JP5001195B2 - Entrance management system - Google Patents

Description

The present invention relates to an entry management system that uses a plurality's co-face authentication equipment to authenticate more than one person of entry seekers at the same time.

  Conventionally, in an entrance management system provided in a building or the like, an individual's physical characteristics, for example, the iris of the face or eyes are read to confirm the identity of the person, and only when the person is a pre-registered admitted person Is allowed to enter. In this room management system, when a plurality of people enter the room at a time, the physical characteristics are read one by one to confirm the identity of the person. There is a problem.

  On the other hand, it is conceivable to allow a plurality of people to enter the room at the same time only by confirming one person's identity. However, if multiple persons are allowed to enter the room at the same time only by confirming one person's identity, there is a possibility that a suspicious person will be mixed into the plurality of persons. Therefore, for example, in the entrance management system disclosed in Patent Document 1, an input means for inputting the number of people entering the room and a counting means for counting the number of people actually entering the room are provided, and after the identity of one person is confirmed, the room is entered. The door is unlocked, and then the entrance door is locked when the number of persons counted by the counting means reaches the number of persons input from the input means.

JP 2001-40963 A Japanese Patent Laid-Open No. 10-134188 JP 2007-164668 A

  However, in the entrance management system disclosed in Patent Document 1 described above, one person is verified, but the other person is not verifying the identity, but only checking the number of persons entering the room. There is a security problem because it is not checked whether the person who did not perform registration is registered as an entry permitter. Further, when a person who is not registered as a person who enters the room enters the room, there is no evidence for identifying the person, so that there is a disadvantage that the person cannot be monitored or identified thereafter.

The present invention has been made in order to solve such problems, and the object of the present invention is a person who is not registered as an entry permitter when checking whether or not to permit entry of a plurality of persons at once. in but it is possible to limit without much stricter entry be mixed, to provide the incoming room management system that Ru can store specific information of entry who are not registered as entry permitter .

In order to achieve such an object, an entrance management system according to the present invention includes an imaging unit that simultaneously captures a face image of a plurality of applicants, and all members included in the image from the images captured by the imaging unit. Multiple-person simultaneous face comprising: a face image extracting means for extracting a face image of the person and a face image collating means for collating each face image extracted by the face image extracting means with a face image of a person permitted to enter the room The authentication device, the control device that unlocks the entrance door only when it is determined that the authentication is successful by the multiple face authentication device, and the actual number of people who want to enter the room located in a predetermined area in front of the entrance door The simultaneous face authentication apparatus for detecting the number of persons includes a number of face images extracted by the face image extracting means and a real number of persons who wish to enter the room detected by the real number detecting means. Match If not, the means for determining that authentication is unsuccessful, the number of face images of all the members extracted by the face image extracting means, and the actual number of people who want to enter the room detected by the actual number detecting means match, and the face A first authentication unit that determines that the authentication is successful when the face image collation unit confirms a match with the face image of the person permitted to enter the room for all the face images extracted by the image extraction unit, and the face image extraction unit extracts The number of face images of all the persons and the actual number of persons who wish to enter the room detected by the actual number detection means match, and the face image matching means does not match the face image of the person permitted to enter the room. When it is confirmed that a face image is mixed, at least a face image that does not match is recorded, and a second authentication unit that determines that authentication is successful is provided.

