Authentication device, authentication system, authentication method, and program

Abstract

Provided is an authentication device comprising: a detection unit that detects a person from each of a first image and a second image, the first image and the second image being images of an area within a predetermined range based on a gate (60) and captured by, respectively, a first camera (410) and a second camera (412) which are positioned apart by a predetermined distance from each other in the width direction of the gate; a subject identification unit that, on the basis of the parallax between the first image and the second image, identifies a person closest to the gate as a subject to be authenticated from at least one of persons detected from the first image and persons detected from the second image; and an authentication unit that controls biometric authentication of the subject to be authenticated.

Description

[Technical Field]

[0001] The present disclosure relates to an authentication device, an authentication system, an authentication method, and a non-transitory computer-readable medium. [Background technology]

[0002] A method has been proposed for determining whether or not a person approaching a gate is permitted to enter by so-called walk-through authentication, which performs biometric authentication on the person. For example, Patent Document 1 discloses an authentication system that selects candidate feature information for a moving subject to be authenticated using feature information extracted from biometric information acquired at a first location, and performs authentication processing using the selected candidate and biometric information acquired just before the ticket gate at the station. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2020 / 084748 Summary of the Invention [Problem to be solved by the invention]

[0004] In walk-through authentication, multiple people may appear in the captured image, and it is necessary to accurately capture the person to be authenticated from the captured image.

[0005] In view of the above-mentioned problems, an object of the present disclosure is to provide an authentication device, an authentication system, an authentication method, and a non-transitory computer-readable medium that are capable of accurately identifying a person to be authenticated and performing authentication processing. [Means for solving the problem]

[0006] An authentication device according to one aspect of the present disclosure includes a detection means for detecting a person from each of a first image and a second image, wherein each of the first image and the second image is an image of a shooting area within a predetermined range based on the gate, taken by a first camera and a second camera arranged a predetermined distance apart in the width direction of the gate; a target identification means for identifying a person closest to the gate from at least one of the people detected from the first image and the people detected from the second image based on the parallax between the first image and the second image, as a person to be authenticated; and an authentication means for controlling biometric authentication of the person to be authenticated.

[0007] An authentication system according to one aspect of the present disclosure includes a first camera and a second camera arranged a predetermined distance apart in a width direction of a gate, and an authentication device. The authentication device includes a detection means for detecting a person from each of the first image and the second image, where the first image and the second image are images captured by the first camera and the second camera, respectively, of a photographing area within a predetermined range based on the gate, an object identification means for identifying, as an authentication subject, from at least one of the people detected from the first image and the people detected from the second image, the person closest to the gate based on the parallax between the first image and the second image, and an authentication means for controlling biometric authentication of the person to be authenticated.

[0008] An authentication method according to one aspect of the present disclosure includes detecting a person from each of a first image and a second image, wherein each of the first image and the second image is an image of a shooting area within a predetermined range based on the gate, taken by a first camera and a second camera arranged a predetermined distance apart in the width direction of the gate; identifying a person closest to the gate from at least one of the people detected from the first image and the people detected from the second image based on the parallax between the first image and the second image as a person to be authenticated; and controlling biometric authentication for the person to be authenticated.

[0009] A non-transitory computer-readable medium according to one aspect of the present disclosure stores a program for causing a computer to execute: a detection process for detecting a person from each of a first image and a second image, wherein each of the first image and the second image is an image of a shooting area within a predetermined range based on the gate, taken by a first camera and a second camera positioned a predetermined distance apart in the width direction of the gate; a target identification process for identifying a person closest to the gate from at least one of the people detected from the first image and the people detected from the second image based on the parallax between the first image and the second image as a person to be authenticated; and an authentication process for controlling biometric authentication for the person to be authenticated. [Effects of the Invention]

[0010] The present disclosure makes it possible to provide an authentication device, an authentication system, an authentication method, and a non-transitory computer-readable medium that are capable of accurately identifying an authentication target and performing authentication processing. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a block diagram showing a configuration of an authentication device according to a first embodiment. [Figure 2] 1 is a flowchart showing an example of the flow of an authentication method according to the first embodiment. [Figure 3] FIG. 10 is a block diagram showing the overall configuration of an authentication system according to a second embodiment. [Figure 4] FIG. 10 is a block diagram showing the configuration of an authentication terminal according to a second embodiment, together with a first camera and a second camera. [Figure 5] FIG. 10 is a block diagram showing the configuration of an authentication device according to a second embodiment. [Figure 6] 10 is a flowchart showing the flow of face information registration processing according to the second embodiment. [Figure 7] 10 is a flowchart showing the flow of authentication processing according to the second embodiment. [Figure 8] FIG. 10 is a diagram showing a first image captured by a first camera according to the second embodiment. [Figure 9] FIG. 10 is a diagram showing a second image captured by a second camera according to the second embodiment. [Figure 10] FIG. 10 is a diagram for explaining an example of a position calculation method according to the second embodiment. [Figure 11] 10 is a flowchart showing the flow of face matching processing according to the second embodiment. [Figure 12] FIG. 10 is a diagram for explaining an example of a position calculation method according to a modified example of the second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. In each drawing, the same or corresponding elements are designated by the same reference numerals, and for clarity of explanation, duplicate explanations will be omitted as necessary.

