JP5813829B1 - Crime prevention system - Google Patents

Abstract

A crime prevention system for estimating a person requiring attention from among persons visiting a facility and notifying that a person requiring attention has come to the facility is provided. A crime prevention system including a caution person estimation device 1 that estimates whether or not a person imaged by a camera C installed in a facility is a cautionary person. The caution person estimation device 1 enters the facility. A registered person database 15 that stores the face image of the permitted person, a face authentication collating unit 12, and an unregistered person database 16 that stores, as the attention candidate person, the person who is determined not to match the admission permitted person by the face authentication collating unit 12. And have. The face authentication collation unit 12 performs face authentication collation processing for collating the acquired captured image with the image of the attention candidate person stored in the unregistered person database 16, and determines that the captured image matches the image of the attention candidate person. If the visit history of the candidate candidate for attention satisfies a predetermined requirement, the candidate candidate for attention is estimated to be a person requiring attention. [Selection] Figure 1

Description

  The present invention relates to a crime prevention system, for example, estimating whether or not a person imaged with a camera installed at the entrance of a facility is a person who needs attention, and the person responsible for the facility comes to the facility manager. It is related with the crime prevention system which notifies that it is.

  Conventionally, face authentication verification is performed in which a face image of a person visiting the facility is captured by a camera installed at the entrance of the facility, and the face image of the captured person is compared with the face image of a registered person (registrant). An entry / exit management device is known that permits entry into a facility when it is determined that the person who performs the process and visits the facility is a registrant. Such an entrance / exit management device is disclosed in Patent Document 1, for example.

  The entrance / exit management device described in Patent Document 1 captures a pedestrian's heel image with a camera while a pedestrian is passing through a traffic zone toward a facility door, and registers the captured face image in advance. It is determined whether or not the pedestrian is a pre-registered person by collating with the face image of the person who has been registered, and the door is unlocked if it is determined as a pre-registered person Yes.

JP 2005-146709 A

By the way, incidents where suspicious persons intrude frequently occur in facilities such as schools, and it is strongly desired to effectively prevent suspicious persons from entering. Therefore, when a person suspected of being a suspicious person arrives at the facility, the person who is suspected of being a suspicious person is discovered, and the person who is suspected of being a suspicious person has come to the facility to the manager responsible for the guard of the facility If there is a system that can be notified, the alert level of the person responsible for management can be increased. In other words, the supervisor can effectively prevent a suspicious person from entering a school or a company by monitoring a person suspected of being a suspicious person as a person requiring attention. However, currently no such system is known.
In addition, the above-described prior art entrance / exit management device can unlock the entrance when a pre-registered person arrives at the entrance. However, the person who is suspected of being a suspicious person among the visitors to the facility can be unlocked. It does not find out and notify the facility manager that the suspicious person is at the facility.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a crime prevention system for estimating a person requiring attention from among persons visiting the facility and notifying that the person requiring attention is coming to the facility. There is to do.

  The present invention made to solve the above problems is a crime prevention system comprising a caution person estimation device that estimates whether or not a person imaged by a camera installed in a facility is a cautionary person, wherein the caution person estimation The apparatus acquires a registered person database storing face images of persons permitted to enter the facility, a captured image of the face of the person captured by the camera, and the acquired captured image and stored in the registered person database. A face authentication collation unit that performs a first face authentication collation process for collating with a face image of an admission permitted person, and a person of a captured image that is determined not to match the face image of the admission permitted person by the first face authentication collation process And an unregistered person database that stores the captured image in association with visit history information of the attention candidate person in association with the captured image, and the face authentication verification unit includes the first face authentication When it is determined by the joint process that the captured image does not match the face image of the person permitted to enter, a second face that collates the acquired captured image with the image of the attention candidate person stored in the unregistered person database. When authentication verification processing is performed and it is determined by the second face authentication verification processing that the captured image matches the image of the attention candidate person, the image of the matching attention candidate person stored in the unregistered person database The associated visit history information is updated, and when the updated visit history information satisfies a predetermined requirement, the person in the acquired captured image is estimated to be a person requiring attention, and the person requiring attention comes to the facility. It is characterized by outputting information indicating that the

As described above, in the present invention, the person of the captured image that is determined not to match the face image of the person permitted to enter by the first face authentication collation process is set as the candidate candidate person, and the captured image of the candidate candidate person is used as the visit history information. An unregistered person database that is stored in association with each other is provided. According to the present invention, whenever the attention candidate person registered in the unregistered person database visits the facility (every time it is captured by the camera), the visit history information of the attention candidate person is updated, and the updated visit A candidate for attention is estimated as a person requiring attention when the history information satisfies a predetermined requirement. That is, in the present invention, the crime prevention system automatically stores persons other than the permitted entry person among the persons coming to the facility as caution candidate persons, and takes suspicious behavior such as the caution candidate person visiting the facility many times. The person is estimated as a person requiring attention, and information indicating that the person requiring attention is coming to the facility is output.
Therefore, according to the present invention, for example, when there is a suspicious person who visits a facility many times to preview a criminal act or the like, or visits a school or the like many times for voyeurism, such a suspicious person A person is estimated to be a person who needs attention, and the fact that the person who needs attention is coming to the facility is output. As a result, according to the present invention, it is possible to effectively prevent a suspicious person from entering a facility such as a school or a hospital.