According to the present invention, the actual number of people who want to enter the room located in a predetermined area in front of the entrance door is detected, while the face images of a plurality of people who want to enter the room are taken at the same time. The face images of all the members included in the image are extracted from the images, and each of the extracted face images is collated with the face image of the person permitted to enter the room registered in advance. Here, it is confirmed that the number of extracted face images of all the persons coincides with the detected actual number of people who want to enter the room, and that all the extracted face images match the face images of the person permitted to enter the room. Then, it is determined that the authentication is successful, and the entrance door is unlocked. In addition, when the match with the face image of the person permitted to enter is not confirmed for all the extracted face images, and the face image that matches the face image of the person permitted to enter the room is mixed, At least a face image that does not match is recorded, it is determined that the authentication is successful, and the entrance door is unlocked. As a result, not only when the extracted face images of all persons match the face image of the person permitted to enter the room, but also when there are mixed face images that do not match and face images that do not match, it is determined that authentication is successful. Therefore, when checking whether or not to allow a plurality of people to enter the room at once, even if there are people who are not registered as entry permitters, it is possible to restrict entry without making it so severe.
In the present invention, it is determined that the authentication is unsuccessful when the number of the extracted face images of all the members does not match the detected actual number of people who want to enter the room. In this way, when capturing the face images of multiple people who want to enter the room at the same time, for example, if there is a person who is hidden in the shadow of the person or crouched so that it does not appear intentionally Since the number of face images of all the members and the actual number of people who wish to enter the room do not match, even if the face images of all the members extracted from the captured images match the face image of the person who has entered the room, Even if a face image that matches the face image and a face image that does not match are mixed, it is not determined that the authentication is successful.

In the entrance / exit management system of the present invention , the multiple face authentication device includes a face image matching means and a face image that does not match the face image of the person permitted to enter the room by the face image matching means as the second authentication means. If the matching face image is a face image of a specific person registered in advance, at least a face image that does not match is recorded, and a means for determining that authentication is successful is provided. Also good. In this way, for example, only when the administrator's face image is confirmed as a specific person, it can be determined that the authentication is successful, and identification of a room resident who is not registered as a room admitted person Information can be stored.

In the entrance / exit management system of the present invention , the multiple face authentication device extracts the face images of all persons included in the image from the captured image, but the face image may be a two-dimensional face image. It may be a three-dimensional face image. In the case of a two-dimensional face image, for example, there is a possibility that the face of a person image in a poster pasted on a wall is erroneously extracted as a face image. On the other hand, when a three-dimensional face image is used, only an actual human face having irregularities is extracted, and there is no possibility that the face of the person image in the poster is erroneously extracted as a face image.

  When the image to be extracted is a three-dimensional face image, as imaging means, a first imaging means for simultaneously imaging face images of a plurality of people who want to enter the room from one direction, and a direction different from the first imaging means. A second imaging unit that simultaneously captures face images of a plurality of people who want to enter the room, and as a face image extracting unit, a three-dimensional image from an image captured by the first imaging unit and an image captured by the second imaging unit; 3D image creation means for creating the image, and 3D face image extraction for extracting all 3D face images included in the image from the 3D image created by the 3D image creation means as the face images of all members Means may be provided.

According to the present invention, while detecting the actual number of people who want to enter the room in the predetermined area in front of the entrance door, the face images of a plurality of people who want to enter the room are picked up at the same time. All face images included in the image are extracted, and each extracted face image is collated with the face image of the person permitted to enter the room in advance. If the person's actual number matches and all the extracted face images match the face image of the person permitted to enter the room, it is determined that the authentication is successful, and all the extracted face images enter the room. If a match with the face image of the permitter is not confirmed, and a face image that matches the face image of the person permitted to enter the room is mixed with a face image that does not match, at least the face image that does not match is recorded and authenticated. Faces of all who are judged successful and extracted If the number of people detected and the actual number of people who want to enter the room do not match, it is determined that the authentication is unsuccessful. Not only when matching face images and face images that do not match are mixed, it is determined that authentication is successful, and when checking whether to allow multiple people to enter the room at the same time, Even if there are people who are not registered as entry permitters, it is possible to restrict entry without making it so severe.

  Hereinafter, the present invention will be described in detail with reference to the drawings.

[ Reference Example 1 ]
FIG. 1 is a diagram showing a main part of a reference example (reference example 1) before entering into the description of an embodiment of a room entrance management system using a multiple-person simultaneous face authentication apparatus according to the present invention. In the figure, 1 is a multiple face recognition device, 2 is a control device, 3 is an entrance door, and 4 is an electric lock provided for the entrance door. The multi-person simultaneous face authentication device 1 and the electric lock 3 are connected via the control device 2.