[0013] <Embodiment 1> Next, a first embodiment of the present disclosure will be described. Fig. 1 is a block diagram showing a configuration of an authentication device 10 according to the first embodiment. The authentication device 10 is one or more devices that recognize a person who is closest to the gate among people who are scheduled to enter a gate as an authentication target for biometric authentication.

[0014] The distance from the gate to a person can be estimated based on the size of the person's facial area or the size of their entire body area included in the captured image, but since the size of such areas and height vary from person to person, it is difficult to accurately determine the distance. In this embodiment 1, the authentication device 10 determines the distance from the gate to a person based on the parallax of the two cameras.

[0015] A first camera and a second camera (not shown) are installed near the gate, and are spaced a predetermined distance apart in the width direction of the passageway within the gate, i.e., in the direction perpendicular to the direction of travel of the approaching person.

[0016] The authentication device 10 includes a detection unit 13, an object identification unit 14, and an authentication unit 15.

[0017] The detection unit 13 is also referred to as a detection means. The detection unit 13 detects people from each of the first image and the second image. The first image is an image generated by the first camera capturing an image of a predetermined range of a shooting area based on the gate. The second image is an image generated by the second camera capturing an image of a predetermined range of a shooting area based on the gate.

[0018] The target identification unit 14 is also referred to as a target identification means. Based on the parallax between the first image and the second image, the target identification unit 14 identifies the person closest to the gate from at least one of the people detected in the first image and the people detected in the second image as the person to be authenticated.

[0019] The authentication unit 15 is also referred to as an authentication means. The authentication unit 15 controls biometric authentication for the identified person to be authenticated.

[0020] Here, biometric authentication refers to face authentication, iris authentication, gait authentication, or other biometric authentication. Therefore, biometric information refers to face information, iris information, gait information, or other biometric information. In the following description, biometric information refers to face information, but is not limited to this. In the following, face information is used as a general term for face images and facial feature information extracted from face images. Controlling biometric authentication may mean that the authentication device 10 performs biometric authentication. Alternatively, it may mean that the authentication device 10 requests biometric authentication from another device (not shown) and receives the biometric authentication result from that device.

[0021] 2 is a flowchart showing an example of the flow of the authentication method according to the first embodiment. First, the authentication device 10 acquires a first image captured by a first camera and a second image captured by a second camera (S10). For example, if the authentication device 10 is connected to the first and second cameras, the authentication device 10 may acquire the first and second images directly from the first and second cameras. If another authentication terminal (not shown) is connected to the first and second cameras, the authentication device 10 may receive the first and second images from the authentication terminal, or may receive cropped images of the person area or face area included in each of the first and second images from the authentication terminal.

[0022] Next, the detection unit 13 detects a person from each of the first image and the second image by image recognition (S11). Next, the object identification unit 14 identifies the person closest to the gate as the person to be authenticated based on the parallax between the first image and the second image (S12).

[0023] Next, the authentication unit 15 controls biometric authentication for the identified person to be authenticated (S13). For example, the authentication unit 15 may perform face authentication by extracting features from a facial region of the identified person to be authenticated and comparing the extracted features with pre-registered features.

[0024] Then, the authentication unit 15 outputs the authentication result (S14).

[0025] As described above, the authentication device 10 according to the first embodiment makes it possible to accurately capture the person to be authenticated using the parallax between the two cameras and perform the authentication process when performing biometric authentication at the time of passing through a gate.

[0026] Furthermore, using two cameras can reduce blind spots. For example, even if a person enters from the side in front of the gate, the authentication device 10 can capture the person closest to the gate with a relatively large margin before they reach the gate. This can avoid delays in authentication timing.

[0027] <Embodiment 2> Next, a second embodiment of the present disclosure will be described. The second embodiment is a specific example of the first embodiment. Fig. 3 is a block diagram showing the overall configuration of an authentication system 1 according to the second embodiment. The authentication system 1 is used to pass through a gate 60 using face authentication.

