JP6587594B2 - Face authentication crime prevention system - Google Patents

Description

  The present invention relates to a face authentication crime prevention system for performing face authentication of a visitor.

  Perform face authentication of the visitor to the building, determine whether the visitor is a registrant of the security system, such as a resident of the building, and if the visitor is a registrant, check the entrance of the building There is technology to unlock electric locks.

  However, if the control for unlocking the electric lock is executed only by confirming the presence of the registrant, the electric lock is unlocked even when an unregistered person accompanies the registrant. Since unregistered persons are not necessarily harmless, there is a possibility that undesired persons may be invited into the building in some cases.

  Patent Document 1 discloses a crime prevention system that estimates a person requiring attention from persons visiting a facility and notifies that a person requiring attention is coming to the facility. The crime prevention system described in Patent Literature 1 estimates that the person in the image is a person requiring attention when the captured image of the person captured by the camera matches the image of the attention candidate person stored in the unregistered person database.

Japanese Patent Laid-Open No. 2006-9234

  However, in the crime prevention system described in Patent Document 1, since a person who is not stored in the unregistered person database is not immediately estimated as a person who needs attention, the unlocking of the electric lock is performed when there is a registrant together with the unregistered person. There was a risk of accepting.

  The crime prevention system described in Patent Document 1 considers an unregistered person to be an important person when an unregistered person comes to the front door several times. There was a problem that the electric lock would be unlocked if accompanied.

  The present invention has been made in consideration of the above facts, and provides a face authentication crime prevention system capable of appropriately controlling locking and unlocking of an electric lock of a building when an unregistered person accompanies a registrant. With the goal.

  A first aspect for solving the above problems is provided in a photographing unit that photographs a visitor and obtains image data for authentication and a door of the entrance, and locks and unlocks the door based on the authentication. When face recognition is performed using the electric lock and the image data acquired by the photographing unit, and when the visitor is a registrant, the electric lock is unlocked, and the registrant is accompanied by an unregistered person Includes a control unit that controls the locking and unlocking of the electric lock according to the age of the registrant and the age of the unregistered person calculated from the image data.

  According to the first aspect, when an unregistered person accompanies the registrant, the registrant's age is more than a certain degree, or the registrant's age and the age of the unregistered person calculated from the image data are very small. If there is no difference, the electric lock can be unlocked.

  According to a second aspect, the control unit is configured so that the registrant's age is less than a predetermined age, and the electric lock is activated when the calculated age of the unregistered person is larger than the registrant's age by a predetermined value or more. Control to maintain the locked state.

  According to the second aspect, when the registrant is a preschool child or the like and the difference between the registrant's age and the registrant's age calculated from the image data is significant, the electric lock remains locked. Can be.

  According to a third aspect, the control unit determines that the registrant's age is less than a predetermined age and the unregistered person's age is greater than the registrant's age by a predetermined value or more. The registrant's image and warning information are transmitted to a predetermined portable information terminal.

  According to the third aspect, when the registrant is a preschool child or the like and the difference between the registrant's age and the age of the registrant calculated from the image data is significant, the registrant's guardian The image of the unregistered person and the warning information can be transmitted to the portable information terminal owned by.

  As described above, the face authentication crime prevention system according to the present invention can appropriately control locking and unlocking of an electric lock of a building when an unregistered person accompanies a registrant.

It is the schematic which shows an example of the building provided with the face authentication crime prevention system which concerns on embodiment of this invention. It is a block diagram which shows schematic structure of the controller in the face authentication crime prevention system which concerns on embodiment of this invention. It is the schematic which showed an example of the age estimation of the visitor in the face authentication crime prevention system which concerns on embodiment of this invention. It is the flowchart which showed an example of the process of the face authentication crime prevention system which concerns on embodiment of this invention.

  Hereinafter, an example of an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a schematic diagram illustrating an example of a building 100 including a face authentication crime prevention system 10 according to an embodiment of the present invention. As shown in FIG. 1, the face authentication crime prevention system 10 includes a controller 30, which is a control device, an interphone 44, a camera 40, a human sensor 42, and an electric lock 50 that locks the front door 52.

  As will be described later, the controller 30 is a kind of computer including a CPU (Central Processing Unit), and controls the camera 40, the interpersonal sensor 42, the interphone 44, and the electric lock 50. The controller 30 is connected to the cloud server 70 via the network 60 so as to be communicable. Information transmitted from the controller 30 to the cloud server 70 via the network 60 is configured to be transmitted to a portable information terminal 80 such as a smartphone.

  The camera 40 acquires moving images of the registered person 32 and the unregistered person 34 who have approached the building 100. The camera 40 may always shoot the entrance of the building 100 for 24 hours. In this embodiment, the camera 40 starts shooting when the person sensor 42 detects a visitor, and the intercom 44 button is pressed. If it is, the authentication of the registrant and the process of estimating the age and sex of the visitor may be started.