  In this room entry management system, the multiple-person simultaneous face authentication device 1 includes a CPU 1A, a RAM 1B, a ROM 1C, a storage device 1D, a CCD camera 1E, a display device 1F, and interfaces 1G to 1I.

The CPU 1A operates according to a program stored in the ROM 1C while accessing the RAM 1B. The ROM 1C stores a multiple face authentication program . In addition, a face image of a person who is permitted to enter the room is stored in the storage device 1D. In this example, the number of registered entry permitters is N, and one two-dimensional face image is stored for each of the N entry permitters.

FIG. 8 shows a functional block diagram of the main part of the multiple-person simultaneous face authentication device 1 in the room entrance management system of the first reference example . In this entrance management system, the multiple-person simultaneous face authentication device 1 includes an imaging unit 1-1 including a camera that simultaneously captures face images of a plurality of people who want to enter a room, and an image captured from the image captured by the imaging unit 1. A face image extraction unit 1-2 for extracting the face images of all members included in the image, a storage unit 1-3 for storing a two-dimensional face image of a person permitted to enter the room, and a face image extraction unit 1-2. A face image collation unit 1-4 that collates each face image (extracted face image) extracted by the above with a face image (registered face image) of the person permitted to enter the room stored in the storage unit 1-3, and all the extracted faces The first authentication unit 1-5 that determines that the authentication is successful when it is confirmed that the image matches the registered face image, and the extracted face image that does not match the extracted face image that matches the registered face image may coexist. If confirmed, at least the extracted face image does not match After it recorded in the memory 1-7, and a second authentication unit that determines the authentication success 1-6. In the multiple face authentication device 1, the face image extraction unit 1-2, the face image collation unit 1-4, the first authentication unit 1-5, and the second authentication unit 1-6 are realized as processing functions of the CPU 1A. Is done.

  Hereinafter, with reference to the flowchart shown in FIG. 2, the processing operation executed by the CPU 1A in accordance with the multiple face authentication program will be described.

  Now, it is assumed that a plurality of people who wish to enter the room stand in front of the CCD camera 1E and perform an operation to the effect that they want to enter the multiple-face simultaneous face authentication device 1. In response to this, the CPU 1A sends an imaging command to the CCD camera 1E, and simultaneously captures the face images of a plurality of people who want to enter the room standing in front of the CCD camera 1E (step S101).

  The CPU 1A captures an image captured by the CCD camera 1E, and extracts the face images of all members included in the image from the captured image (step S102). In this example, for example, using the technique disclosed in Patent Document 2, a two-dimensional face image included in the image is extracted. At this time, the number of extracted two-dimensional face images (hereinafter referred to as extracted face images) is t.

  Next, the CPU 1A sets n = 1 (step S103), and reads the face images (two-dimensional face images (hereinafter referred to as registered face images)) of N room-permitted persons stored in the storage device 1D. (Step S104). Then, after confirming that n> t is not satisfied (NO in step S105), the collation (1: N collation) between the n = 1 extracted face image and the N registered face images is performed (step S106).

  In the collation between the n = 1 extracted face image and the N registered face images, if the collation score exceeds the threshold th (YES in step S107), n of the N registered face images = It is considered that a face image matching the first extracted face image has been confirmed, and the NO. OK image NO. (Step S108), n = n + 1 = 2 (step S109), the process returns to step S105, and the n = 2nd extracted face image is compared with the N registered face images (step S106).

  In the collation between the n = 2nd extracted face image and the N registered face images, if the collation score exceeds the threshold th (YES in step S107), n of the N registered face images = It is considered that a face image matching the second extracted face image is confirmed, and the NO. OK image NO. (Step S108), n = n + 1 = 3 (step S109), and the process returns to step S105.

  In the collation between the n = 2nd extracted face image and the N registered face images, if the collation score does not exceed the threshold th (NO in step S107), n of the N registered face images = It is considered that a face image matching the second extracted face image has not been confirmed, and n = NO of the second extracted face image. No. of NG image. (Step S110), n = n + 1 = 3 (step S109), and the process returns to step S105.