[0028] The authentication system 1 includes an authentication device 10a, an authentication terminal 40, a gate 60, a first camera 410, and a second camera 412. The authentication device 10a and the authentication terminal 40 are communicatively connected via a network N. The network N is a wired or wireless communication network.

[0029] The gate 60 may be a ticket gate at a station, a gate at an airport, or a gate at a concert hall or amusement park, etc. The gate 60 includes a first gate body 61 and a second gate body 62.

[0030] The first gate body 61 and the second gate body 62 are spaced apart at a predetermined interval in a direction perpendicular to the traveling direction of pedestrians. Two cameras (a first camera 410 and a second camera 412) are disposed on each of the first gate body 61 and the second gate body 62. In other words, the first camera 410 and the second camera 412 are spaced apart at a predetermined interval in a direction perpendicular to the traveling direction of pedestrians.

[0031] The first camera 410 and the second camera 412 constantly capture images of the area around the gate 60, generating a first image and a second image, respectively. The authentication terminal 40 and the authentication device 10 then work together to identify the person to be authenticated from the people included in the first and second images using parallax, and perform facial authentication on the person to be authenticated. The gate 60 then opens or closes based on the results of the facial authentication.

[0032] Although the first camera 410 and the second camera 412 are described as being disposed on the first gate body 61 and the second gate body 62, respectively, they may be disposed regardless of the positions of the first gate body 61 and the second gate body 62. The distance between the first camera 410 and the second camera 412 is preferably 0.5 m or more and 1.5 m or less, and more preferably 0.7 m or more and 1.0 m or less. When the distance is 0.5 m or more and 1.5 m or less, the common area of ​​the fields of view of the two cameras can be secured widely while maintaining parallax, thereby improving the accuracy of identifying the person to be authenticated. Furthermore, when the distance is 0.7 m or more and 1.0 m or less, the balance between the common area of ​​the fields of view of the cameras and parallax is more optimal, thereby further improving the accuracy of identifying the person to be authenticated.

[0033] Here, opening or closing the gate 60 means rotating or sliding the flap 63 to an open or closed state. However, opening or closing the gate 60 may also mean unlocking or locking the flap 63, or simply displaying or outputting a sound indicating whether passage is permitted or not. Note that the flap 63 is not essential for the gate 60.

[0034] The authentication terminal 40 is a terminal device connected to the first camera 410 and the second camera 412. The authentication terminal 40 acquires a first image and a second image from the first camera 410 and the second camera 412. The authentication terminal 40 transmits the first image and the second image to the authentication device 10a and requests facial authentication from the authentication device 10a. The authentication terminal 40 receives the result of the facial authentication from the authentication device 10a and controls the opening and closing of the gate 60 based on the result of the facial authentication.

[0035] The gate 60 and the authentication terminal 40 may be configured as separate devices or may be configured as an integrated unit. The gate 60 may not have an opening / closing mechanism. For example, the gate 60 may be a pole-type gate rather than having two separate bodies, or may simply be a passageway. In this case, the authentication terminal 40 may output a pass-through permission / denial signal instead of controlling the opening / closing of the gate 60.

[0036] The authentication device 10a is a specific example of the authentication device 10 of embodiment 1. In response to receiving a facial authentication request from the authentication terminal 40, the authentication device 10a identifies the person closest to the gate as the person to be authenticated based on the first image and the second image. Then, the authentication device 10a performs facial authentication on the facial area of ​​the person to be authenticated and transmits the result of the facial authentication to the authentication terminal 40.

[0037] 4 is a block diagram showing the configuration of authentication terminal 40 according to the second embodiment, together with first camera 410 and second camera 412. Authentication terminal 40 includes a storage unit 420, a communication unit 430, a display unit 440, and a control unit 460.

[0038] The storage unit 420 is a storage device that stores programs for realizing each function of the authentication terminal 40. The communication unit 430 is a communication interface with the network N. The display unit 440 is a display device. The control unit 460 controls the hardware that the authentication terminal 40 has.

[0039] For example, the control unit 460 receives first and second images from the first and second cameras 410 and 412 via the communication unit 430 and transmits the first and second images to the authentication device 10a. In this case, the control unit 460 may constantly receive the images from the first and second cameras 410 and 412. However, the control unit 460 may control the first and second cameras 410 and 412 so that they capture images of a predetermined area based on the gate 60 at approximately the same time, thereby acquiring the first and second images in response to the capture. "Capturing images at approximately the same time" may mean that the capture timing is the same, or the difference in capture timing may be less than a predetermined time. Furthermore, the control unit 460 controls the opening and closing of the gate 60 based on the result of face authentication in response to the communication unit 430 receiving the result of face authentication from the authentication device 10a.