  Further, the face authentication collation unit determines the person of the captured image when the second face authentication collation process determines that the captured image does not match the image of the attention candidate person stored in the unregistered person database. It is desirable to select a candidate for attention and store the visit history information of the candidate candidate for attention in association with the captured image in the unregistered person database. The predetermined requirement is preferably the number of visits to the facility within a predetermined period.

According to the above configuration, when a person that does not match the person stored in the registered person database and does not match the attention candidate person stored in the unregistered person database comes to the facility, the captured image of the person is displayed. It is stored in the unregistered person database as an image of the attention candidate person. In other words, in the present invention, when a person other than the admission permitted person visits the facility for the first time, the crime prevention system automatically registers the person as an attention candidate person.
Therefore, according to the present invention, if a facility user, a facility employee, a person in the vicinity of the facility, etc. is registered in the registered person database, it is registered as a caution candidate person when a person unrelated to the facility comes If the caution candidate person has taken a suspicious action such as coming to the facility many times, the person can be warned as a cautionary person.

  The attention person estimation device is connected to a server device via a network, and when the face authentication collation unit estimates that the person in the captured image is a person of interest, the server device is notified of the attention When the captured image of the person estimated to be a person is transmitted and the server device receives the captured image transmitted by the attention person estimation device, the server device requires the attention to the information terminal registered in advance via the network. It is desirable to notify that the person is in the facility and to provide information on the person requiring attention including the captured image transmitted from the attention person estimation device.

As described above, in the crime prevention system of the present invention, when it is estimated that the person visiting the facility is the person who needs attention, the person who needs attention comes to the facility in the information terminal registered in the server device in advance. In addition, the information about the person requiring attention including the captured image of the person estimated as the person requiring attention is provided.
Therefore, if the information terminal (information such as information terminal identification information and e-mail address) of the person responsible for management of the facility guard is registered in the server device, the person responsible for the suspicious person suspected to be the person responsible for management. Can be notified that the person is in the facility, and a captured image of the person who needs attention can be provided. In other words, the person in charge of the facility can easily confirm which person should be monitored because the face of the person requiring attention can be visually confirmed when the person requiring attention comes to the facility. Further, in a facility having a plurality of managers, a plurality of managers can simultaneously see and grasp the images of people who need attention, so that an effective crime prevention system is established.

  According to the present invention, it is possible to provide a crime prevention system for estimating a person requiring attention from among persons visiting a facility and notifying that a person requiring attention is coming to the facility.

It is the schematic diagram which showed the whole structure of the crime prevention system of this embodiment. It is the schematic diagram which showed the registered person database used for the crime prevention system of this embodiment. It is the schematic diagram which showed the unregistered person database used for the crime prevention system of this embodiment. It is the schematic diagram which showed the terminal database used for the crime prevention system of this embodiment. It is the flowchart which showed the procedure of the setting process of the smart phone used with the crime prevention system of this embodiment. It is the flowchart which showed the procedure of the attention person estimation process which the attention person estimation apparatus of this embodiment performs. It is the flowchart which showed the procedure of the process in which the information provision server of this embodiment provides the information regarding a person requiring attention to the smart phone of the person in charge of management of a facility. It is the schematic diagram which showed the login screen of the smart phone used by this embodiment. It is the schematic diagram which showed the screen which displays the information regarding the person requiring attention which the crime prevention system of this embodiment provides to a smart phone.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. First, the structure of the crime prevention system of this embodiment is demonstrated using FIGS.

  As shown in FIG. 1, the crime prevention system according to the present embodiment includes a camera (for example, a CCD camera) C installed in a facility such as a school, an apartment house, a hospital, or a company, and a person photographed by the camera C requires attention. And the information providing server (server device) 3 connected to the attention person estimating device 1 via the network NW (Internet).

In addition, when the facility is a school (or kindergarten or nursery school), the camera C is installed at the main gate, back gate, or bicycle parking lot of the school (or kindergarten or nursery school) and approaches the front, back gate, or bicycle parking lot. It is set to be able to image the face.
For example, when the facility is an apartment house such as an apartment, the camera C is installed at the entrance (entrance) of the apartment house, the stairway entrance, the parking lot entrance, etc., and the entrance of the apartment house, the staircase entrance, the parking lot entrance, etc. It is set so that the face of a person approaching can be imaged.
For example, when the facility is a hospital, the camera C is installed at the front gate, back gate, emergency entrance, and night entrance of the hospital so that the face of a person approaching the main gate, back gate, emergency entrance, and night entrance can be imaged. Is set to

The attention person estimation device 1 is installed in a guard room or the like of a facility where the camera C is installed, and is configured to exchange various information with the information providing server 3 via the network NW. In addition, the smartphone (personal digital assistant) SP is held by a person in charge of management such as a guard who is resident in the facility.
Further, a push notification server 7 for performing push notification to the smartphone SP is connected to the network NW.

  The attention person estimation apparatus 1 is connected to a display unit 5 configured by a liquid crystal display or the like and an input unit 6 configured by a keyboard, a mouse, or the like. And the display part 5 receives the image information transmitted from the attention person estimation apparatus 1, and displays the image information.

  In addition, the attention person estimation device 1 acquires a captured image of a person captured by the camera C (captured image including a face portion), and according to the processing procedure shown in FIG. A process of estimating whether or not the person is an attention person is performed. Further, when the attention person estimation device 1 estimates that the person in the captured image is a person requiring attention, the display unit 5 displays “information indicating that a person requiring attention is present at the facility (for example, text information)”. And “a captured image in which a person requiring attention is displayed” and “information indicating that a person in need of attention has come to the facility” and “a captured image in which a person requiring attention is captured” are displayed on the information providing server 3. Then, “facility information” described later is transmitted.

