and / or effects are provided by some embodiments and the features described in the specification and drawings respectively, but not all of them are necessarily provided to obtain one or more identical features.

Brief Description of the Drawings

[0012]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Mode for Carrying Out the Invention

[0013] Hereinafter, preferred embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. In the present specification and drawings, components having substantially the same function are denoted by the same reference numerals, and redundant description is omitted.

[0014] (Embodiment) FIG. 1 is a diagram showing a configuration example of a face authentication system according to the present disclosure. The face authentication system 100 according to the present embodiment is, for example, a system that controls a gate (such as an entrance gate or a ticket gate) installed at an entrance / exit of a facility such as an airport, a station, or an event venue. In the face authentication system 100 according to the present embodiment, illustratively, the management of entry and exit of users who use the facility is executed by face authentication. For example, when a user passes through the gate and enters the facility, it is determined by face authentication whether the user is a person permitted to enter the facility. Also, when a user passes through the gate and exits the facility, it is determined by face authentication whether the user is a person permitted to exit the facility. Note that "face authentication" may be regarded as a concept included in "verification using a face image".

[0015] The face authentication system 100 includes a gate control device 20 and a face authentication server 200. The face authentication system 100 also includes a plurality of cameras 1 for face photography, a QR code (registered trademark) reader 2, a passage management photoelectric sensor 3, an opening / closing door mechanism 4, an entry guidance indicator 5, a passage guidance LED (Light Emitting Diode) 6, and a guidance display 7. The face authentication system 100 also includes a speaker 8, an interface board 9, an interface driver 10, a network hub 30, and the like.

[0016] The gate control device 20 is connected to the network hub 30 and can communicate with the server 200 via the network hub 30 and the network 300. The server 200 performs processing related to face authentication. Therefore, the server 200 may be referred to as the face authentication server 200. The gate control device 20 is a device that controls gates (such as entrance gates and ticket gates) installed in facilities such as airports, railway stations, and event venues. The gate control device 20 controls the opening and closing door mechanism 4 of the gate. For example, for a person permitted by face authentication, the gate is opened. On the other hand, for a person who fails in face authentication, the gate is closed.

[0017] In face authentication, for example, information on face images of hundreds of thousands to tens of millions of individuals is used. This information is recorded in, for example, the face authentication server 200. Hereinafter, the information used for face authentication may be referred to as "authentication information" or "collation information". For example, the authentication information may be registered in the face authentication server 200 in advance through the usage procedures of users who use the face authentication service.

[0018] The collation device 21 is communicably connected to the face authentication server 200 via the network 300. The collation device 21 collates the face image of a person passing through the gate with the face images of the population included in the registered authentication information to authenticate the person passing through the gate.

[0019] Collation means determining whether the pre-registered face image matches the face image of the person passing through the gate, or whether the pre-registered face image and the face image of the person passing through the gate are face images of the same person, by comparing the registered face image with the face image of the person passing through the gate.

[0020] On the other hand, authentication means proving to the outside (for example, the gate) that a person with a face image that matches the pre-registered face image is the person himself / herself (in other words, a person who may be permitted to pass through the gate).

[0021] However, in the present disclosure, "collation" and "authentication" may be used interchangeably.

[0022] For example, the collation process compares the feature points of the face images for each individual registered in advance with the feature points extracted from the face image detected by the face detection process to identify who the face is in the image data. The gate control device 20 controls the gate (for example, the opening and closing operation of the opening and closing door mechanism 4) according to the result of this authentication. Note that the collation device 21 only needs to be arranged so as to be communicable with the face authentication server 200, and may be incorporated inside the gate control device 20, or may be provided outside the gate control device 20.

[0023] The QR code reader 2 reads a QR code including information for identifying a person passing through the gate. For example, among the people passing through the gate, those who perform entry and exit management without using face authentication have the QR code reader 2 read the QR code to perform authentication.

[0024] The passage management photo - electric sensor 3 detects whether a person has entered the gate and whether a person permitted to pass through the gate has finished passing through the gate. For example, the passage management photo - electric sensor 3 may be provided at a plurality of positions including a location for detecting whether a person has entered the gate and a location for detecting whether a person has finished passing through the gate. The passage management photo - electric sensor 3 is connected to the gate control device 20 via, for example, the interface board 9. Note that the method for detecting the entry and passage of a person is not limited to the method using a photo - electric sensor, and other methods such as monitoring the movement of a person photographed by a camera installed on the ceiling or the like can also be realized. That is, the photo - electric sensor is an example of a sensor for passage management, and other sensors may be used.

[0025] The opening and closing door mechanism 4 is connected to the gate control device 20 via, for example, the interface board 9.

[0026] The entry guidance indicator 5 notifies whether passage through the gate 400 is permitted. The entry guidance indicator 5 is connected to the gate control device 20 via, for example, the interface driver 10.

[0027] The passage guidance LED 6 emits light in a color corresponding to the state of the gate 400, for example, to indicate whether the gate 400 is in a passable state.

[0028] The guidance display 7 displays, for example, information regarding passage permission or not.

[0029] The speaker 8 generates a sound indicating passage permission or not, for example.

[0030] Next, with reference to FIG. 2, the hardware configurations of the face authentication server 200 and the collation device 21 will be described. FIG. 2 is a diagram showing an example of the hardware configurations of the face authentication server 200 and the collation device 21.

[0031] The face authentication server 200 includes a processor 601, a memory 602, and an input / output interface 603 used for transmitting various information. The processor 601 is an arithmetic device such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit). The memory 602 is a storage device realized using, for example, a RAM (Random Access Memory) or a ROM (Read Only Memory). The processor 601, the memory 602, and the input / output interface 603 are connected to a bus 604 and perform the transfer of various information via the bus 604. The processor 601 realizes the functions of the face authentication server 200 by, for example, reading programs, data, etc. stored in the ROM onto the RAM and executing the processing.

[0032] The verification device 21 includes a processor 701, a memory 702, and an input / output interface 703 used for transmitting various information. The processor 701 is an arithmetic device such as a CPU or a GPU. The memory 702 is a storage device realized using a RAM, a ROM, etc. The processor 701, the memory 702, and the input / output interface 703 are connected to a bus 704, and various information is transferred via the bus 704. The processor 701 realizes the functions of the verification device 21 by, for example, reading programs, data, etc. stored in the ROM onto the RAM and executing processing.

[0033] Next, the functions of the face authentication server 200 and the verification device 21 will be described with reference to FIG. 3, and the arrangement example of the camera 1 will be described with reference to FIG. 4. FIG. 3 is a diagram showing a functional configuration example of the face authentication server 200 and the verification device 21. FIG. 4 is a diagram showing an installation example of a plurality of cameras 1 at the gate.

[0034] The gate 400 includes, for example, three cameras 1 (camera 1-1, camera 1-2, and camera 1-3).