[0040] 5 is a block diagram showing the configuration of the authentication device 10 according to the embodiment 2. The authentication device 10 includes a storage unit 100, a memory 110, a communication unit 120, an input unit 130, and a control unit 140.

[0041] The storage unit 100 is a storage device such as a hard disk, a flash memory, etc. The storage unit 100 stores a program 101 and a face information DB 102. The program 101 is a computer program in which the processing by the authentication device 10a is implemented.

[0042] The face information DB 102 stores a user ID for identifying a user and facial feature information of the user in association with each other. The facial feature information is a set of feature points extracted from a face image.

[0043] The memory 110 is a volatile storage device such as a RAM (Random Access Memory), and is a storage area for temporarily storing information when the control unit 140 is operating. The communication unit 120 is a communication interface with the network N. The input unit 130 is an input device. Note that the input unit 130 is not essential for the authentication device 10a.

[0044] The control unit 140 is a processor, i.e., a control device, that controls each component of the authentication device 10a. The control unit 140 loads the program 101 from the storage unit 100 into the memory 110 and executes the program 101. In this way, the control unit 140 realizes the functions of a registration unit 141, an acquisition unit 142, a detection unit 143, an object identification unit 144, and an authentication unit 145.

[0045] The registration unit 141 is also referred to as a registration means. When registering face information, the registration unit 141 issues a new user ID. The registration unit 141 also detects a face area from a registration image for face information registration, extracts features from the face area, and generates face feature information. The registration unit 141 associates the issued user ID with the face feature information for registration extracted from the registration image, and registers them in the face information DB 102.

[0046] The acquisition unit 142 is also referred to as an acquisition means. The acquisition unit 142 receives and acquires the first image and the second image from the authentication terminal 40 via the communication unit 120. The acquisition unit 142 supplies the received first image and second image to the detection unit 143.

[0047] The detection unit 143 is an example of the detection unit 13 of embodiment 1. The detection unit 143 detects a person's face area from each of the first image and the second image received from the authentication terminal 40, and supplies each face area to the object identification unit 144. If the first image or the second image includes multiple people, the detection unit 143 detects a face area for each person and supplies the face area to the object identification unit 144.

[0048] The target identification unit 144 is an example of the target identification unit 14 of the first embodiment. First, the target identification unit 144 identifies the facial region of the same person included in the first image and the second image based on the facial region of each person acquired from the detection unit 143. The first image and the second image are images captured at approximately the same time. Next, for a person presumed to be the same person in the first image and the second image, the target identification unit 144 measures the distance between the person and the gate based on the parallax between the first camera 410 and the second camera 412. The parallax may be calculated based on the position of the person's facial region in the first image and the position of the person's facial region in the second image. Then, the target identification unit 144 identifies the person closest to the gate as the person to be authenticated based on the calculated parallax. The target identification unit 144 supplies the facial region of the person to be authenticated included in the first image or the second image to the authentication unit 145.

[0049] The authentication unit 145 is an example of the authentication unit 15 of the first embodiment. The authentication unit 145 extracts feature points from the face region acquired from the object identification unit 144 and sets the extracted feature points as facial feature information for authentication. The authentication unit 145 then performs facial authentication using the facial feature information. Specifically, the authentication unit 145 compares the facial feature information extracted from the face region of the person to be authenticated with the facial feature information in the face information DB 102. As a result, the authentication unit 145 identifies a user ID whose facial feature information matches. Note that the facial feature information matches (matches) when the degree of match is equal to or greater than a predetermined value. The authentication unit 145 transmits the result of the facial authentication to the authentication terminal 40 via the communication unit 120.

[0050] FIG. 6 is a flowchart showing the flow of face information registration processing according to the second embodiment. First, the registration unit 141 of the authentication device 10 acquires a registration image of the user included in a face registration request (S21). For example, the registration unit 141 receives the face registration request from a user terminal (not shown) used by the user via the network N. Next, the registration unit 141 detects a face area included in the registration image (S22). Next, the registration unit 141 extracts feature points from the face area detected in step S22 and sets the extracted feature points as face feature information (S23). Finally, the registration unit 141 issues a user ID and associates the user ID with the face feature information to register it in the face information DB 102 (S24). Note that the authentication device 10 may receive face feature information from a source of the face registration request and register it in the face information DB 102 in association with the user ID.

[0051] 7 is a flowchart showing the flow of authentication processing according to embodiment 2. The following describes a case where the optical axes of first camera 410 and second camera 412 are approximately parallel to the traveling direction of a passerby, that is, parallel stereo vision.