In addition, when the information providing server 3 receives “information indicating that a person requiring attention”, “captured image”, and “facility information” are received from the attention person estimation device 1, the smartphone SP ( The smartphone SP held by the facility guard or the like is notified via the network NW that a person requiring attention is coming to the facility, and includes a “captured image of the person requiring attention”. Information is provided (see FIG. 9).
Hereinafter, each apparatus which comprises the crime prevention system of this embodiment is demonstrated in detail.

First, the configuration of the attention person estimation device 1 will be described.
The attention person estimation device 1 includes a control unit 10, an image processing unit 11, a communication processing unit 12, a face authentication verification unit 13, a registered person database 15, and an unregistered person database 16.

  The hardware configuration of the attention person estimation device 1 is not particularly limited. For example, a central processing unit (CPU), a main storage device, an auxiliary storage device, an input / output I / F, and a communication I / F are included. It is comprised by the computer (one or several computer) provided.

Further, the auxiliary storage device stores a program for realizing the functions of the respective units (the control unit 10, the image processing unit 11, the communication processing unit 12, and the face authentication collation unit 13). The auxiliary storage device stores a registered person database 15 and an unregistered person database 16.
The auxiliary storage device stores information (facility information) related to the facility where the attention person estimation device 1 is installed. In the following, the facility information is information including a facility ID (unique data composed of alphanumeric characters) identifying the facility, a facility name, and a contact information (address, telephone number, e-mail address, etc.) Take as an example. In addition, the facility ID is assigned by a provider that provides the security system service of the present embodiment.

  And the function of each part (the control part 10, the image process part 11, the communication process part 12, and the face authentication collation part 13) of the attention person estimation apparatus 1 loads the program stored in the auxiliary storage device into the main storage device. It is realized by executing.

Next, each component of the attention person estimation device 1 will be described.
The registered person database 15 is a database that stores information on a person (admission permitted person) who is permitted to enter the facility. The registered person database 15 includes “gender”, “age”, “name”, and “attribute” of the permitted person. Correspondingly, image data obtained by capturing the face of the person permitted to enter is stored. This admission person includes, for example, facility users (students and parents if the facility is a school), facility employees (faculty and administrative staff if the facility is a school), neighbors living in the facility, facilities It is envisaged that there are traders in and out of the city.

  Specifically, as shown in FIG. 2, the registered person database 15 includes a field 15a for registering gender, a field 15b for registering age, a field 15c for registering the name of the person permitted to enter, It is composed of a record having a field 15d for registering attribute information and a field 15e for registering image data (for example, image data in jpeg format) showing the face of the person permitted to enter (consisting of a plurality of records). ) For example, when the facility is a school, the attribute information registered in the field 15d is information on students, parents of students, faculty members, office workers, residents in the vicinity of the school (neighbors), and the like.

  In addition, the unregistered person database 16 sets a person who has been determined not to match the person permitted to enter by the face authentication collation process among persons photographed by the camera C (person who has come to the facility) as a candidate for attention, and the candidate for attention. It is a database that stores information about people. The unregistered person database 16 is configured to store visit history information on the facility of the attention candidate person in association with the image of the attention candidate person (an image showing a face).

Specifically, as shown in FIG. 3, the unregistered person database 16 includes a field 16a for registering the attention candidate person ID and image data (for example, image data in jpeg format) showing the face of the attention candidate person. It consists of a record having a field 16b to be registered, a field 16c for registering a visit date and time to the facility, and a field 16d for registering the number of visits to the facility during a predetermined period (consisting of a plurality of records). ).
Note that the attention candidate person ID registered in the field 16a is unique data (data composed of alphanumeric characters or the like) set for each attention candidate person. The predetermined period is appropriately set by the security system administrator. For example, “3 hours (current date to 3 hours before)”, “1 day (current date to 24 hours before) Period) "," one week (current date and time to 7 days before) "

The control unit 10 shown in FIG. 1 controls the overall operation of the attention person estimation device 1 and accepts various data inputs from the input unit 6.
In addition, the image processing unit 11 acquires a human image (an image including a face) captured by the camera C and performs image processing such as color adjustment. The communication processing unit 12 exchanges various information with the information providing unit 31 of the information providing server 3 via the network NW.

  The face authentication verification unit 13 acquires a captured image of the person on which image processing has been performed by the image processing unit 11, and performs face authentication verification on the acquired captured image and the image data registered in the registered person database 15. To determine whether the captured image of the person who has come to the facility matches the image data of the permitted entry person stored in the registered person database 15.

Further, when it is determined by the first face authentication verification process that the acquired captured image does not match the image data of the person permitted to enter, the face authentication verification unit 13 registers the acquired captured image and the unregistered person database 16. A face authentication collation process (referred to as “second face authentication collation process”) for collating the image data of the attention candidate person that has been performed is performed.
Then, when it is determined that the captured image matches the image of the attention candidate person as a result of the second face authentication verification process, the face authentication verification unit 13 matches the data registered in the unregistered person database 16. The history information (visit date / time, number of visits) of the attention candidate person determined to have been updated is updated. When the face authentication collation unit 13 updates the unregistered person database 16, first, a new visit date / time is additionally registered in the field 16c, and then the visit date / time and the current date / time registered in the field 16c. The number of visits for a predetermined period is calculated, and the calculated number of visits is registered (overwritten) in the field 16d. The current date and time is acquired from a clock (not shown) built in the attention person estimation device 1.