  Here, for example, if the number t of extracted face images is 2, n> t (YES in step S105), the CPU 1A proceeds to the process in step S111.

  In step S111, if the match of all the extracted face images with the registered face image is confirmed (YES in step S111), the CPU 1A determines that the authentication is successful and displays “authentication OK” on the screen of the display device 1F. It is displayed (step S112). Further, a predetermined signal is transmitted to the control device 2 to notify that the authentication is successful (step S113). Upon receiving the notification of successful authentication, the control device 2 sends an unlocking command to the electric lock 4 and unlocks the entrance door 3. Further, the CPU 1A determines NO. Of the OK image stored in step S108. No. of the read OK image is read out. Information such as the shooting date and time is attached to the extracted face image corresponding to, and the extracted face image is filed as a history of face images of persons permitted to enter the room (step S114).

  In step S111, if the match with the registered face image is not confirmed for all the extracted face images (NO in step S111), the CPU 1A includes a face image that matches the face image of the person permitted to enter in the extracted face image. It is checked whether or not face images that do not match are mixed (step S115).

  Here, if it is confirmed that a face image that matches the face image of the person permitted to enter the room and a face image that does not match are mixed (YES in step S115), it is determined that the authentication is successful, and “authentication OK”. Is displayed on the screen of the display device 1F (step S116). Further, a predetermined signal is transmitted to the control device 2 to notify that the authentication has been successful (step S117). Upon receiving the notification of successful authentication, the control device 2 sends an unlocking command to the electric lock 4 and unlocks the entrance door 3.

  Further, the CPU 1A determines NO. Of the OK image stored in step S108. And NO. Of the NG image stored in step S110. No. of the read OK image is read out. NO. Of the extracted face image and NG image corresponding to. The extracted face image corresponding to is paired with the shooting date and time, information for distinguishing the OK image / NG image, and the like are attached, and the face image of the person permitted to enter the room is filed (step S118). In addition, an alarm is issued to inform that there are a mixture of persons who are not permitted to enter the room and those who are not registered as a room-permitted person (step S119).

  On the other hand, when it is confirmed that the face image that matches the face image of the person permitted to enter the room is not mixed with the face image that does not match, and that all the extracted face images do not match the face image of the person permitted to enter the room. (NO in step S115), the CPU 1A determines that the authentication is unsuccessful, and displays “authentication NG” on the screen of the display device 1F (step S120). In addition, the NG image NO. No. of the read NG image is read out. The extracted face image corresponding to is attached with information such as the shooting date and time and is filed as a face image history of the person who refused to enter the room (step S121). Further, an alarm is issued to inform that there is no person who is allowed to enter the room (step S122).

  In this example, the number t of the extracted face images is two. Similarly, when t ≧ 3, all the extracted face images are confirmed to match the registered face images, and the registered face images It is determined that the authentication is successful when a matching face image and a non-matching face image are mixed, and when all the extracted face images do not match the registered face image, it is determined that the authentication is unsuccessful.

  Further, in this authentication success / failure determination, when a face image that matches the registered face image and a face image that does not match are mixed, the face image determined not to match is left as an NG image. In this case, it is determined that the authentication is successful, but the face image stored as an NG image is sent to a monitoring device provided in a guardhouse or the like so that it can be determined whether or not it is a legitimate visitor. When it happens, it becomes a clue to identify the person by referring to the remaining NG image. In addition, since the OK image is stored as a pair in this NG image, it is possible to determine the person who has entered the room together.

In this way, in Reference Example 1 , not only when the extracted face images of all persons match the face image of the person permitted to enter the room, but also when the matching face image and the non-matching face image are mixed. Whether or not to allow multiple people to enter the room at a time by judging that authentication has succeeded with conditions and judging that the authentication is unsuccessful only when the extracted face images of all the members do not match the face image of the person permitted to enter the room In such a case, even if there are people who are not registered as entry permitters, it is possible to limit entry without making it so severe, and it is possible to store the specific information of those who are not registered as entry permitters.