[0035] The three cameras 1 each photograph a person moving toward the gate 400 in the traveling direction X of the arrow X in FIG. 4. In FIG. 4, the arrow X indicates a path along which there is a flow of people from the region A1 (first region) via the region A2 (third region) to the region A3 (second region), and at least a part of the gate 400 is arranged in the region A3.

[0036] The camera 1-1 photographs the face of a person existing in the region A1 at a position separated from the gate 400 by a certain distance. The region A1 is provided upstream of the gate 400 in the traveling direction X. The region A1 is, for example, a region from a position 1.5 m away from the support part T that supports the cameras 1-1 and 1-2 of the gate 400 to a position 3.0 m away from the support part T. The camera 1-1 photographs a person existing in the region A1 that is relatively far from the gate 400. Hereinafter, the camera 1-1 may be referred to as a "long-distance camera". The image captured by the camera 1-1 is input to the processing unit 102.

[0037] Cameras 1-3 are second cameras that capture the faces of people present in area A3, which is closer to gate 400 than area A1. Area A3 is, for example, an area extending 50 cm in the direction opposite to the traveling direction X of people from support T. Hereinafter, cameras 1-3 may be referred to as "close-range cameras".

[0038] Cameras 1-2 are third cameras that capture the faces of people present in area A2, which is between area A1 and area A3. Area A2 is, for example, an area extending from 1.5 m to 50 cm from support T. Hereinafter, cameras 1-2 may be referred to as "mid-range cameras".

[0039] Note that the imaging ranges of these cameras 1 are not limited to the above examples. For example, at least a part of the imaging ranges of each camera 1 may overlap with each other. For example, the area captured by camera 1-3, which is a close-range camera, is not limited to the range of area A3, and may include, for example, the entire range or a partial range of area A2 within the range of area A3. Also, in FIG. 4, an example where the imaging ranges of each camera 1 are adjacent in the traveling direction (arrow X) is shown, but there may be a gap between the imaging ranges of each camera 1.

[0040] However, since area A3 is an area for authenticating people who are about to pass through gate 400, it may be a downstream area from the location where it is determined that a person has entered gate 400. For example, when gate 400 determines whether a person has entered using passage management photoelectric sensor 3, area A3 is set as the downstream area from the location where passage management photoelectric sensor 3 detects the entry of a person.

[0041] Also, the imaging ranges of cameras 1 illustrated in FIG. 4 are merely conceptually representative of the ranges within which each camera 1 can acquire clear images to the extent that face verification can be performed, such as focal length and angle of view, and it is not intended to exclude the possibility that an image corresponding to a portion outside the said area may be captured.

[0042] Also, the installation position of the camera 1 is not limited to the above example. For example, instead of being attached to the gate 400, the long-distance camera (camera 1-1) may be provided at a position away from the gate 400 and may capture the area A1. Also, for example, instead of being attached to the gate 400, the medium-distance camera (camera 1-2) may be provided at a position away from the gate 400 and may capture the area A2. Also, for example, instead of being attached to the gate 400, the short-distance camera (camera 1-3) may be provided at a position away from the gate 400 and may capture the area A3.

[0043] Also, instead of the camera 1 attached to the gate 400, for example, the face authentication server 200 and the verification device 21 may use images captured by cameras for other purposes such as surveillance cameras.

[0044] Also, the shooting frame rate, the number of shootings (the number of times of face image recording), the maximum number of face detections, etc. of these cameras are set according to the type of the gate 400, the installation location of the camera, etc.

[0045] The verification device 21 includes a communication unit 101 that communicates with the face authentication server 200 via the network 300, a buffer 103 that temporarily records various information, and a processing unit 102. The processing unit 102 performs processes such as face authentication and face verification of people who can pass through the gate 400.

[0046] The face authentication server 200 includes a communication unit 202 that communicates with the verification device 21 via the network 300, a face registration database (DB) 203 that manages authentication information, and a processing unit 201. The authentication information includes, for example, information on face images of each of hundreds of thousands to tens of millions of users.

[0047] Next, the operation overview and operation details of the face authentication server 200 and the verification device 21 will be described.

[0048] FIG. 5 is a diagram for explaining the outline of the operation of the face authentication system 100. The face authentication server 200 detects the area of a person's face (face image) from an image captured by a long-distance camera, and narrows down the collation candidates from the face images included in the face registration DB 203 by collating them with the face images included in the face registration DB 203. Hereinafter, the process of narrowing down the collation candidates using the image captured by the long-distance camera may be described as "long-distance narrowing-down search". In FIG. 5, the long-distance narrowing-down search corresponds to the first narrowing-down search (primary narrowing-down search).

[0049] For example, the face authentication server 200 calculates a score indicating the similarity between two face images, and narrows down the collation candidates based on the calculated score. The similarity between two face images indicates the likelihood that the two face images are of the same person.

[0050] For example, the face authentication server 200 calculates the score between the face image detected from the image captured by the long-distance camera and each of the face images included in the face registration DB 203. Then, the face authentication server 200 buffers the N 1 highest-scoring face images (N 1 is an integer of 1 or more) into the collation candidate list ML. For example, in FIG. 5, an example where N 1 = 6 is shown. Note that the process of detecting a person's face image from the image captured by the long-distance camera may be executed by the collation device 21.

[0051] As a result of the long-distance narrowing-down search, the collation candidate list ML includes, for example, face images (candidate face images) narrowed down from the face images in the face registration DB 203. For example, in FIG. 5, for each of the six face images captured by the long-distance camera, six candidate face images are included. Then, the face authentication server 200 transmits the collation candidate list ML to the collation device 21. The collation candidate list ML is transmitted to the collation device 21. The collation candidate list ML is an example of collation candidates narrowed down using the face image captured by the long-distance camera.

[0052] The matching device 21 detects a human face image from an image captured by the medium-distance camera, and narrows down the matching candidates from the face images included in the matching candidate list ML by matching the detected face image with the face images included in the matching candidate list ML. Hereinafter, the process of narrowing down the matching candidates using the image captured by the medium-distance camera may be described as medium-distance narrowing search. In FIG. 5, the medium-distance narrowing search corresponds to the second narrowing search (secondary narrowing search).

[0053] For example, the matching device 21 calculates a score between the face image detected from the image captured by the medium-distance camera and each of the candidate face images included in the matching candidate list ML for each of the candidate face images included in the matching candidate list ML. Then, the matching device 21 buffers the N 2 number (N 2 is an integer of 1 or more) of candidate face images with higher scores into the matching candidate list SL. For example, in FIG. 5, an example of N 2 = 2 is shown. Note that N 2 may be smaller than N 1 .

[0054] As a result of the medium-distance narrowing search, the matching device 21 acquires N 2 number of candidate face images that are matching candidates from the matching candidate list ML and buffers them into the matching candidate list SL. In the example of FIG. 5, the matching candidate list SL includes, for example, three candidate face images for each of the face images of two persons captured by the medium-distance camera. Note that when the narrowing search cannot be performed, the matching device 21 may request the face authentication server 200 to perform the narrowing search.