[0052] First, the acquisition unit 142 receives and acquires the first image and the second image from the authentication terminal 40 via the communication unit 120 (S31). Next, the detection unit 143 detects a person's face area from each of the first image and the second image (S32). Then, the object identification unit 144 estimates a correspondence relationship between the people detected in both the first image and the second image (S33).

[0053] The processes shown in S32 and S33 will now be described in detail with reference to Figures 8 and 9. Figure 8 is a diagram showing a first image 500 captured by the first camera 410 according to the second embodiment. Figure 9 is a diagram showing a second image 502 captured by the second camera 412 according to the second embodiment.

[0054] For example, the detection unit 143 detects human face areas PA1, PA2, and PA3 from the first image 500. The detection unit 143 also detects human face areas PB1, PB2, and PB3 from the second image 502.

[0055] Next, in S33, the object identification unit 144 searches for the same person in the first image and the second image using various methods, and estimates the correspondence between the people detected in both images. Examples of methods for searching for the same person include the following methods 1 and 2.

[0056] (Method 1) The target identification unit 144 may infer that the same person appears in the first image 500 and the second image 502 based on the positional relationship between the people in the first image 500 and the positional relationship between the people in the second image 502. The direction from the bottom to the top of each of the first image 500 and the second image 502 is defined as the +v direction, and the direction from the left to the right is defined as the +u direction. In this case, the target identification unit 144 may determine that PA1, PA2, and PA3 are arranged in this order in the +v direction of the first image 500, and that PB1, PB2, and PB3 are arranged in this order in the +v direction of the second image 502. This allows the target identification unit 144 to infer that the facial regions of PA1 and PB1, PA2 and PB2, and PA3 and PB3, which are in the same order in the +v direction, belong to the same person. In the above example, the target identification unit 144 inferred that the same person appears based on the order in the +v direction. However, in addition to or instead of this, the target identification unit 144 may also infer based on the order in the +u direction. This makes it easier to search for the same person.

[0057] (Method 2) The target identification unit 144 may extract feature amounts of the facial region of each person included in the first image 500 and feature amounts of the facial region of each person included in the second image 502. The target identification unit 144 may then search for the same person included in the first image 500 and the second image 502 based on the difference in feature amounts of the facial region between each person included in the first image 500 and each person included in the second image 502. This makes it possible to easily search for the same person and to smoothly perform the subsequent authentication process.

[0058] The target identification unit 144 associates the facial regions of people who are estimated to be the same person by the above-described method 1 or 2 with each other, and proceeds to the next step.

[0059] Continuing the explanation, returning to Fig. 7, the target identification unit 144 calculates the position of each person based on the parallax between the first image 500 and the second image 502 (S34).

[0060] The process shown in S34 will now be described in detail with reference to Fig. 10. Fig. 10 is a diagram for explaining an example of a method for calculating a position according to embodiment 2. The principle of triangulation is used to calculate the position.

[0061] Here, the width direction of the passageway inside gate 60 is defined as the x-axis direction, and the direction of travel of pedestrians is defined as the y-axis direction. The x-axis direction and the y-axis direction are perpendicular to each other. The distance between first camera 410 and second camera 412 in the x-axis direction is L [m]. The directions of the optical axes OA1 and OA2 of first camera 410 and second camera 412 are approximately parallel to the y-axis direction. Point P in the figure indicates a feature point of the target person, and the coordinates of point P are (x p ,y p )

[0062] When the first camera 410 captures an image of a target person, point P in real space is projected onto point M1(u1,v1) in the first image 500. When the second camera 412 captures an image of a target person, point P in real space is projected onto point M2(u2,v2) in the second image 502.

[0063] In S34, first, the object identification unit 144 calculates the disparity D between the left and right cameras. Specifically, the object identification unit 144 may calculate the disparity D using the following formula. D=|u2-u1|

[0064] Then, the target identification unit 144 calculates the coordinates (x p ,y p ) using the parallax D, the distance L between the left and right cameras, and the focal length f. Specifically, the target identification unit 144 calculates the coordinates (x p ,y p ) may be calculated. x p =L / D y p =(f×L) / D

[0065] The target identification unit 144 can calculate the positions of all people commonly included in the first image 500 and the second image 502 by repeating the process of calculating the parallax and the process of calculating the coordinates of each person. Note that the target identification unit 144 does not necessarily have to calculate the positions of all people. For example, the target identification unit 144 may omit calculating the positions of people whose face regions are equal to or smaller than a predetermined size. This makes the process more efficient.

[0066] 7, the description will be continued. Based on the calculated position of the person, the object identification unit 144 identifies the person closest to the gate 60 as the person to be authenticated (S35). Then, the object identification unit 144 supplies information on the face area of ​​the identified person to be authenticated to the authentication unit 145.