  In the flowchart shown in FIG. 2, if it is confirmed in step S111 that a face image that matches the face image of the person permitted to enter the room and a face image that does not match are mixed, it is immediately determined that the authentication is successful. I did it. On the other hand, as in the flowchart shown in FIG. 3, step S123 is provided between step S115 and step S116, and it is confirmed whether or not the face image that matches the face image of the person permitted to enter the room is an administrator. If the user is an administrator, it may be determined that the authentication is successful.

  In the flowchart shown in FIG. 2, the OK image is left in step S114 or step S118. However, the OK image is not necessarily left. In other words, only the NG image may be left in step S118 or step S121.

  In the flowchart shown in FIG. 2, a two-dimensional face image is extracted from the image captured by the CCD camera 1E in step S102, but a three-dimensional face image is used instead of a two-dimensional face image. You may make it extract. In the case of a two-dimensional face image, for example, there is a possibility that the face of a person image in a poster pasted on a wall is erroneously extracted as a face image. On the other hand, if the face image to be extracted is a three-dimensional face image, only the actual human face with unevenness is extracted, and the face of the person image in the poster may be erroneously extracted as a face image. This disappears.

[ Reference Example 2 ]
FIG. 4 shows a main part of the room entry management system when a three-dimensional face image is extracted as Reference Example 2 . In the case of extracting a three-dimensional face image, two CCD cameras 1E1 and 1E2 are provided. The CCD camera 1E1 simultaneously captures face images of a plurality of people who want to enter the room from one direction, and the CCD camera 1E2 and the CCD camera 1E1. Takes images of the faces of a plurality of applicants from different directions at the same time.

Further, in the process according to the simultaneous face authentication program of the CPU 1A, a three-dimensional image is created from an image captured by the CCD camera 1E1 and an image captured by the CCD camera 1E2, for example, using the technique disclosed in Patent Document 3. Then, all the three-dimensional face images included in the created three-dimensional image are extracted as the face images of all members. Further, the storage device 1D stores a three-dimensional face image as a registered face image. In this reference example 2 as well, in the same manner as in reference example 1 , the authentication success / failure is determined according to the flowcharts shown in FIG. 2 and FIG.

FIG. 9 shows a functional block diagram of a main part of the multiple-person simultaneous face authentication device 1 in the room entrance management system (FIG. 4) of the reference example 2 . In this entrance management system, the multiple-person simultaneous face authentication device 1 includes a first imaging unit 1-9 including a camera that simultaneously captures face images of a plurality of people who want to enter the room from one direction, and the first imaging unit 1 The second imaging unit 1-10 including a camera that simultaneously captures face images of a plurality of people who want to enter the room from a direction different from -9, the image captured by the first imaging unit 1-9, and the second imaging unit 1 The 3D image creating unit 1-11 that creates a 3D image from the image captured by the image -10, and all three included in the image from the 3D image created by the 3D image creating unit 1-11 A three-dimensional face image extracting unit 1-12 for extracting a three-dimensional face image as a face image of all members, a storage unit 1-13 for storing a three-dimensional face image of a person who has been registered in advance, and a three-dimensional face Each face image (extracted face image) extracted by the image extracting unit 1-9 is When the face image collating unit 1-14 that collates with the face image (registered face image) of the person permitted to enter the room stored in the memory unit 1-10, and a match with the registered face image for all the extracted face images is confirmed If it is confirmed that the first authentication unit 1-15 that determines that the authentication is successful and the extracted face image that does not match the extracted face image that matches the registered face image are mixed, the extracted face image that does not match at least the memory And a second authentication unit 1-16 that determines that the authentication is successful. In this multiple-person simultaneous face authentication device 1, a 3D image creation unit 1-11, a 3D face image extraction unit 1-12, a face image collation unit 1-14, a first authentication unit 1-15, and a second authentication The unit 1-16 is realized as a processing function of the CPU 1A. Further, the three-dimensional image creation unit 1-11 and the three-dimensional face image extraction unit 1-12 constitute a face image extraction unit in the present invention.