  For example, the interpersonal sensor 42 transmits an ultrasonic wave, and when the transmitted ultrasonic wave is reflected by a visitor such as the registrant 32 and the unregistered person 34, the presence of the visitor is detected by detecting the reflected wave. To detect. Alternatively, a laser sensor that irradiates a laser and detects the presence of the visitor from the change in wavelength and phase of the laser reflected by the visitor may be used. Furthermore, the infrared sensor which detects the infrared rays which a human body emits, and detects the presence of a visitor may be used.

  As will be described later, the controller 30 performs face authentication of the visitor and controls locking / unlocking of the electric lock based on the age estimation of the visitor or the height difference between the registrant 32 and the unregistered person 34. .

  FIG. 2 is a block diagram showing a schematic configuration of the controller 30 in the face authentication crime prevention system 10 according to the present embodiment. The controller 30 includes a CPU 12, an HDD (Hard Disk Drive) 14, a RAM (Random Access Memory) 16, a network I / F unit 18, a ROM (Read Only Memory) 20, an external I / F unit 26, including. The controller 30 includes a display unit 22, an operation input unit 24, and a bus 28, and has functions of a terminal for inputting information and a terminal for displaying information.

  The CPU 12 governs the overall operation of the controller 30. Face authentication program, age and gender estimation program, program for calculating height difference between registered user 32 and unregistered user 34, program for controlling electric lock 50, communication program with cloud server 70, OS (Operating System) and It is a non-volatile storage device in which data and the like for execution of each program are recorded. The HDD 14 also stores identification information, age, and the like of face authentication of a user such as a resident who is the registrant 32. The RAM 16 is a volatile storage device in which an OS, a program, or data is expanded. The network I / F unit 18 is for connecting to the network 60, and includes a NIC (Network Interface Card) and its driver. The ROM 20 is a non-volatile storage device that stores a boot program that operates when the controller 30 is activated. The display unit 22 displays information related to the face authentication crime prevention system 10 to the user. The operation input unit 24 is used when a user inputs operations and information of the face authentication crime prevention system 10, and as an example, an input device such as a touch panel and a keyboard and a pointing device such as a trackball, a pen tablet or a mouse, Furthermore, a microphone for voice input is included. The bus 28 is used when information is exchanged. The external I / F unit is an interface to which the camera 40, the interpersonal sensor 42, the interphone 44, and the electric lock 50 are connected.

  FIG. 3 is a schematic diagram showing an example of age estimation of visitors in the face authentication crime prevention system 10 according to the present embodiment. The face authentication crime prevention system 10 according to the present embodiment is based on the image data of the face portions of the registrant 32 and the unregistered person 34 acquired by the camera 40, and the age of the registrant 32 and the unregistered person 34 who are visitors. Estimate gender.

Inference by NN (neural network) is used to estimate age and sex from image data. As an example, the age and sex of a visitor are estimated by the following processing.
(1) Edge extraction from face part image data.
(2) A skin color region is extracted from the image data of the face portion.
(3) Extract an area corresponding to the skin color area extracted in (2) from the edge image generated in (1).
(4) Calculate cumulative histograms in the vertical and horizontal directions from the edge image corresponding to the skin color area.
(5) The number of pixels of the cumulative histogram where the luminance is equal to or higher than a predetermined threshold is set as NN input data.
(6) Based on the input data and the teacher data, the age and sex of the visitor are estimated by the NN.

  As the teacher data, for example, data specifying the number of pixels of the cumulative histogram in which the luminance is equal to or higher than a threshold value by age and gender is used. In general, young people like children have fewer facial irregularities and less skin dullness than adults, so the number of pixels in the cumulative histogram where the luminance exceeds a predetermined threshold tends to be higher as younger people. is there. In addition, women tend to have less facial irregularities and less dull skin than men.

  Through the processing described above, in this embodiment, as shown in FIG. 3, the age gender determination result 36 of the registrant 32 and the age gender determination result 38 of the unregistered person 34 are output. The determination processes shown in the above (1) to (6) are examples. Since there are various methods for estimating the age and sex of a person from image data, the age and sex of the person may be estimated using a method other than the above as appropriate.

  FIG. 4 is a flowchart showing an example of processing of the face authentication crime prevention system 10 according to the present embodiment. In step 400, the face authentication of the visitor is performed from the image data acquired by the camera 40. There are various face authentication methods, but image data is edge-processed, and the edge arrangement and direction of the constituent elements such as eyebrows, nose, lips, and eyes included in the edge-processed image data are used as input data. Authenticate against the data. If the determination in step 400 is affirmative, the procedure proceeds to step 406. If the determination in step 400 is negative, the electric lock 50 is kept locked in step 402, the image of the visitor (unregistered person 34) is recorded in step 404, and the process is terminated.