Further, the face authentication collation unit 13 determines that the number of visits within a predetermined period of the updated attention candidate person is a predetermined number or more (for example, five or more when the predetermined period is set to one day). The person of the captured image that matches the attention candidate person is estimated to be a person requiring attention.
Further, when the face authentication collating unit 13 estimates that the person in the captured image is a person requiring attention, the display unit 5 displays “information indicating that a person requiring attention is present at the facility (for example, text information)”. And “a captured image of a person requiring attention”, “a captured image of a person requiring attention”, “a captured image of a person requiring attention”, and the information providing server 3 via the communication processing unit 12. “Facility information (facility information registered in advance)” is transmitted.

  If it is determined that the captured image does not match the attention candidate person as a result of the second face authentication verification process, the face authentication verification unit 13 newly sets the person in the captured image as the attention candidate person in the unregistered person database 16. sign up. Specifically, the face authentication verification unit 13 assigns a new “attention candidate person ID” to the person in the captured image, and associates the captured image and history with the “notified candidate person ID” in the unregistered person database 16. Register information (visit date / time, number of visits). At this time, the date and time when the captured image was captured (or the date and time when the face authentication collation unit 13 acquired the captured image) is registered in the field 16c, and “once” is registered in the field 16d.

Next, the configuration of the information providing server 3 will be described.
The information providing server 3 includes a control unit 30, an information providing unit 31, a terminal database 35, and a caution person database 36.
The hardware configuration of the information providing server 3 is not particularly limited, and includes, for example, a central processing unit (CPU), a main storage device, an auxiliary storage device, an input / output I / F, and a communication I / F. It is composed of computers (one or a plurality of computers).
The auxiliary storage device stores programs for realizing the functions of the respective units (the control unit 30 and the information providing unit 31). A terminal database 35 and a caution person database 36 are stored in a predetermined area of the auxiliary storage device.
And the function of each part (the control part 30, the information provision part 31) is implement | achieved by loading the said program stored in the auxiliary storage device to the main storage device, and executing it.

Next, each component of the information providing server 3 will be described.
The terminal database 35 includes an identifier for identifying the smartphone SP held by the manager in association with the “user ID (unique data) and login password assigned to the manager (such as a guard)” of the facility. This is a database that stores a device token (terminal identification information) that is included in association with facility information of a facility to which the manager in charge belongs.

  Specifically, as shown in FIG. 4, the terminal database 35 includes a field 35a for registering a user ID, a field 35b for registering a password, a field 35c for registering a device token, and facility information (facility ID, facility And a field 35d for registering (name and contact address) (consisting of a plurality of records).

  The attention person database 36 is a database for registering “a captured image of a person estimated to be a person requiring attention” transmitted from the face authentication verification unit 13, and corresponds to date / time information (reception date / time of a captured image). In addition, “a captured image of a person estimated to be a person requiring attention” and “facility information visited by a person requiring attention” are stored.

1 controls the operation of the information providing server 3 as a whole.
When the information providing unit 31 receives the “information indicating that a person requiring attention”, “the captured image of the person requiring attention”, and “facility information” transmitted from the attention person estimation device 1, the terminal database 35 is received. To identify the smartphone SP associated with the facility information. In addition, the information providing unit 31 notifies the identified smartphone SP that a person requiring attention is visiting the facility (see FIG. 9A), and the “imaging” transmitted from the person-of-care estimating apparatus 1 A person-of-interest information providing site 120 (FIG. 9B) including “image” is provided.
Further, the information providing unit 31 registers the received “captured image” and “facility information” in the visited person database 36 in association with the date / time information (year, month, day, time).

Next, processing contents performed by the crime prevention system of the present embodiment will be described.
First, the setting process of the smartphone SP for using the crime prevention system of the present embodiment will be described with reference to FIGS. 5 and 8.
FIG. 5 is a flowchart showing the procedure of the smartphone setting process used in the security system of this embodiment. FIG. 8 is a schematic diagram showing a login screen of a smartphone used in the present embodiment.

In order to use the crime prevention system according to the present embodiment, a manager (user) such as a facility guard installs a dedicated application program (the crime prevention system cooperation application 20) on his / her smartphone SP, It is necessary to set the smartphone SP so that it can function in cooperation with the security system.
And when the security system cooperation application 20 is installed in the smart phone SP, initial setting is performed, and the security system cooperation application 20 is activated, the smart phone SP receives a push notification from the information providing server 3. At the same time, the access information (site address) of the watchable person information providing site 120 (FIG. 9B) provided by the information providing server 3 is presented, and the watchable person information providing site 120 is displayed on the display unit 3 of the smartphone SP. Can be displayed.