[0067] The authentication unit 145 performs face matching based on the information on the facial area of ​​the person to be authenticated (S36). As a result, the authentication unit 145 obtains a face authentication result. At this time, the authentication unit 145 may perform face matching on both the facial area of ​​the person to be authenticated included in the first image and the facial area of ​​the person to be authenticated included in the second image, or may perform face matching on either one of them. The authentication unit 145 then transmits the face authentication result to the authentication terminal 40 (S37).

[0068] 11 is a flowchart showing the flow of the face matching process (S36 in FIG. 7) according to the second embodiment. First, the authentication unit 145 extracts facial feature information from the face area of ​​the person to be authenticated (S41). Next, the authentication unit 145 matches the extracted facial feature information with the facial feature information in the face information DB 102 (S42). If the degree of match of the facial feature information is equal to or greater than a predetermined value (Yes in S43), the authentication unit 145 determines that the face authentication has been successful (S44). On the other hand, if the degree of match of the facial feature information is less than the predetermined value (No in S43), the authentication unit 145 determines that the face authentication has failed (S45).

[0069] Note that, although methods 1 and 2 were given as methods for searching for the same person in S33, the present invention is not limited to these. As shown in FIG. 10, for example, if it is found that point P in real space is projected onto point M1(u1, v1) in the first image 500, it is projected somewhere on line LN2 where v = v1 in the second image 502. Therefore, the object identification unit 144 can efficiently search for the same person by limiting the search range to line LN2. Conversely, if it is found that point P in real space is projected onto point M2(u2, v2) in the second image 502, it is projected somewhere on line LN1 where v = v2 in the first image 500. Therefore, the object identification unit 144 can efficiently search for the same person by limiting the search range to line LN1.

[0070] Furthermore, in S35, the object identification unit 144 identifies the person closest to the gate 60 as the person to be authenticated, but instead, the object identification unit 144 may identify the person closest to the gate 60 and who has entered a predetermined passage area as the person to be authenticated. For example, the object identification unit 144 may identify the person closest to the gate 60 who has entered within a predetermined distance from the gate 60 as the person to be authenticated. This makes it possible to reduce the number of matching processes that do not contribute to determining whether the gate 60 has been passed, thereby enabling efficient authentication.

[0071] 7 and 10 have been described with reference to parallel stereoscopic vision, the angle between the optical axis of first camera 410 and the optical axis of second camera 412 may be less than 90 degrees. In other words, first camera 410 and second camera 412 may be disposed so that the optical axis of first camera 410 and the optical axis of second camera 412 are close to each other toward the center of the front.

[0072] 12 is a diagram illustrating an example of a method for calculating the position of point P in the case of such stereoscopic viewing. The optical axis OA1 of the first camera 410 and the optical axis OA2 of the second camera 412 are directed toward the center forward. Therefore, the common area of ​​the field of view can be made wider than in the case of FIG. 10.

[0073] In this case as well, point P in real space is projected onto point M1 (u1, v1) in the first image 500, and point P in real space is projected onto point M2 (u2, v2) in the second image 502. In this case, the object identification unit 144 may transform the projection plane (first image 500 and second image 502) using a rotation matrix so that the optical axes OA1 and OA2 become parallel. Then, the object identification unit 144 determines the coordinates (x p ,y p ) may be calculated.

[0074] In Fig. 12, the two cameras are located at different positions, so one camera can see the other, and the other camera is projected onto points E1 and E2. In Fig. 12, line LN1 passes through points E1 and M1, and line LN2 passes through points E2 and M2. Lines LN1 and LN2 are called epipolar lines.

[0075] When two cameras capture the facial regions of the same person, the facial regions lie on each other's epipolar lines. Therefore, even if methods 1 and 2 illustrated in S33 are not used, as long as the positional relationship between the two cameras is known, the target identification unit 144 can use a method similar to the method for searching for the same person described in FIG. 10. That is, if it is determined that point P in real space is projected onto point M1(u1, v1) in the first image 500, the target identification unit 144 may limit the search range to line LN2. Also, if it is determined that point P in real space is projected onto point M2(u2, v2) in the second image 502, the target identification unit 144 may limit the search range to line LN1. This allows for efficient searching for the same person.

[0076] In this way, the authentication system 1 according to the second embodiment achieves the same effects as those of the first embodiment. It has been difficult to measure the distance between the passerby and the gate by image processing, especially when the passerby is a child or in a wheelchair, but the second embodiment measures the distance using parallax, improving measurement accuracy.