[0055] Next, the matching device 21 performs face authentication processing by matching the face image corresponding to the face captured by the short-distance camera with the matching candidate list SL. When the face authentication fails, the matching device 21 requests the face authentication server 200 to perform the narrowing search.

[0056] In this way, before a person enters the gate 400, the face recognition system 100 performs narrowing down of collation candidates. By narrowing down the collation candidates before the person enters the gate 400, in the face recognition process performed when the person enters the gate 400, the collation candidates are narrowed, so that the face recognition process can be speeded up.

[0057] In addition, in FIG. 5, the collation candidate list ML may include the results of multiple long-distance narrowing-down searches. For example, the results of long-distance narrowing-down searches performed for each of the images captured by the long-distance camera at multiple time points may be included in the collation candidate list ML.

[0058] Also, in FIG. 5, the collation candidate list SL may include the results of multiple medium-distance narrowing-down searches. For example, the results of long-distance narrowing-down searches performed for each of the images captured by the medium-distance camera at multiple time points may be included in the collation candidate list SL.

[0059] In addition, the information (for example, face images) included in the collation candidate list ML and the collation candidate list SL may be deleted after a predetermined time has elapsed since the information was added to the list. In addition, as a condition for deleting the information from the list, for example, it may be considered that it is detected by the passage management photo sensor 3 that the person corresponding to the information has finished passing through the gate. Also, when the gate 400 is for managing entry and exit to and from a closed space (such as a building or public transportation), the information may be deleted at the timing when it is detected that the person corresponding to the information has exited the closed space.

[0060] FIG. 6 is a flowchart for explaining an operation example of the face recognition system 100. The collation device 21 detects a face from an image captured by the long-distance camera (step S1). The collation device 21 transmits a request for long-distance face narrowing search to the face authentication server 200 (step S2). In the request for long-distance face narrowing search, for example, the collation device 21 transmits the face image detected from the image captured by the long-distance camera to the face authentication server 200. Note that the collation device 21 may transmit the data of the face image, or may extract the data regarding the feature points used for the search from the data of the face image and transmit the extracted data. When transmitting the data of the face image, since the data of the face image can be processed by the face authentication server 200, face collation can be performed by an arbitrary collation method. On the other hand, when extracting and transmitting the data regarding the feature points, the capacity of the data to be transmitted can be suppressed.

[0061] In step S3, the processing unit 201 of the face authentication server 200 waiting for receiving the search request receives the request (step S4) and executes a long-distance narrowing search (step S5). The processing unit 201 buffers it in the collation candidate list ML (step S7). The processing unit 201 transmits the collation candidate list ML to the collation device 21 (step S8).

[0062] The collation device 21 detects a face image from the image captured by the medium-distance camera (step S9). The collation device 21 executes a medium-distance narrowing search by collating the face image detected from the image captured by the medium-distance camera with the candidate face images in the collation candidate list ML (step S10). The collation device 21 buffers the candidate face images narrowed down from the collation candidate list ML in the collation candidate list SL (step S11).

[0063] The collation device 21 detects a face image from the image captured by the short-distance camera (step S12). The collation device 21 performs face authentication processing by collating the face image detected from the image captured by the short-distance camera with the candidate face images in the collation candidate list SL (step S13).

[0064] Specifically, the verification device 21 verifies the face image detected from the image captured by the short-distance camera with the candidate face images in the verification candidate list SL. As a result of the verification, if the face image detected from the image captured by the short-distance camera corresponds to any one of the face images in the verification candidate list SL, the verification device 21 determines that the person captured by the short-distance camera can pass through the gate 400.

[0065] As a result of the verification, if the face image detected from the image captured by the short-distance camera does not correspond to the face images in the verification candidate list SL, the verification device 21 determines that the person captured by the short-distance camera is not permitted to pass through the gate 400.

[0066] For example, if one of the scores calculated from the face image detected from the image captured by the short-distance camera and the face images in the verification candidate list SL is equal to or greater than the threshold value, the verification device 21 determines that the face image in the verification candidate list SL corresponding to the score equal to or greater than the threshold value corresponds to the face image detected from the image captured by the short-distance camera.

[0067] In addition, if the score calculated from the face image detected from the image captured by the short-distance camera and the face images in the verification candidate list SL is less than the threshold value, the verification device 21 may determine that the face image detected from the image captured by the short-distance camera does not correspond to the face images in the verification candidate list SL. Also, if there are multiple face images indicating scores equal to or greater than the threshold value, the verification device 21 may determine that the face image detected from the image captured by the short-distance camera does not correspond to the face images in the verification candidate list SL. In this case, if it cannot be narrowed down to one person, it is determined that the face verification has failed, so a strict verification result can be obtained.

[0068] Conversely, when there are multiple face images showing scores equal to or above the threshold value, it may be determined that the face image detected from the image captured by the short-distance camera corresponds to any of the face images in the collation candidate list SL. In this case, even when face collation occurs for a person such as a twin who is difficult to narrow down to one person, it is possible to prevent stopping the flow of people. Even in this case, since at least a score equal to or above the threshold value is obtained, a certain degree of reliability can be ensured. Therefore, clearly, face collation will not succeed for a person who is clearly not a collation candidate (for example, a person with a face image not registered in the face registration DB203).

[0069] If face authentication is successful in S13 and the person captured by the short-distance camera can pass through the gate 400 (step S14, Yes), the collation device 21 generates result information R indicating that passage is possible. Then, the gate control device 20 gives a gate opening instruction based on this result information R (step S15). In step S16, the gate opening state is maintained until the person finishes passing through the gate 400. After the person passes through the gate 400, a gate closing instruction is given (step S17), and the processing after step S12 is repeated.

[0070] If the person captured by the short-distance camera cannot pass through the gate 400 (step S14, No), the collation device 21 generates result information R indicating that passage is not possible. Then, the gate control device 20 performs the processing of step S17 based on this result information R.

[0071] In FIG. 6, an example is shown where the collation device 21 can create a collation candidate list SL by medium-distance narrowing-down search. However, there may be a case where the result of the medium-distance narrowing-down search process by the collation device 21 is not appropriate and the collation candidate list SL cannot be created. The case where the result of the medium-distance narrowing-down search process is not appropriate includes, for example, the case where the score between the face image detected from the image captured by the medium-distance camera and the candidate face image in the collation candidate list ML is less than the threshold value. For example, the case where the result of the medium-distance narrowing-down search process is not appropriate occurs when a person not included in the image of the long-distance camera is included in the image of the medium-distance camera.

[0072] Figure 7 is a flowchart for explaining an operation example when the result of the medium-distance narrowing-down search process is inappropriate. In the following, the explanation of the same processes as those with each step number shown in FIG. 6 will be omitted, and different processes will be explained.