Specifically, first, the user of the smartphone SP installs the crime prevention system cooperation application 20 on the smartphone SP from an application server (for example, an Apple Apple application server) (not shown), and performs initial setting (S1).
Here, the initial setting will be described. When the user activates the crime prevention system cooperation application 20 installed on the smartphone SP, a registration request is transmitted to the push notification server 7 via the OS (Operating System) of the smartphone SP. In response to this registration request, the push notification server 7 generates a device token of the smartphone SP that has requested registration, and transmits the device token to the smartphone SP. When the smartphone SP receives the device token from the push notification server 7, the smartphone SP displays an initial setting screen (not shown) provided by the security system cooperation application 20, and the user ID (unique data composed of alphanumeric characters or the like). ), Password, and facility information (facility ID, facility name, and facility contact information) are entered. Upon receiving the user ID, password, and facility information, the smartphone SP registers “user ID, password, and facility information” in the information providing server 3 (terminal database 35). Further, the smartphone SP stores the device token from the push notification server 7 in a built-in memory.
It is assumed that the user of the smartphone SP has acquired the facility information that needs to be input in the initial setting from the provider that provides the security system service of the present embodiment prior to the initial setting.

  When the user starts the crime prevention system cooperation application 20 after finishing the initial setting of S1, the smartphone SP displays the login screen 100 shown in FIG. 8 (S2). A user inputs a user ID and a password on the login screen 100 to perform login processing (S3, S4).

Specifically, when receiving the input of “user ID and password” from the user, the smartphone SP transmits “user ID and password” to the information providing unit 31 of the information providing server 3. If the “user ID and password” sent from the smartphone SP matches the terminal database 35 and the “user ID and password” is correct, the information providing unit 31 sends the login success screen 101 shown in FIG. 8 to the smartphone SP. Then, the login success screen 101 is displayed on the smartphone SP, and the process proceeds to S5.
On the other hand, if the received “user ID and password” does not match that registered in the terminal database 35 as a result of the collation, the information providing unit 31 transmits the login failure screen 102 shown in FIG. 8 to the smartphone SP. Then, the login failure screen 102 is displayed on the smartphone SP, the process returns to S3, and the store clerk is urged to re-enter the “user ID and password”.

In S <b> 5, the smartphone SP transmits its device token (device token acquired from the push notification server 7) to the information providing unit 31 of the information providing server 3.
Further, when the information providing unit 31 of the information providing server 3 receives the device token transmitted from the smartphone SP, the information providing unit 31 receives the device token in the record in which the “user ID and password” input in S3 is registered in the terminal database 35. The registered device token is registered (S6).

  When the process of S6 ends, the smartphone SP functions in cooperation with the security system by the security system cooperation application 20. That is, after the process of S6 is completed, information indicating that a person requiring attention is pushed to the facility is pushed to the smartphone SP every time the person-of-care estimating apparatus 1 estimates that a person requiring attention is present at the facility. In addition to being notified, access information to the person-of-interest information providing site 120 (FIG. 9B) including an image of the person requiring attention is provided.

Next, the process which the crime prevention system of this embodiment performs is demonstrated using FIGS.
FIG. 6 is a flowchart showing the procedure of the caution person estimation process performed by the caution person estimation apparatus of the present embodiment. FIG. 7 is a flowchart illustrating a processing procedure in which the information providing server according to the present embodiment provides information about the person requiring attention to the smartphone of the facility manager. FIG. 9 is a schematic diagram showing a screen for displaying information on a person requiring attention provided to the smartphone by the crime prevention system of the present embodiment.

First, the process performed by the attention person estimation device 1 will be described with reference to FIG.
As illustrated, the image processing unit 11 of the attention person estimation device 1 acquires a captured image (an image including a person's face) of a person approaching the facility captured by the camera C (S10). A face image is extracted from an image including a face, and image processing such as color adjustment is performed on the extracted face image (S11).
Note that two or more persons may enter the imaging range of the camera C, and the camera C may capture images of a plurality of persons. In this case, the image processing unit 11 extracts face images of a plurality of persons, and performs image processing such as color adjustment for each extracted face image.

Next, the face authentication collation unit 13 acquires “a face image obtained by imaging a person who has come to the facility” which has been subjected to image processing by the image processing unit 11, and the gender and age of the person who has come to the facility from the acquired face image Is estimated (S12). For example, the face authentication verification unit 13 estimates from the acquired face image that the visitor is “male” and “age of about 20-30 years old”.
Note that the “sex and age estimation process” performed in S12 uses a known technique, and thus a detailed description thereof is omitted.

Next, the face authentication collation unit 13 accesses the registered person database 15 and extracts a record in which the gender and age estimated in S12 are registered, and image data of the permitted entry person registered in the extracted record ( Face image) and a captured image (face image) of a person who came to the facility captured by the camera C acquired from the image processing unit 11 by face authentication verification processing (first face authentication verification processing), It is determined whether or not the image data of the person and the face image of the person captured by the camera C match (S13).
By this step, “determination that the captured image of the person who came to the facility matches the image of the permitted person stored in the registered person database 15” and “the captured image of the person who came to the facility is the registered person database. The determination result of any one of “determination that the image does not match the image of the permitted entry person stored in 15” is obtained.
In addition, since the face authentication collation process performed by S13 uses a well-known technique, detailed description is abbreviate | omitted.

  As a result of the processing of S13, when “determination that the captured image of the person who came to the facility matches the image of the permitted person stored in the registered person database 15” is made, that is, in the extracted record If there is a record in which image data matching the acquired captured image (face image) is registered, it is not necessary to monitor the person in the captured image, and the process returns to S10. As a result of the processing of S13, when “determination that the captured image (face image) of the person who came to the facility does not match the image of the permitted person stored in the registered person database 15” is made, that is, If there is no record in which the image data that matches the acquired captured image is registered in the extracted record, in order to determine whether or not the person of the captured image is a person requiring attention, S15 Proceed to the subsequent processing (S14).