[0077] It should be noted that the present disclosure is not limited to the above-described embodiments and may be modified as appropriate within the scope of the present disclosure. For example, in the above-described second embodiment, the authentication device 10a and the authentication terminal 40 are separate devices, but they may be configured as the same device.

[0078] The first camera 410 and the second camera 412 may also capture images in response to a motion sensor detecting the presence of a person near the gate 60. The authentication terminal 40 may also transmit a face authentication request to the authentication device 10a in response to detecting, by image recognition or a motion sensor, that a person is included in the first image and the second image.

[0079] Any processing in the present disclosure can be realized by causing a processor to execute a computer program.

[0080] In the above examples, the program includes instructions (or software code) that, when loaded into a computer, cause the computer to perform one or more functions described in the embodiments. The program may be stored on a non-transitory computer-readable medium or a tangible storage medium. By way of example and not limitation, computer-readable medium or tangible storage medium includes random-access memory (RAM), read-only memory (ROM), flash memory, solid-state drive (SSD) or other memory technology, CD-ROM, digital versatile disc (DVD), Blu-ray® disc or other optical disk storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage device. The program may also be transmitted on a transitory computer-readable medium or communication medium. By way of example and not limitation, transitory computer-readable medium or communication medium includes electrical, optical, acoustic, or other forms of propagated signals.

[0081] A part or all of the above-described embodiments can be described as, but not limited to, the following supplementary notes. (Appendix 1) a detection means for detecting a person from each of the first image and the second image, wherein the first image and the second image are images of a photographic area within a predetermined range based on the gate, photographed by a first camera and a second camera arranged at a predetermined distance in the width direction of the gate; an object specifying means for specifying, as an object to be authenticated, a person closest to a gate from at least one of a person detected from the first image and a person detected from the second image based on a parallax between the first image and the second image; an authentication means for controlling biometric authentication of the person to be authenticated; An authentication device comprising: (Appendix 2) The target identification means identifies the person who is closest to the gate and who has entered a predetermined passage area as the person to be authenticated. 10. The authentication device of claim 1. (Appendix 3) The target identification means measures the distance between a person presumed to be the same person included in a first image and a second image photographed at the same time or with a difference in photographing time less than a predetermined time, based on the parallax between the first camera and the second camera, and the person and the gate. 3. An authentication device according to claim 1 or 2. (Appendix 4) The target identification means estimates the same person included in the first image and the second image based on a feature amount of a face region of each person included in the first image and a feature amount of a face region of each person included in the second image. 10. The authentication device according to claim 3. (Appendix 5) The object identification means estimates the same person included in the first image and the second image based on a positional relationship between people included in the first image and a positional relationship between people included in the second image. 10. The authentication device according to claim 3. (Appendix 6) a first camera and a second camera arranged a predetermined distance apart in the width direction of the gate; an authentication device; the first camera generates a first image by capturing an image of a capturing area within a predetermined range based on the gate; the second camera generates a second image by capturing an image of a capturing area within a predetermined range based on the gate; The authentication device a detection means for detecting a person from each of the first image and the second image; an object specifying means for specifying, as an object to be authenticated, a person closest to a gate from at least one of a person detected from the first image and a person detected from the second image based on a parallax between the first image and the second image; an authentication means for controlling biometric authentication of the person to be authenticated; An authentication system comprising: (Appendix 7) The predetermined distance is equal to or greater than 0.5 m and equal to or less than 1.5 m. 1. The authentication system described in Appendix 6. (Appendix 8) 7. The authentication system of claim 6, wherein an angle between the optical axis of the first camera and the optical axis of the second camera is less than 90 degrees. (Appendix 9) an authentication terminal that acquires the first image and the second image from the first camera and the second camera, respectively, transmits the first image and the second image to the authentication device, and receives a face authentication result from the authentication device; 10. An authentication system as described in Appendix 6 or 7. (Appendix 10) Detecting a person from each of a first image and a second image, wherein each of the first image and the second image is an image of a photographing area within a predetermined range based on the gate, photographed by a first camera and a second camera arranged at a predetermined distance in the width direction of the gate; Identifying, as an authentication target, a person closest to a gate from at least one of the people detected from the first image and the people detected from the second image based on the parallax between the first image and the second image; controlling biometric authentication of the person to be authenticated; An authentication method comprising: (Appendix 11) a process of detecting a person from each of a first image and a second image, wherein the first image and the second image are images of a photographic area within a predetermined range based on the gate, photographed by a first camera and a second camera arranged at a predetermined distance apart in the width direction of the gate; an object identification process for identifying, as an object to be authenticated, a person closest to a gate from at least one of a person detected from the first image and a person detected from the second image based on a parallax between the first image and the second image; an authentication process for controlling biometric authentication of the person to be authenticated; A non-transitory computer-readable medium storing a program for causing a computer to execute the above. [Explanation of symbols]