  In step 406, it is determined whether or not the registrant 32 is a specific registrant who is a preschool child under 5 years old. If the determination in step 406 is affirmative, it is determined in step 408 whether an unregistered person is accompanied. The under 5 years old is an example, and the registrant 32 may be regarded as a specific registrant when the registrant 32 is under a predetermined age of 4 to 10 years old.

  If the determination in step 408 is affirmative, it is determined in step 410 whether or not the accompanying person is older than a predetermined registrant (for example, five years old) by the age estimation process described above. Since data such as age is stored in the HDD 14 of the controller 30, the specific registrant calculates the age by referring to the storage. Since the accompanying person is an unregistered person, the age is estimated from the image data by the above-described age estimation by NN.

  If a negative determination is made in steps 406, 408, and 410, the electric lock 50 is unlocked in step 412 and the process is terminated.

  If the determination in step 410 is affirmative, gender estimation based on the above-described NN is performed in step 414 to determine whether or not the accompanying person is a male. If the determination in step 414 is affirmative, the electric lock 50 is not unlocked in step 416, the image is recorded in step 418, and the recorded image is transferred to the portable information terminal 80. In addition, a voice announcement such as “There are unregistered people and cannot be unlocked” is given to the accompanying person, and warning information is transmitted to the portable information terminal 80. The portable information terminal 80 that has received the alert information displays a pop-up indicating the alert on the screen, and notifies the guardian of the specific registrant who is the owner of the portable information terminal 80 that there is an emergency situation. Then, after step 418, the process ends.

  If the determination in step 414 is negative, the electric lock 50 is not unlocked in step 420, the image is recorded in step 422, and the recorded image is transferred to the portable information terminal 80. Furthermore, a voice announcement such as “unregistered people are mixed and cannot be unlocked” is given to the accompanying person, and caution information is transmitted to the portable information terminal 80. The mobile information terminal 80 that has received the warning information requires a pop-up display indicating a warning on the screen, and notifies the guardian of the specific registrant who is the owner of the mobile information terminal 80 that the situation requires warning. To do. Then, after step 422, the process ends.

  In the process of FIG. 4, the age is estimated in step 410, but the heights of the specific registrant and the accompanying person are calculated from the image data of the camera 40, and the accompanying person's height is a predetermined ratio with respect to the specific registrant. In the above case, an affirmative determination may be made in step 410. As an example, if the accompanying person's height is 40% or more of the specific registrant (140% or more of the specific registrant's height), an affirmative determination is made in step 410.

  Since the calculation for height calculation is lighter than the calculation for age estimation, affirmative or negative determination can be made quickly in step 410.

  As described above, according to the present embodiment, when the accompanying person of the specific registrant who is a preschool child is an unregistered person and the age difference is large, the electric lock 50 is protected without being unlocked. Since information indicating that the person should be wary is transmitted to the person who is not desired, it is possible to prevent an undesired person from being invited into the building 100 and to protect the specific registrant by contacting the police.

  Further, instead of estimating the age from the face image data, whether or not the electric lock 50 can be unlocked may be determined based on the height difference between the specific registrant and the accompanying person as described above. The calculation of the height difference can be handled even if the CPU 12 of the controller 30 has a low specification because the load on the computer is lighter than the age estimation.

  The controller 30 of the present embodiment corresponds to the control unit described in the claims, and the camera 40 corresponds to the imaging unit described in the claims.

10 Face authentication crime prevention system 12 CPU
14 HDD
16 RAM
18 Network I / F section 20 ROM
22 Display unit 24 Operation input unit 26 External I / F unit 28 Bus 30 Controller 32 Registrant 34 Unregistered person 36, 38 Age sex determination result 40 Interphone 40 Camera 42 Human sensor 44 Interphone 50 Electric lock 52 Door 60 Network 70 Cloud server 80 Mobile information terminal 100 Building

Claims (3)

A photographing unit that photographs a visitor and obtains image data for authentication;
An electric lock that is provided on the front door and locks and unlocks the door based on authentication;
Face authentication is performed based on the image data acquired by the photographing unit. If the visitor is a registrant, the electric lock is unlocked. If the registrant is accompanied by an unregistered person, the registrant A control unit for controlling the locking and unlocking of the electric lock according to the age of the user and the age of the unregistered person calculated from the image data;
A face authentication crime prevention system.   The control unit maintains the state where the electric lock is locked when the age of the registrant is less than a predetermined age and the calculated age of the unregistered person is greater than the registrant's age by a predetermined value or more. The face authentication crime prevention system of Claim 1 which performs control to perform.   The control unit, when the age of the registrant is less than a predetermined age, and the calculated age of the unregistered person is larger than the age of the registrant by a predetermined value or more, the image and warning of the unregistered person The face authentication crime prevention system of Claim 2 which transmits information to a predetermined | prescribed portable information terminal.