[0073] When the result of the medium-distance narrowing-down search process is inappropriate (step S100, No), the collation device 21 transmits a medium-distance narrowing-down search request to the face authentication server 200 (step S101). Here, the collation device 21 may transmit the face image detected from the image captured by the medium-distance camera to the face authentication server 200.

[0074] When the face authentication server 200 receives a medium-distance narrowing-down search request (step S102), it performs a medium-distance narrowing-down search process (step S103) and transmits the result of the medium-distance narrowing-down search process to the collation device 21 (step S104). Note that in the medium-distance narrowing-down search process in step S103, the face authentication server 200 collates the face image detected from the image captured by the medium-distance camera with the candidate face images included in the face registration DB 203.

[0075] When the collation device 21 receives the result of the medium-distance narrowing-down search process (step S105), it buffers this result in the collation candidate list SL (step S106).

[0076] When the medium-distance narrowing-down search process can be performed (step S100, Yes), the process of step S106 is performed.

[0077] According to the flowchart of FIG. 7, even when the result of the medium-distance narrowing-down search process by the collation device 21 is inappropriate and the collation candidate list SL cannot be created, the face authentication process (see FIG. 6) can be implemented by requesting the face authentication server 200 to perform a re-search.

[0078] Note that the verification device 21 may request the face authentication server 200 to perform a re-search when the face authentication process fails and it is determined in the passage determination of S14 in FIG. 6 that passage is not possible. For example, when the score between the face image detected from the image captured by the short-distance camera and the face images in the verification candidate list SL is less than the threshold value, the face authentication server 200 may be requested to perform a re-search. For example, when a person not included in the image of the medium-distance camera is included in the image of the short-distance camera, the face authentication server 200 may be requested to perform a re-search. When the verification device 21 requests the face authentication server 200 to perform a re-search, the verification in the verification device 21 is delayed by the time required for the communication to obtain the verification candidate list again from the face authentication server 200. However, since the possibility of a re-search request occurring is low, even if such processing is performed, the processing speed of the verification device 21 is improved compared to the case where no narrowing down is performed at all.

[0079] FIG. 8 is a flowchart for explaining an operation example when the face authentication process fails. Hereinafter, the explanation of the same processes as those of each step number shown in FIG. 6 will be omitted, and different processes will be explained.

[0080] When the person captured by the short-distance camera cannot pass through the gate 400 (step S14, No), the verification device 21 transmits a short-distance search request to the face authentication server 200 (step S201). The verification device 21 may transmit the face image detected from the image captured by the short-distance camera to the face authentication server 200.

[0081] The face authentication server 200 receives the short-distance search request (step S202) and performs a short-distance search process (step S203). For example, the face authentication server 200 identifies, for example, one face image detected from the image captured by the short-distance camera and the face registration DB 203, and transmits the processing result to the verification device 21 (step S204). In the short-distance face narrowing-down search process in step S203, the face authentication server 200 compares the face image detected from the image captured by the short-distance camera with the face images included in the face registration DB 203.

[0082] When the matching device 21 receives the result of the short-distance face narrowing search process (step S205), it performs face authentication processing in the same manner as the process in step S13 by matching the face image detected from the image captured by the short-distance camera with this result, and determines whether passage is permitted in the same manner as in step S14 (step S206).

[0083] As a result, when the person captured by the short-distance camera can pass through the gate 400 (step S206, Yes), the matching device 21 generates result information R indicating that passage is possible. Thereby, the processes after step S15 are performed.

[0084] If the person captured by the short-distance camera cannot pass through the gate 400 as a result of the matching (step S206, No), the matching device 21 generates result information R indicating that passage is not possible. Thereby, the gate door is not opened (step S17).

[0085] In addition, in FIG. 8, when the person captured by the short-distance camera cannot pass through the gate 400 (step S14, No), an example is shown in which the matching device 21 transmits a short-distance search request to the face authentication server 200, but the present disclosure is not limited to this. For example, instead of transmitting a short-distance search request to the face authentication server 200, the matching device 21 may match the face image detected from the image captured by the short-distance camera with the candidate face images in the matching candidate list ML. Since the matching candidate list ML is buffered in the matching device 21, matching within the range of the candidate face images included in the matching candidate list ML can be executed without communicating with the face authentication server 200. Therefore, if the matching within the range of the candidate face images included in the matching candidate list ML is successful, the matching process can be speeded up by the amount of communication that can be omitted. However, since the matching candidate list ML is a relatively large list, depending on the communication speeds of the matching device 21 and the network 300, it may be faster to complete the process by transmitting a short-distance search request to the face authentication server 200. Therefore, if the size of the matching candidate list ML is variable, it may be switched whether to transmit a short-distance search request or to perform matching with the candidate face images in the matching candidate list ML according to the size.

[0086] Also, in FIG. 8, the information returned by the face authentication server 200 in response to a short-distance search request is the result of the short-distance face narrowing-down search process, and the collation device 21 performs a collation process with that result (steps S205, S206). However, if the processing capacity of the face authentication server 200 has a margin, the face authentication server 200 may issue a conclusion on the success or failure of face collation and transmit the success or failure as it is. In this case, the collation device 21 determines whether to pass according to the received result of the success or failure of face collation and performs the gate opening / closing process.

[0087] FIG. 9 is a diagram showing a state in which a person is photographed by a long-distance camera. The long-distance camera (camera 1-1) can photograph the face of a person existing in the area A1 at a position a certain distance away from the gate 400. Therefore, by using the face images of a plurality of people who may enter the gate 400, a plurality of people who may pass through the gate 400 can be roughly narrowed down from the registered collation information with a large population. In this way, by performing pre-narrowing using the face images photographed by the long-distance camera, the face authentication process performed when a person enters the gate 400 can be speeded up.

[0088] FIG. 10 is a diagram showing a state in which a person is photographed by a medium-distance camera. The medium-distance camera (camera 1-2) can photograph the face of a person existing in the area A2 closer to the gate 400 than the area A1. Therefore, for example, immediately before a person enters the gate 400, one or a plurality of people with a higher possibility of passing through the gate 400 can be narrowed down from the collation candidate list ML. By this narrowing down, the face authentication process performed when a person enters the gate 400 can be made faster. Also, even when there are people who cannot be captured by the long-distance camera, by executing the process shown in FIG. 7, collation using the medium-distance camera is possible. People who cannot be captured by the long-distance camera are, for example, people who cut in to the gate 400, people who approach behind the previous passenger and enter the gate 400, and the like.

[0089] FIG. 11 is a diagram showing a state in which a person is photographed by a short-distance camera. In the short-distance camera (Camera 1-3), since the face image of a person passing through the gate 400 can be clearly photographed, face authentication processing can be performed with high accuracy. Also, since the collation target is narrowed down in advance, collation can be performed at high speed, and it may be possible to collate even a person with a relatively high walking speed. Further, compared to the case of performing face authentication processing for a large number of populations, the burden of face authentication processing is reduced, so that an inexpensive CPU with low processing power can be used. Also, since the burden of face authentication processing is reduced, the resolution of the short-distance camera can be increased, and the authentication accuracy can be further improved.