In S15, the face authentication collation unit 13 accesses the unregistered person database 16, extracts a record in which image data associated with the gender and age estimated in S12 is registered, and the extracted record field 16b. The face authentication collation process (second face) is performed on the image data (face image) of the attention candidate registered in the image and the captured image (face image) of the person who came to the facility captured by the camera C and acquired from the image processing unit 11. Authentication verification processing) is performed, and it is determined whether the image data of the attention candidate matches the captured image of the person who has come to the facility.
By this step, “determination that the captured image of the person who came to the facility matches the image of the attention candidate person stored in the unregistered person database 16” and “the captured image of the person who came to the facility is not registered. One of the determination results “determination that the image does not match the image of the attention candidate person stored in the person database 16” is obtained.
Note that the face authentication collation process performed in S135 is a known technique as in S13, and thus detailed description thereof is omitted.

  As a result of the processing of S15, when “determination that the captured image of the person who has come to the facility matches the image of the attention candidate person stored in the unregistered person database 16” is made, that is, the extracted record If there is a record in which image data that matches the acquired captured image is registered, the process proceeds to S18. In addition, as a result of the processing of S15, if “determination that the captured image of the person who has come to the facility does not match the image of the attention candidate person stored in the unregistered person database 16” is made, that is, extracted. If there is no record in which no image data matching the acquired captured image is registered, the process proceeds to S17 (S16).

In S17, the person of the acquired captured image (face image) is newly registered in the unregistered person database 16 as a candidate for attention.
Specifically, the face authentication verification unit 13 assigns a new “attention candidate person ID” to the person of the acquired captured image and associates it with the assigned “attention candidate person ID” in the unregistered person database 16. The captured image and history information (visit date / time, number of visits) are registered. Then, when the process of S17 ends, the process returns to S10.
Such a processing procedure is adopted in the crime prevention system according to the present embodiment, when a person other than an entrance-permitted person visits the facility for the first time, the person is not suddenly estimated as a person requiring attention. This is because it is registered as a person and can be estimated as a person requiring attention when there is a suspicious behavior such as frequent visits to the facility thereafter.

Next, the processing after S18 will be described.
In S <b> 18, the face authentication matching unit 13 records history information (in the data registered in the unregistered person database 16) of records in which attention candidate persons determined to have been matched by the second face authentication matching process are registered ( (Visit date, number of visits) is updated, and the process proceeds to S19.
Specifically, the face authentication collation unit 13 adds the visit date / time to the field 16c of the record in which the candidate attention person determined to have been matched in S15 is registered, and then determines from the visit date / time registered in the field 16c. The number of visits in the period is newly calculated, and the number of visits in the field 16d is updated to the newly calculated number of visits.

  In S19, the face authentication collation unit 13 refers to “the number of visits in a predetermined period” of the record in which the history information is updated, and whether or not the updated number of visits of the attention candidate person in the predetermined period is a predetermined number or more. If it is not the predetermined number of times or more, the process returns to the process of S10 without identifying the candidate for attention as a person requiring attention, and if it is the predetermined number of times or more, the process proceeds to the process of S20.

  In S20, the face authentication collation unit 13 estimates that the candidate for attention (person in the captured image) of the record whose history information has been updated is a person requiring attention, and proceeds to the processing in S21.

In S <b> 21, the face authentication verification unit 13 displays “information indicating that a person requiring attention has arrived at the facility (for example, text information)” and “captured image of a person estimated to be a person requiring attention” on the display unit 5. Is output. Thereby, “information indicating that a person requiring attention has arrived at the facility” and “a captured image of a person estimated to be a person requiring attention” are displayed on the display unit 5.
Further, the face authentication verification unit 13 sends “information indicating that a person requiring attention is present at the facility”, “captured image of the person requiring attention”, and “facility information” to the information providing server 3 via the communication processing unit 12. "Is transmitted (output), and the process returns to S10. In the present embodiment, the face authentication verification unit 13 includes the location information of the camera C in the “facility information” (when the facility is a school, installation location information of the main gate, the back gate, etc.) and transmits it to the information providing server 3. To do.

  As described above, in the crime prevention system according to the present embodiment, when a person other than an entrance-permitted person visits a facility for the first time, the person is determined as a candidate for attention and is associated with the image of the attention candidate. History information including the visit date and time of the candidate person's facility is registered. In addition, according to the crime prevention system of this embodiment, every time a candidate for attention visits the facility, the history information of the candidate for attention is updated, and when the updated history information satisfies a predetermined requirement, the candidate for attention is a person who needs attention. It is estimated. That is, in this embodiment, the crime prevention system automatically estimates a person with an unnatural visit history as a person requiring attention from among persons coming to the facility. Therefore, according to the present embodiment, for example, when there is a suspicious person who visits a facility many times for a preview of criminal acts or the like or visits a school many times for voyeurism It can be estimated that a suspicious person is a person requiring attention, and effective crime prevention can be expected by monitoring the person requiring attention.

  Next, processing performed by the information providing server 3 and processing performed by the smartphone SP will be described.

  As illustrated in FIG. 7, the information providing unit 31 of the information providing server 3 transmits “information indicating that a person requiring attention”, “a captured image of a person requiring attention” transmitted from the attention person estimation device 1, and When “facility information” is received (S30), the smartphone SP associated with the received facility information is identified using the received information and the terminal database 35, and a push notification is sent to the identified smartphone SP. Through the server 7, “notice that a person requiring attention has arrived at the facility” and “access information to the person-of-care of information providing site” are notified by push (S 31).