Embodiment
In the reference examples 1 and 2 described above, when simultaneously capturing the face images of a plurality of people who want to enter the room, for example, when there is a person who hides in the shadow of the person or crouches so that it does not appear intentionally, It may be determined that authentication is successful. On the other hand, as shown in FIG. 5, the real number detection means 5 for detecting the real number tR of the person who wants to enter the room located in the predetermined area AR1 in front of the entrance door 3 is provided, and a multi-face simultaneous face authentication device is provided. 1, the number t of all face images extracted from the captured face image is compared with the actual number tR of the person who wants to enter the room detected by the real number detection means 5, and authentication is performed when the two do not match. If it is determined as unsuccessful, it is possible to prevent the authentication from being determined as successful. The system configured as described above is the room entry management system of the present embodiment.

  For example, as the real number detection means 5, a CCD camera is provided on the upper surface of the area AR1. An image of the head of the person who wants to enter the room is extracted from the image captured by the CCD camera, and the actual number tR of persons who wish to enter the room located in the area AR1 is counted from the extracted image of the head. In the multiple-person simultaneous face authentication device 1, in the process according to the multiple-person simultaneous face authentication program of the CPU 1A (FIG. 2), for example, as shown in FIG. 6, step S124 is provided between step S105 and step S111. The number t of the person who wants to enter the room and the actual number tR of the person who wants to enter the room are compared, and if the number t of the extracted face images and the actual number tR of the person who wants to enter the room do not match (NO in step S124), (Steps S120, S121, S122). As a result, even if the extracted face image of everyone matches the registered face image, or even if a face image that matches the face image of the person permitted to enter the room and a face image that does not match are mixed, it is determined that the authentication is successful. It is possible to prevent the suspicious person from entering the room.

  In FIG. 6, step S124 is provided between step S105 and step S111. However, as shown in FIG. 7, step S124 may be provided between step S102 and step S103. In this way, after the face image is extracted from the image captured by the CCD camera 1E, if the number t of the extracted face images does not match the actual number tR of people who want to enter the room, it is immediately determined that the authentication has failed. Will be.

It is a figure which shows the principal part of the reference example (reference example 1) before starting description of embodiment of the entrance management system using the multiple person simultaneous face authentication apparatus which concerns on this invention. It is a flowchart which shows the processing operation according to the multiple person simultaneous face authentication program which CPU of the multiple person simultaneous face authentication apparatus in this room entrance management system performs. It is a flowchart in the case where it is confirmed whether or not the face image that matches the face image of the person permitted to enter the room is an administrator when matching / mismatching are mixed. It is a figure which shows the principal part of the entrance management system ( reference example 2 ) at the time of making it extract a three-dimensional face image. It is a figure which shows the example which provided the real number detection means which detects the real number of persons who wish to enter the room located in the predetermined area in front of the entrance door. It is a flowchart which shows an example of the processing operation according to the multiple person simultaneous face authentication program which CPU of a multiple person simultaneous face authentication apparatus at the time of providing a real number detection means. It is a flowchart which shows another example of the processing operation according to the multiple person simultaneous face authentication program which CPU of a multiple person simultaneous face authentication apparatus at the time of providing a real number detection means. It is a functional block diagram of the principal part of the multiple-person simultaneous face authentication apparatus in the entrance management system of Reference Example 1 . It is a functional block diagram of the principal part of the multiple-person simultaneous face authentication apparatus in the entrance management system of the reference example 2 .