[0090] FIG. 12 is a diagram for explaining the relationship between the walking speed of a person passing through the gate G and face authentication processing. In the case of a gate where the pass / fail of passage is determined by face authentication, the length of the gate is determined by the relationship between the face authentication processing time and the time (performance) from the time when the door opening command of the opening / closing door mechanism Dr is transmitted to the time when the door opening is completed. The following explains this relationship.

[0091] In FIG. 12, the position SP indicates the position where the face authentication process of a person passing through the gate G starts. In order to prevent the gate from being allowed to pass through by the face of a person who has not entered the gate G, it is desirable that this position be a position after the person has entered the gate G. For example, in the case of the gate 400 described above, it corresponds to the position where the entry into the gate 400 is detected by the passage management photo - electric sensor 3. Also, the position LP indicates the limit position for issuing a command to open and close the gate G. In the case of the gate G provided with the physical opening and closing door mechanism Dr, a certain time is required until this opening and closing operation is completed. It is difficult to significantly shorten this time (door opening process time) from the viewpoints of technical performance and safety. Therefore, the length L2 reflecting the standard walking speed of a person in this time is the shortest length of the gate G when it is assumed that the time required for the opening and closing determination is zero. On the other hand, in a configuration where the opening and closing determination of the opening and closing door mechanism 4 is made according to the result of the face authentication process as in the present embodiment, the face authentication process must be completed at least by the time the person reaches the position LP. Therefore, the length from the opening and closing door mechanism Dr of the gate G to the position SP where the face authentication process of a person passing through the gate G starts is the sum of the length L1 reflecting the time required for the face authentication process (face authentication process time) and the length L2.

[0092] As a specific example, when the walking speed is 3.6 km / h, the face authentication process time is 200 msec, and the door opening process time is 600 msec, the distance from the gate end (gate entrance) to the opening and closing door mechanism Dr may be set to 800 mm or more. As described above, it is difficult to shorten the length L2. Therefore, in order to shorten the gate length, it is required to shorten the length L1, that is, to shorten the time required for the face authentication process.

[0093] In addition, when the number of registered users reaches 1 million to 10 million, there is a possibility of delay in face verification. For example, when the False Acceptance Rate (FAR) is 0.001%, there is a possibility that one person different from the user may be selected as a candidate out of 100,000 people. Even when the FAR is 0.0001%, there is a possibility that one person different from the user may be selected as a candidate out of 1 million people. In order to lower the FAR, more complex face verification processing such as evaluation from various viewpoints is required, so the processing time for face verification tends to be long.

[0094] Also, in the case of the gate 400 having a narrow passage where two or more people cannot pass through simultaneously, such as a ticket gate, it becomes difficult for a person who has been refused passage to retreat if there is a queue behind. Therefore, it is ideal for authentication to be completed when a person enters the gate 400.

[0095] Thus, for any purpose such as reducing the size of the gate, lowering the FAR, or ensuring convenience in a narrow passage, speeding up face verification is required.

[0096] According to the verification device 21 according to the present disclosure, by using a long-distance camera to capture a face image of a person approaching the gate 400, before the person enters the gate 400, a narrowing-down search can be performed using the face image from among collation information with a large population. As a result, the face authentication process using a short-distance camera can be speeded up.

[0097] FIG. 13 is a diagram for explaining a modified example of a gate. The arch-shaped gate 400 shown in FIG. 13 includes a reader 500 that reads a code such as a QR code containing information for identifying a person passing through the gate 400, and a detection camera 501 that detects companions and the like. Further, the arch-shaped gate 400 shown in FIG. 13 includes two cameras for each of long distance, medium distance, and short distance. Camera 1-1A and camera 1-1B are two long-distance cameras, and are arranged separately on the left and right on the pillar portion of the housing of the gate 400. Camera 1-2A and camera 1-2B are two medium-distance cameras, and are arranged separately on the left and right on the pillar portion of the housing of the gate 400. Camera 1-3A and camera 1-3B are two short-distance cameras, and are arranged separately on the left and right on the pillar portion of the housing of the gate 400. By arranging one camera for each distance on the left and right, even if the face of a person trying to pass through the gate faces either the left or the right, a frontal face suitable for collation can be captured by one of the cameras.

[0098] In FIG. 13, one long-distance camera, one medium-distance camera, and one short-distance camera are all arranged on the left and right, but it is not necessarily necessary to arrange one camera for each distance on the left and right. For example, in an environment where a certain degree of bias is expected in the direction a person faces, the cameras in the direction opposite to that direction are limited to auxiliary uses, so it is not necessary to have all of long distance, medium distance, and short distance. As another example of omitting one of the left or right cameras for any distance, there may be cases where it is difficult to mount a camera on either the left or the right due to the shape of the gate, the internal structure of the gate, design considerations, etc. Note that the farther the distance, the less the difference in face orientation affects the shooting result. Therefore, when omitting one of the left or right cameras, it is less affected to omit the camera for a farther distance.

[0099] By arranging a plurality of cameras for long-distance, medium-distance, and short-distance use in this way, even when the face of a person passing through the gate 400 is facing in a direction different from the traveling direction, it is easy to obtain a face image capturing the front of the face. Note that the shape of the gate 400, the positions of the cameras, and the number of cameras are not limited to those in the above-described embodiment.

[0100] In the above-described embodiment, an example in which two narrowing-down search processes are performed has been described. However, it may be configured to perform one narrowing-down search process (either a long-distance narrowing-down search process or a medium-distance narrowing-down search process).

[0101] According to this configuration example, since one narrowing-down search process is performed, the processing time for narrowing down can be shortened, and the number of cameras for capturing images used in the narrowing-down search process can be reduced. Therefore, the configuration of the system is simplified, and while reducing the cost associated with constructing the system, it is possible to speed up the face authentication process.

[0102] Also, according to this configuration example, since the narrowing-down search process is performed by the face authentication server 200, it is possible to handle an enormous amount of collation information that cannot be stored in the collation device 21. Further, since the face authentication server 200 can be shared by the gates 400 set at a plurality of bases, it is possible to speed up the face authentication process and effectively utilize resources.

[0103] In this configuration example, the narrowing-down search is performed once in the face authentication server 200. However, after the narrowing-down search in the face authentication server 200, the second narrowing-down search process may be performed by the collation device 21. In this case, the processing unit 102 of the collation device 21 uses the third face image captured by the medium-distance camera that captures the faces of people existing in the area A2 between the area A1 and the area A3 to further narrow down the face images of people who can pass through the gate 400 from the face images acquired from the face authentication server 200. The processing unit 102 of the collation device 21 collates the second face image with the narrowed-down face image. For example, when the size of the collation candidate list ML is large, it may take time to transmit from the face authentication server 200 to the collation device 21, or it may take time for the collation device 21 to perform the narrowing-down search process using the collation candidate list ML due to its performance. In such a case, with the above-described configuration, it is possible to further speed up the face authentication process. When the size of the collation candidate list ML is sufficiently small, as in the above configuration example, it may be possible to speed up the process by having the face authentication server 200 transmit the collation candidate list ML to the collation device 21 at an early stage and having the collation device 21 perform the second narrowing-down. Based on these, depending on the size of the collation candidate list ML, it may be possible to switch whether to perform the second narrowing-down search process in the face authentication server 200 or in the collation device 21.