  Specifically, in S31, the information providing unit 31 refers to the terminal database 35, selects a record in which the “facility information” received in S30 is registered, and extracts a device token registered in the selected record. To do. Further, the information providing unit 31 uses the information received in S <b> 30 to generate cautionary person information providing site information including “captured images of cautionable persons”. In addition, the information providing unit 31 includes push notification information to be presented by the push notification (“notice that a person requiring attention is present at the facility” and “access information to the person requiring attention information providing site 120”) and a device token. An included notification package is generated, and the notification package is transmitted to the push notification server 7.

  When the push notification server 7 receives the notification package, the push notification server 7 “provides a person to be watched at the facility” and “provides attention to the person to be watched” for the smartphone SP specified by the device token included in the notification package. Information on access to site 120 "is transmitted.

  When the information providing unit 31 transmits the notification package to the push notification server 7, the information providing unit 31 then associates the “facility information”, “required” with the current date (year, month, day, time) in the attention person database 36. The “captured image of the caution person” is registered (S32). This current date and time is obtained from a clock (not shown) built in the information providing server 3.

Next, in S31, via the push notification server 7, the smartphone SP to which “information that requires attention” and “access information to the information site requiring attention” are sent from the information providing unit 31. Will be described.
Specifically, when the smartphone SP receives from the push notification server 7 “noticeable person is coming to the facility” and “access information to the person-of-care information providing site 120”, the received information is sent to the smartphone SP. Using, the notice person notice screen 110 shown in FIG. 9A is displayed on the display unit of the smartphone SP (S40). Note that the smartphone SP vibrates or sounds when it receives information from the push notification server 7.

Note that, in the illustrated notice person screen 110 requiring attention, an area 111 in which information indicating “a noticeable person is present” is presented to the facility, and a view button for accessing the notice person information providing site 120 is displayed. 112.
With this configuration, every time a person requiring attention visits the facility, the smartphone manager SP in charge of the facility is notified in real time that the person who needs attention is coming to the facility.

Then, when the person in charge of management who views the notice-noticeable person notification screen 110 displayed on the smartphone SP presses the View button 112 provided on the notice-noticeable person notification screen 110, the smartphone SP is connected to the information providing server 3. The information providing unit 31 is accessed, and a transmission request for the person-of-interest information providing site is requested (S41).
When the information providing unit 31 of the information providing server 3 receives a transmission request from the smartphone SP for the caution person information providing site 120, the information providing unit 31 transmits the requested person information providing site information generated in S31 to the requested smartphone SP. Transmit (S33).

  Next, the smartphone SP receives the cautionary person information providing site information from the information providing unit 31, and uses the received cautionary person information providing site information to display the cautionary person information providing site 120 (FIG. 9). (B) is displayed (S42).

  As shown in the drawing, in the person-of-interest information providing site 120, an area 121 for displaying a captured image of the person who needs attention, the number of visits of the person who needs attention to the facility, and position information of the captured image ( For example, an area 122 for presenting location information such as a main gate of a school and a parking lot of a hospital is provided. In addition, the person information providing site 120 requiring attention is provided with a button 123 for accepting the completion of the correspondence.

With the above configuration, every time a person requiring attention visits the facility, the person in charge of the facility (guard, etc.) is notified that the person requiring attention is coming to the facility, and the captured image of the person requiring attention is displayed. Be presented. In addition, the location information of the facility where the person in need of attention came to the manager.
That is, according to the present embodiment, the person in charge of the facility (guard, etc.) can visually check the face of the person requiring attention when the person requiring attention comes to the facility. Can be easily grasped. In addition, in a facility where there are a plurality of managers (guards, etc.), a plurality of managers can simultaneously see and grasp the images of people who need attention, so an effective crime prevention system is established. In addition, since the location information of the facility where the person requiring attention has been transmitted to the person in charge of management, the person who needs attention can be found early.

In the crime prevention system of the present embodiment, the estimated captured image of the person who needs attention is automatically stored in the information providing server 3 (S32 in FIG. 7).
Therefore, for example, in the case where an incident (an incident such as theft, voyeurism, arson) caused by a suspicious person occurs in a facility other than the facility where the camera C of the crime prevention system according to the present embodiment is installed, the information is stored in the information providing server 3. The images that are displayed can be an important document for police investigation and can help resolve the incident.

  In addition, this invention is not limited to embodiment mentioned above, A various change is possible within the range of the summary.

  For example, in the above configuration, the attention person estimation apparatus 1 may be connected to a lamp (for example, an LED lamp) and a speaker. In this case, the caution person estimation device 1 turns on (or blinks) a warning lamp and generates “voice information indicating that a person requiring attention is present” when it is estimated that a person requiring attention has arrived. The audio information may be output from a speaker.

  In the above-described embodiment, only the number of visits for one predetermined period is registered in the field 16c of the unregistered person database 16, but the number of visits for a plurality of periods can be registered. A threshold value may be determined to determine whether or not the candidate for attention is a person requiring attention.

In the above-described embodiment, even if a person other than the admission permitted person visits the facility, if it is not estimated that the person needs attention, alert processing (display on the display unit 5 or display on the smartphone SP) is performed. Push notification) is not performed, but the present invention is not limited to this. When it is determined by the first face authentication collation process that the person who has come to the facility is a person other than the person permitted to enter, the fact that a person other than the person permitted to enter is displayed on the display unit 5 or pushed to the smartphone SP. You may make it notify.
Further, in the above case, when a plurality of persons come into the imaging range of the camera C and the plurality of persons are imaged together by the camera C, even one of the plurality of persons is permitted to enter. If so, the alert process may not be performed.