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Multiple-person simultaneous face authentication apparatus, 1A ... CPU, 1B ... ROM, 1C ... RAM, 1D ... Storage device, 1E, 1E1,1E2 ... CCD camera, 1F ... Display device, 1G-1J ... Interface, 2 ... Control device 3 ... Entrance door, 4 ... Electric lock, 5 ... Real number detection unit, AR1 ... Predetermined area, 1-1 ... Imaging unit, 1-2 ... Face image extraction unit, 1-3 ... Storage unit, 1-4 ... face image collation unit, 1-5 ... first authentication unit, 1-6 ... second authentication unit, 1-7 ... memory, 1-9 ... first imaging unit, 1-10 ... second imaging unit, DESCRIPTION OF SYMBOLS 1-11 ... 3D image preparation part, 1-12 ... 3D face image extraction part, 1-13 ... Memory | storage part, 1-14 ... Face image collation part, 1-15 ... 1st authentication part, 1-16 ... 2nd authentication part, 1-17 ... Memory.

Claims (3)

By means of imaging means for simultaneously capturing the face images of a plurality of people who wish to enter the room, face image extracting means for extracting the face images of all members included in the images from the images taken by the imaging means, and the face image extracting means A multiple-person simultaneous face authentication device comprising face image matching means for matching each extracted face image with a face image of a person permitted to enter the room registered in advance;
A control device that unlocks the entrance door only when it is determined that the authentication is successful by the multiple face authentication device;
An actual number detecting means for detecting the actual number of persons who want to enter the room located in a predetermined area in front of the entrance door;
The multiple-person simultaneous face authentication device
Means for determining that the authentication is unsuccessful if the number of face images of all persons extracted by the face image extraction means and the actual number of persons who wish to enter the room detected by the real number detection means do not match;
The number of face images of all persons extracted by the face image extracting means matches the actual number of people who want to enter the room detected by the real number detecting means, and all the face images extracted by the face image extracting means First authentication means for determining that authentication is successful when the face image collation means confirms a match with the face image of the person permitted to enter the room,
The number of face images of all persons extracted by the face image extracting means matches the actual number of people who want to enter the room detected by the actual number detecting means, and the person permitted to enter the room by the face image checking means. A second authentication unit that records at least a face image that does not match and determines that the authentication is successful when it is confirmed that a face image that matches the face image matches a face image that does not match. An entrance management system . By means of imaging means for simultaneously capturing the face images of a plurality of people who wish to enter the room, face image extracting means for extracting the face images of all members included in the images from the images taken by the imaging means, and the face image extracting means A multiple-person simultaneous face authentication device comprising face image matching means for matching each extracted face image with a face image of a person permitted to enter the room registered in advance;
A control device that unlocks the entrance door only when it is determined that the authentication is successful by the multiple face authentication device;
An actual number detecting means for detecting the actual number of persons who want to enter the room located in a predetermined area in front of the entrance door;
The multiple-person simultaneous face authentication device
Means for determining that the authentication is unsuccessful if the number of face images of all persons extracted by the face image extraction means and the actual number of persons who wish to enter the room detected by the real number detection means do not match;
The number of face images of all persons extracted by the face image extracting means matches the actual number of people who want to enter the room detected by the real number detecting means, and all the face images extracted by the face image extracting means First authentication means for determining that authentication is successful when the face image collation means confirms a match with the face image of the person permitted to enter the room,
The number of face images of all persons extracted by the face image extracting means matches the actual number of people who want to enter the room detected by the actual number detecting means, and the person permitted to enter the room by the face image checking means. If the face image that matches the face image and the face image that does not match are mixed, and the matching face image is a face image of a specific person registered in advance, at least a face image that does not match A room entry management system comprising: a second authentication unit that records and determines that authentication is successful . In the entrance management system according to claim 1 or 2,
The imaging means includes
First imaging means for simultaneously imaging the face images of the plurality of people who want to enter the room from one direction;
A second imaging unit that simultaneously captures the face images of the plurality of people who want to enter the room from a direction different from the first imaging unit;
The face image extraction means includes
3D image creation means for creating a 3D image from the image captured by the first imaging means and the image captured by the second imaging means;
3D face image extracting means for extracting all 3D face images included in the 3D image created by the 3D image creating means as the face images of all the members. Entrance management system .

Images