[0104] Hereinafter, modified examples of the narrowing-down search process, face authentication process, etc. in the face authentication server 200 and the collation device 21 will be described.

[0105] In the above-described long-distance narrowing-down search (primary narrowing-down search) in FIG. 5, an example was described in which the face authentication server 200 buffers a predetermined number (for example, N 1 ) of face images with high scores calculated for each of the face images included in the face registration DB203 in the collation candidate list ML. However, the present disclosure is not limited to this. For example, the face authentication server 200 may buffer, in the collation candidate list ML, the face images corresponding to the scores exceeding the first threshold among the scores calculated for each of the face images included in the face registration DB203.

[0106] In the above-described long-distance narrowing search (primary narrowing search) of FIG. 5, for each face image included in the collation candidate list ML, the collation device 21 buffers a predetermined number (for example, N 2 ) of face images with higher scores from the higher-scoring ones in the collation candidate list SL.

[0107] According to this configuration example, since the upper limit of the face images buffered in the collation candidate list SL can be defined, it is possible to prevent buffer overflow even when a large number of face images with high scores are found.

[0108] However, the present disclosure is not limited to this. For example, the collation device 21 may buffer, in the collation candidate list SL, face images corresponding to scores exceeding a second threshold higher than a first threshold among the scores calculated for each face image included in the collation candidate list ML.

[0109] According to this configuration example, since narrowing is performed based on a threshold, face images corresponding to scores exceeding the threshold are not excluded from the collation candidates, and face images corresponding to scores less than the threshold are excluded from the collation candidates. Therefore, it is possible to speed up the face authentication process and improve the accuracy of the face authentication process.

[0110] In this configuration example, a secondary narrowing search is performed, but it is also possible to omit the secondary narrowing search. An example in this case will be described.

[0111] As a result of the primary narrowing search, when the secondary narrowing search does not need to be performed, the processing unit 102 may execute the face authentication process without performing the secondary narrowing. The case where the secondary narrowing search does not need to be performed may correspond to, for example, a case where the number of face images corresponding to scores exceeding the first threshold is a number (for example, 1) for which the secondary narrowing search does not need to be performed as a result of the primary narrowing search. Alternatively, the case where the secondary narrowing search does not need to be performed may correspond to a case where the difference between the highest score and the Nth (N is an integer of 2 or more) highest score is equal to or greater than a predetermined difference as a result of the primary narrowing search.

[0112] According to this example, the secondary narrowing-down search can be omitted, so that the processing time for narrowing down can be shortened, and the face authentication process can be speeded up.

[0113] In addition, when a sufficiently high matching result is obtained by the above-described primary narrowing-down search and secondary narrowing-down search, the face authentication process may be omitted. An example in this case will be described.

[0114] The case where a sufficiently high matching result is obtained by the primary narrowing-down search and the secondary narrowing-down search may correspond to, for example, a case where the difference between the highest score and the second highest score as a result of the secondary narrowing-down search is equal to or greater than a predetermined difference. In this case, the processing unit 102 of the collation device 21 may generate the result of the secondary narrowing-down search as result information R indicating that authentication has been performed without executing the face authentication process.

[0115] According to this example, since the face authentication process may be omitted, it is possible to reduce the processing time due to the face authentication process.

[0116] In addition, when the scores for all face candidates included in the secondary narrowing-down result are too low, the face image captured by the medium-distance camera may be sent to the face authentication server 200, and the candidates to be buffered may be acquired again. A configuration example in this case will be described.

[0117] When the face image corresponding to the score exceeding the second threshold value as a result of the secondary narrowing-down can be collated with the third face image, the processing unit 102 of the collation device 21 performs the face authentication process without transmitting a plurality of third face images to the face authentication server 200.

[0118] As a result of the second filtering, if the face image corresponding to the score exceeding the second threshold value cannot be matched with the third face image, the processing unit 102 of the matching device 21 transmits a plurality of third face images to the face authentication server 200 and causes the face authentication server 200 to perform filtering of the face images from among the plurality of third face images. The processing unit 102 of the matching device 21 acquires the filtered face images from the face authentication server 200 and further filters the face images of the persons who can pass through the gate 400 from among the acquired face images.

[0119] According to this configuration example, the face image captured by the medium-distance camera can be sent to the face authentication server 200 to obtain candidates for buffering again. Therefore, even if the score of any face candidate included in the second filtering result is too low, the filtering accuracy can be ensured at a certain level or higher, and the face authentication process can be accelerated while suppressing a decrease in the accuracy of the face authentication process.

[0120] Note that the processing unit 102 may be configured to acquire a feature amount indicating a feature of a face included in a face image captured by, for example, a long-distance camera, a medium-distance camera, or a short-distance camera, and use the feature amount to filter face images of persons who can pass through the gate 400 from among a plurality of matching information. Further, the processing unit 102 may be configured to acquire a feature amount indicating a feature of a face included in the face image filtered by the face authentication server 200 and use the feature amount to filter face images of persons who can pass through the gate 400 from among a plurality of matching information. Here, as an example of the feature amount, the color, shape, brightness distribution, etc. of the face can be considered. It may also be a feature amount generated by more complex processing used in the field of machine learning. By using the feature amount, the size of the information exchanged between the face authentication server 200 and the matching device 21 can be suppressed. Further, depending on the feature amount used, the influence of parameters that are likely to change in the actual environment is suppressed, enabling robust face authentication.

[0121] In the above-described embodiment, an example where the authentication information is recorded in the face authentication server 200 has been described, but the present disclosure is not limited thereto. For example, the authentication information may be recorded in the collating device 21 or the gate control device 20. For example, in the collating device 21, if it has a large recording capacity capable of recording information (authentication information) of a large number of face images and a processing ability capable of executing the first narrowing-down search process, the processing unit 102 of the collating device 21 may narrow down face images from among a plurality of collation information using the first face image.

[0122] Also, in the face authentication system 100, the number of the plurality of cameras 1 may be 4 or more. Also, the number of times of narrowing down may be, for example, 3 or more. Specifically, it is configured to photograph each of 4 or more regions with 4 or more cameras. The more the number of times of narrowing down, the more it is possible to handle a large number of face images (for example, when the number of stored images in the face registration DB 203 is large). However, since the time required for re-search in case of failure in determination may become longer as the number of times of narrowing down increases, the threshold value of the score for determination may be lowered as the number of times increases.