[0082] 1. Authentication System 10,10a Authentication device 13 Detector 14 Target Identification Department 15 Authentication Section 100 Storage section 101 Program 102 Face Information DB 110 memory 120 Communications Department 130 Input section 140 Control Unit 141 Registration Department 142 Acquisition Department 143 Detector 144 Target Identification Department 145 Authentication Department 40 Authentication Terminal 410 Camera 1 412 Second Camera 420 Storage section 430 Communications Department 440 Display section 460 Control Unit 60 gates 61 First Gate Body 62 Second Gate Body 63 Flap 64 Reading unit 500 1st image 502 2nd image N Network

Claims

1. a detection means for detecting a person from each of the first image and the second image, wherein the first image and the second image are images of a photographic area within a predetermined range based on the gate, photographed by a first camera and a second camera arranged at a predetermined distance in the width direction of the gate; an object specifying means for specifying, as an object to be authenticated, a person closest to a gate from at least one of a person detected from the first image and a person detected from the second image based on a parallax between the first image and the second image; an authentication means for controlling biometric authentication of the person to be authenticated; Equipped with The target identification means for a person presumed to be the same person included in a first image and a second image captured at the same time or with a difference in capture time less than a predetermined time, measuring a distance between the person and the gate based on the parallax between the first camera and the second camera; The same person included in the first image and the second image is estimated based on a positional relationship between people included in the first image and a positional relationship between people included in the second image. Authentication device.

2. The target identification means identifies the person who is closest to the gate and who has entered a predetermined passage area as the person to be authenticated. The authentication device according to claim 1 .

3. The target identification means estimates the same person included in the first image and the second image based on a feature amount of a face region of each person included in the first image and a feature amount of a face region of each person included in the second image.

3. The authentication device according to claim 1 or 2.

4. a first camera and a second camera arranged a predetermined distance apart in the width direction of the gate; an authentication device; the first camera generates a first image by capturing an image of a capturing area within a predetermined range based on the gate; the second camera generates a second image by capturing an image of a capturing area within a predetermined range based on the gate; The authentication device a detection means for detecting a person from each of the first image and the second image; an object specifying means for specifying, as an object to be authenticated, a person closest to a gate from at least one of a person detected from the first image and a person detected from the second image based on a parallax between the first image and the second image; an authentication means for controlling biometric authentication of the person to be authenticated; and The target identification means for a person presumed to be the same person included in a first image and a second image captured at the same time or with a difference in capture time less than a predetermined time, measuring a distance between the person and the gate based on the parallax between the first camera and the second camera; The same person included in the first image and the second image is estimated based on a positional relationship between people included in the first image and a positional relationship between people included in the second image. Authentication system.

5. The predetermined distance is equal to or greater than 0.5 m and equal to or less than 1.5 m. The authentication system according to claim 4 .

6. the angle between the optical axis of the first camera and the optical axis of the second camera is less than 90 degrees; The authentication system according to claim 4 .

7. A computer comprising: Detecting a person from each of a first image and a second image, wherein each of the first image and the second image is an image of a photographing area within a predetermined range based on the gate, photographed by a first camera and a second camera arranged at a predetermined distance in the width direction of the gate; Identifying, as an authentication target, a person closest to a gate from at least one of the people detected from the first image and the people detected from the second image based on the parallax between the first image and the second image; controlling biometric authentication of the person to be authenticated; Run Identifying the person to be certified means: for a person presumed to be the same person included in a first image and a second image captured at the same time or with a difference in capture time less than a predetermined time, measuring a distance between the person and the gate based on the parallax between the first camera and the second camera; and estimating the same person included in the first image and the second image based on a positional relationship between people included in the first image and a positional relationship between people included in the second image. Authentication method.

8. a process of detecting a person from each of a first image and a second image, wherein the first image and the second image are images of a photographic area within a predetermined range based on the gate, photographed by a first camera and a second camera arranged at a predetermined distance in the width direction of the gate; an object identification process for identifying, as an object to be authenticated, a person closest to a gate from at least one of a person detected from the first image and a person detected from the second image based on a parallax between the first image and the second image; an authentication process for controlling biometric authentication of the person to be authenticated; on the computer, In the target identification process, for a person presumed to be the same person included in a first image and a second image captured at the same time or with a difference in capture time less than a predetermined time, measuring a distance between the person and the gate based on the parallax between the first camera and the second camera; The same person included in the first image and the second image is estimated based on a positional relationship between people included in the first image and a positional relationship between people included in the second image. program.

Images