  In the process of FIG. 6 described above, the sex and the age are estimated from the acquired captured image in S12, but the present invention is not particularly limited thereto. In the processing step shown in FIG. 6, without performing the process of S12, the image data of all the records in the registered person database 15 is subjected to the first face authentication collation process in S13, and all the records in the unregistered person database 16 are processed in S15. The second face authentication collation process may be performed on the image data.

Further, in the above-described embodiment, the information providing server 3 notifies the smartphone SP held by the facility manager of the notification that “a person requiring attention is coming” via the push notification server 7. However, it is not particularly limited to this. The information providing server 3 is composed of a device integrated with the push notification server 7, and the information providing server 3 directly notifies the smartphone SP that “a person requiring attention is coming”. Also good.
Further, the notification that “a person requiring attention is coming” may be sent by e-mail, for example, instead of the push notification. In this case, the information providing server 3 is an electronic device in which “noticeable person is coming” and “access information to the sensitive person information providing site (site address)” are written to the mail address of the smartphone SP. send an email.

  Moreover, the structure of the information provision server 3 of embodiment mentioned above is only an example. The function of the information providing server 3 may be provided in the cloud.

  In the above-described embodiment, the information distribution server 3 provides information such as a noticed person to the smartphone SP of the facility manager, but the smartphone SP is only an example of an information terminal. Absent. The present invention is applied to any information processing terminal (notebook personal computer, tabred type terminal) having a communication function connectable to the network NW.

  In the above-described embodiment, a device token is used as terminal identification information for specifying the smartphone SP (terminal). However, the device token is only an example of terminal identification information. For example, the administrator of the security system may assign unique terminal identification information to each smartphone SP and register the unique terminal identification information in the information providing server 3.

  In the terminal database 35 of the above-described embodiment, the facility information is registered in association with the user ID, password, and terminal identification information (device token), but the user ID and password may not be present (that is, the terminal The facility information only needs to be registered in the database 35 in association with the terminal identification information).

  Further, in the above-described embodiment, when a person other than the admission permitted person has arrived at the facility more than a predetermined number of times, the person is estimated as a person requiring attention. However, the present invention is not limited to this. For example, as history information of the unregistered person database 16, in addition to “visit date and time” and “number of visits”, position information of the camera that captured the candidate for attention (for example, information on school gates, back gates, bicycle parking lots, etc.) In this way, if a candidate for attention is in a plurality of positions within one day, the candidate for attention may be estimated to be a person requiring attention.

C ... Camera SP ... Smartphone

DESCRIPTION OF SYMBOLS 1 ... Attention person estimation apparatus 10 ... Control part 11 ... Image processing part 12 ... Communication processing part 13 ... Face authentication collation part 15 ... Registered person database 16 ... Unattended person estimation apparatus 20 ... Crime prevention system cooperation application 5 ... Display part 6 ... Input section

3 ... Server device 30 ... Control unit 31 ... Information providing unit 35 ... Terminal database 36 ... Attention person database

7 ... Push notification server

Claims (4)

A security system comprising a caution person estimation device that estimates whether a person captured by a camera installed in a facility is a cautionary person,
The attention person estimation device includes:
A registered person database that stores face images of persons allowed to enter the facility;
Face authentication is performed for acquiring a captured image of a person's face captured by the camera, and performing a first face authentication verification process for verifying the acquired captured image and a face image of an entry-permitted person stored in the registered person database. A collation unit;
The person of the captured image that is determined not to match the face image of the admission-permitted person by the first face authentication collation process is set as a candidate for attention, and the visit history information of the facility for the candidate for attention is associated with the captured image. And an unregistered person database for storing
The face authentication verification unit
When it is determined by the first face authentication collation process that the captured image does not match the face image of the person permitted to enter, the acquired captured image and the image of the attention candidate person stored in the unregistered person database are used. When the second face authentication collation process to be collated is performed and it is determined by the second face authentication collation process that the captured image matches the image of the attention candidate person, the matched attention stored in the unregistered person database The visit history information associated with the image of the candidate person is updated, and when the updated visit history information satisfies a predetermined requirement, the person of the acquired captured image is estimated to be a person requiring attention, and the person concerned needs attention A crime prevention system characterized by outputting information indicating that a person is in the facility.   The face authentication verification unit determines a person of the captured image as a candidate for attention when it is determined by the second face authentication verification process that the captured image does not match the image of a candidate for attention stored in the unregistered person database. The crime prevention system according to claim 1, wherein a visitor history information of the attention candidate person is associated with the captured image and stored in an unregistered person database.   The crime prevention system according to claim 1, wherein the predetermined requirement is the number of visits to the facility within a predetermined period. The attention person estimation device is connected to a server device via a network,
When the face authentication verification unit estimates that the person in the captured image is a person requiring attention, the face authentication collation unit transmits a captured image of the person estimated as the person in need of attention to the server device,
When the server device receives the captured image transmitted by the attention person estimation device, the server device notifies a pre-registered information terminal that the person requiring attention is in the facility via the network. The crime prevention system according to any one of claims 1 to 3, wherein information on a person requiring attention including a captured image transmitted from the attention person estimation device is provided.

Images