[0123] Also, the face authentication server 200 may perform the second narrowing-down search process. In particular, when the size of the collation candidate list ML is large, depending on the performance of the network 300 and the collating device 21, it may take time to send the collation candidate list ML to the collating device 21 and have the second narrowing-down search process performed locally. Also, whether or not the face authentication server 200 performs the second narrowing-down search process may be made switchable based on the size of the collation candidate list ML, the communication speed of the network 300, the buffer size of the collating device 21, or the processing ability of the collating device 21. When adding candidates with a score equal to or higher than the threshold value to the collation candidate list ML, since the size of the collation candidate list ML is variable, it is beneficial to perform such switching.

[0124] Also, regarding the case of performing the narrowing-down search process three or more times, the determination of up to which narrowing-down search process should be performed by the face authentication server 200 may be made from the same viewpoint.

[0125] Further, even if a short-distance camera, a medium-distance camera, and a long-distance camera are not provided for each of the plurality of gates 400, for example, the medium-distance camera and the long-distance camera may be shared by the plurality of gates 400.

[0126] Also, in the above-described embodiment, the information used for each narrowing-down and the types of information used when obtaining the authentication result may be made different. For example, it is conceivable to use the feature amount of the face contour for narrowing-down and the feature amount of the face parts when obtaining the authentication result. Since the size of the face image captured by the long-distance camera and the short-distance camera is also different, the accuracy of determination can be further improved by using information suitable for each distance. Also, in the sense of simply comprehensively evaluating from a plurality of viewpoints, it is possible to expect an improvement in accuracy by performing the narrowing-down using different information and the process of obtaining the authentication result.

[0127] On the other hand, the same information may be used as the information used for each narrowing-down and the types of information used when obtaining the authentication result. By doing so, since the evaluation is performed from the same viewpoint in the previous narrowing-down and the current narrowing-down or face authentication process, the occurrence of a deviation in the judgment result can be suppressed. As a result, the frequency of requests to the face authentication server 200 due to face authentication failure can be reduced, and thus high-speed face authentication processing can be expected.

[0128] Also, in the above-described embodiment, the face images included in the collation candidate list obtained as a result of narrowing down may be not the images themselves but their feature amounts (this information is also referred to as "candidate face images"). In particular, for the collation candidate list transmitted from the face authentication server 200 to the collation device 21, a list consisting of feature amounts can suppress the communication volume. However, since it is generally difficult to perform collation with other feature amounts using the information obtained by extracting the feature amounts, when performing collation with different feature amounts at each stage, even if the size of the collation candidate list increases, it is preferable to use the face images themselves as the collation candidate list. Also, in a situation where the size of the collation candidate list hardly affects the communication volume, such as when performing narrowing down a plurality of times within the face authentication server 200, the face images themselves may be used as the collation candidate list, and the feature amounts may be extracted when creating the collation candidate list to be transmitted to the collation device 21.

[0129] Also, in the above-described embodiment, the gate 400 included the opening / closing door mechanism 4, but the means (regulation unit) for restricting the movement of a person when face collation fails is not limited to this. For example, a mechanism for psychologically restricting such as a siren and / or an alarm may be adopted. Also, instead of notifying the person himself / herself who is trying to pass through the gate, a mechanism for indirectly restricting the movement may be adopted by notifying a guard and / or a robot arranged nearby. Note that depending on what kind of regulation unit is adopted, the time from when face collation fails until regulation is performed is different, but in any case of using any means, it is similarly useful to speed up face collation in order to obtain the result of face collation before the person reaches the regulation unit.

[0130] In other words, the means (restriction unit) for restricting the movement of a person when face verification fails is not limited to an example of physically restricting (blocking) the movement of a person, such as the opening and closing door mechanism 4 provided in the middle of the movement path of the person at the gate 400. For example, a specific point (or a specific range) may be set in the gate 400, and the gate 400 may restrict the movement of a person from upstream of the specific point to downstream of the specific point in the movement direction of the person. And in this case, the means of restriction may be a siren and / or an alarm or the like as described above, or may be a notification to a security guard and / or a robot or the like. In this case, the shooting range of each camera may be located upstream of the specific point. For example, in order from the one closer to the specific position, the shooting range of the short-distance camera (for example, region A3 in FIG. 4), the shooting range of the medium-distance camera (for example, region A2 in FIG. 4), and the shooting range of the long-distance camera (for example, region A1 in FIG. 4) may be provided.

[0131] Also, in the above embodiment, the verification device 21 has been described as a device used for a gate that restricts the movement of a person, but it is not limited to this. As long as it is a system for performing face authentication on a person approaching from a distance, it can be applied to any system. In that case, in the above embodiment, the definition of the region A3 (see FIG. 4) where the face authentication process is performed differs depending on the requirements of the system. For example, when applying it to a monitoring system using a monitoring camera that records a person who has passed a specific monitoring point, it is conceivable to set the periphery of the monitoring point (for example, the range upstream of the monitoring point) as the region A3.

[0132] As described above, the verification device 21 verifies a verification candidate narrowed down using a first face image captured by a first camera that captures the region A1, and a second face image captured by a second camera that captures the region A3 where a person can move from the region A1.

[0133] With this configuration, after capturing the face image of a person approaching gate 400, before the person enters gate 400, it is possible to perform a narrowing-down search from the collation information with a large population using the face image. As a result, the face authentication process using the medium-distance camera or the short-distance camera can be speeded up.

[0134] The present disclosure can be realized by software, hardware, or software in cooperation with hardware.

[0135] Each functional block used in the description of the above embodiment is realized, partially or wholly, as an LSI which is an integrated circuit, and each process described in the above embodiment may be controlled, partially or wholly, by one LSI or a combination of LSIs. The LSI may be composed of individual chips, or may be composed of one chip so as to include part or all of the functional blocks. The LSI may be provided with input and output of data. Depending on the degree of integration, the LSI may also be referred to as an IC, a system LSI, a super LSI, or an ultra LSI.

[0136] The method of integrating into an integrated circuit is not limited to LSI, and it may be realized by a dedicated circuit, a general-purpose processor, or a dedicated processor. Further, after manufacturing the LSI, an FPGA (Field Programmable Gate Array) that can be programmed, or a reconfigurable processor that can reconfigure the connection and setting of circuit cells inside the LSI may be used. The present disclosure may be realized as digital processing or analog processing.

[0137] Furthermore, if an integrated circuit technology that replaces the LSI appears due to the progress of semiconductor technology or other derived technologies, naturally, the technology may be used to integrate the functional blocks. The application of biotechnology or the like is possible as an example.

[0138] The present disclosure is applicable to any type of apparatus, device, system (collectively referred to as a communication device) having a communication function. The communication device may include a wireless transceiver (transceiver) and a processing / control circuit. The wireless transceiver may include a receiving unit and a transmitting unit, or may include them as functions. The wireless transceiver (transmitting unit, receiving unit) may include an RF (Radio Frequency) module and one or more antennas. The RF module may include an amplifier, an RF modulator / demodulator, or the like. Non-limiting examples of the communication device include a telephone (mobile phone, smartphone, etc.), a tablet, a personal computer (PC) (laptop, desktop, notebook, etc.), a camera (digital still / video camera, etc.), a digital player (digital audio / video player, etc.), a wearable device (wearable camera, smartwatch, tracking device, etc.), a game console, a digital book reader, a telehealth / telemedicine (remote healthcare / medical prescription) device, a vehicle or mobile transportation means with a communication function (automobile, airplane, ship, etc.), and combinations of the various devices described above.

[0139] The communication device is not limited to being portable or mobile, and includes any type of apparatus, device, system that is not portable or is fixed, such as a smart home device (home appliance, lighting device, smart meter or measuring device, control panel, etc.), a vending machine, and any "Thing" that can exist on the IoT (Internet of Things) network.

[0140] In recent years, in IoT (Internet of Things) technology, CPS (Cyber Physical Systems), which is a new concept of creating new added value through information cooperation between the physical space and the cyber space, has attracted attention. This CPS concept can also be adopted in the above embodiments.

[0141] That is, as a basic configuration of the CPS, for example, an edge server arranged in the physical space and a cloud server arranged in the cyber space are connected via a network, and processing can be distributed and processed by processors mounted on both servers. Here, each processing data generated in the edge server or the cloud server is preferably generated on a standardized platform. By using such a standardized platform, it is possible to improve the efficiency when constructing a system including various types of sensor groups and IoT application software.

[0142] For communication, in addition to data communication by a cellular system, a wireless LAN system, a communication satellite system, etc., data communication by combinations of these is also included.

[0143] In addition, the communication device also includes devices such as a controller and a sensor that are connected or linked to a communication device that executes the communication function described in the present disclosure. For example, a controller and a sensor that generate a control signal or a data signal used by a communication device that executes the communication function of the communication device are included.

[0144] In addition, the communication device also includes infrastructure facilities such as a base station, an access point, and any other devices, devices, and systems that communicate with or control the above-mentioned various types of non-limiting devices.

[0145] As described above, various embodiments have been described with reference to the drawings. Needless to say, the present disclosure is not limited to such examples. It is obvious that those skilled in the art can come up with various modification examples or correction examples within the scope described in the claims, and it is naturally understood that they also belong to the technical scope of the present disclosure. Also, within the scope not departing from the gist of the disclosure, the components in the above embodiments may be arbitrarily combined.

[0146] As described above, specific examples of the present disclosure have been described in detail, but these are merely examples and do not limit the scope of the claims. The technology described in the claims includes various modifications and changes of the specific examples exemplified above.

Industrial Applicability

[0147] One embodiment of the present disclosure is suitable for an apparatus or system that performs collation (or authentication) using a face image.

Explanation of Signs

[0148] 1, 1-1, 1-2, 1-3 Camera 2 QR Code Reader 3 Passage Management Photoelectric Sensor 4 Opening and Closing Door Mechanism 5 Entrance Guide Indicator 6 Passage Guide LED 7 Guide Display 8 Speaker 9 Interface Board 10 Interface Driver 20 Gate Control Device 21 Collation Device 30 Network Hub 100 Face Authentication System 101, 202 Communication Unit 102, 201 Processing Unit 103 Buffer 200 Face Authentication Server 203 Face Registration DB 300 Network 400 Gate 500 Reader 501 Detection Camera 601, 701 Processor 602, 702 Memory 603, 703 Input / Output Interface 604, 704 Bus

Claims

1. A matching device for regulating the flow of people, a processing unit that performs a second face image matching using a first candidate face image narrowed down from the multiple face images included in the face image database based on a result of a first face image matching using a face image included in a first image photographed of the first area and the multiple face images included in the face image database, and a face image included in a second image photographed of the second area, on a route along which there is a flow of people heading from a first area toward a second area located upstream of a position where the flow of people should be restricted; a communication unit that outputs a result of the second face image matching; A matching device comprising:

2. The regulation is at least one of a siren, an alarm, a guard, or a robot. The collation device according to claim 1 .

3. the processing unit narrows down the first candidate face images using a face image included in a third image obtained by capturing a third area between the first area and the second area, and compares the narrowed down second candidate face images with a face image included in the second image. The collation device according to claim 1 .

4. The communication unit acquires the first candidate face image from a server provided outside the matching device, the matching device acquires the second candidate face image by narrowing down the first candidate face image within the matching device; The collation device according to claim 3.

5. the processing unit determines a score indicating a similarity between two face images between each of N1 (N1 is an integer equal to or greater than 2) first candidate face images and the third image; the processing unit determines the second candidate face image by narrowing down the first candidate face image to a top N2 candidate images (N2 is an integer equal to or greater than 1 and smaller than N1) based on the N1 scores. The collation device according to claim 3.

6. the processor determines a score indicating a similarity between the two face images between each of the first candidate face images and the third image; the processing unit determines the second candidate face image by narrowing down the first candidate face image to a candidate having the score equal to or greater than a second threshold; the score of the first candidate face image with respect to a face image included in the first image is equal to or greater than a first threshold value; The first threshold is lower than the second threshold. The collation device according to claim 3.

7. When the number of the first candidate face images is less than a third threshold, the processing unit does not determine the second candidate face image. The collation device according to claim 6.

8. When the number of the second candidate face images is less than a fourth threshold, the processing unit does not perform matching with a face image included in the second image. The collation device according to claim 6.

9. When the scores between the first candidate face image and the third image are each less than a fifth threshold, the processing unit requests a server provided outside the matching device to determine the second candidate face image. The collation device according to claim 6.

10. When the processing unit fails to match the facial image included in the second image, the processing unit requests a server provided outside the matching device to match the facial image included in the second image with a plurality of facial images included in the facial image database. The collation device according to claim 1 .

11. A matching system for regulating the flow of people, comprising: a first camera that captures an image of a flow of people from a first area toward a second area located upstream of a position where the flow of people should be regulated; a second camera for photographing the second area; a first matching device that performs a first face image matching using a face image included in a first image captured by the first camera and a plurality of face images included in a face image database; a second matching device that performs a second facial image matching using a first candidate facial image narrowed down from the plurality of facial images included in the facial image database based on a result of the first facial image matching and a facial image included in a second image captured by the second camera; A matching system comprising:

12. A matching method for regulating the flow of people, comprising: a second face image matching is performed using a first candidate face image narrowed down from the plurality of face images included in the face image database based on a result of a first face image matching performed on a route along which a flow of people flows from a first area toward a second area located upstream of a position where the flow of people should be regulated, and a face image included in a second image taken of the second area, using a face image included in a second image taken of the second area; outputting a result of the second face image matching; Matching method.

Images