JP6440327B2 - Crime prevention system, crime prevention method, and robot - Google Patents

Description

  The present invention relates to a crime prevention system, a crime prevention method, and a robot for preventing shoplifting in a store.

  Shoplifting damage is having a major impact on management at retail stores such as bookstores and supermarkets, and the annual shoplifting damage in Japan is said to be 1 trillion yen. In order to reduce shoplifting damage, security cameras have been installed, security guards have been placed, and store staff has been increased. In recent years, there has been proposed a crime prevention system that detects a person to be alerted by performing face authentication collation on a facial image of a person in a store taken by a security camera (for example, Patent Document 1).

Patent No. 5747116

  However, in the security system using a security camera that has been proposed in the past, the installation position of the security camera is fixed, so the store is photographed covering the entire store area due to restrictions on the installation position and number of installations and the blind spot of the camera. It is difficult.

  Accordingly, an object of the present invention made in view of the above problems is to provide a crime prevention system, a crime prevention method, and a robot capable of photographing a visitor regardless of the position of the visitor. is there.

In order to solve the above problems, a security system according to an embodiment of the present invention is:
A security system comprising at least one robot that moves autonomously, a server, and a communication terminal,
The robot is
A vehicle that can move autonomously to approach the store visitor;
A shooting section for shooting the video of the store visitor;
A communication unit for transmitting the video to the server;
With
The server
A server communication unit that receives the video from the robot;
A server control unit for determining whether the store visitor is a person to be alerted based on the video;
With
The server control unit transmits an alarm to the communication terminal by the server communication unit when the store visitor determines that the customer is an attention target person.

In addition, the security system according to an embodiment of the present invention,
The server is
It has a person authentication database that stores face images related to the target person,
The server control unit extracts a face image of the store visitor from the video, and determines whether the store visitor is a person to be alerted by face authentication based on the person authentication database.

In addition, the security system according to an embodiment of the present invention,
The person authentication database is encrypted.

In addition, the security system according to an embodiment of the present invention,
The server control unit extracts the visitor's behavior pattern from the video, and determines whether or not the visitor is an attention target person based on the behavior pattern.

In addition, the security system according to an embodiment of the present invention,
The robot further includes:
A voice dialogue function unit for voice dialogue with the store visitor,
The communication unit transmits the voice of the customer collected by the voice interaction function unit to the server;
The server communication unit receives the voice from the robot,
The server control unit determines whether or not the store visitor is an attention target person based on the video and the audio.

In addition, the security system according to an embodiment of the present invention,
The server control unit determines a warning level for the attention target person based on the video, and the warning includes information relating to the warning level.

In addition, the security system according to an embodiment of the present invention,
When the robot receives the tracking instruction, the robot tracks the store visitor by the moving means,
The image of the store visitor is photographed by the photographing unit.

The crime prevention system according to an embodiment of the present invention further includes a fixed camera provided in the store,
The fixed camera transmits the captured second video to the server,
The server control unit determines whether or not the store visitor is an attention target person based on the second video. If the store control unit determines that the store visitor is an attention target person, the server communication unit The tracking instruction is transmitted to the robot.

The crime prevention system according to an embodiment of the present invention includes a plurality of robots,
The plurality of robots communicate via the server and autonomously move so that the distance between the robots becomes a predetermined distance or more.

The crime prevention system according to an embodiment of the present invention further includes a remote monitoring terminal,
The server transmits the video to the remote monitoring terminal.

In addition, the crime prevention method according to an embodiment of the present invention,
A security method by a security system comprising an autonomously moving robot, a server, and a communication terminal,
The robot
A moving step for autonomously moving to approach the store visitor according to the location of the store visitor;
Photographing a video of the store visitor by a photographing unit provided in the robot;
Transmitting the video to the server;
The server is
Receiving the video from the robot;
Determining whether the store visitor is a person to be alerted based on the video;
If it is determined that the store visitor is a person to be alerted, sending an alarm to the communication terminal;
including.

Moreover, the robot according to one embodiment of the present invention is:
A vehicle that can move autonomously to approach the store visitor;
A shooting section for shooting the video of the store visitor;
A communication unit for transmitting the video to a server;
With
When the tracking instruction is received from the server, the store visitor is tracked by the moving means.

  According to the crime prevention system, the crime prevention method, and the robot of the present invention, the customer can be photographed regardless of the location of the customer.

1 is a block diagram of a crime prevention system according to Embodiment 1. FIG. It is a block diagram of the robot of FIG. It is a block diagram of the server of FIG. It is a figure which shows an example of the person authentication database memorize | stored in the server of FIG. It is a block diagram of the communication terminal of FIG. It is a figure which shows an example of the alarm screen displayed by the communication terminal of FIG. It is a figure which shows another example of the alarm screen displayed by the communication terminal of FIG. It is a figure which shows a mode that the robot is arrange | positioned in the predetermined position in a shop. It is a figure which shows a mode that the robot moved from the predetermined position in a shop. 3 is a flowchart illustrating an operation of the crime prevention system according to the first embodiment. It is a figure which shows an example of a store visitor record database. It is a figure which shows a mode in the case of providing a fixed camera. It is a figure which shows a mode that the some robot is arrange | positioned in the shop. It is a block diagram in the case of providing a remote monitoring terminal. It is a block diagram of the remote monitoring terminal of FIG.

  Embodiments of the present invention will be described below.

(Embodiment 1)
FIG. 1 is a schematic diagram of a security system 10 according to an embodiment of the present invention. The security system 10 is used in a store such as a bookstore or a convenience store. Hereinafter, a case where the security system 10 is used in a bookstore will be described. A security system 10 according to an embodiment of the present invention includes at least one robot 1 that moves autonomously, a server 2, and a communication terminal 3. The robot 1 and the server 2 are connected by a network N such as the Internet. In FIG. 1, only one robot 1 is shown, but the present invention is not limited to this, and the number of robots may be two or more. In FIG. 1, five servers 2 are shown. However, the number of servers 2 is not limited to this, and an arbitrary number of servers 2 may be provided or one. 1 shows one communication terminal 3, the number of communication terminals 3 is not limited to this, and may be two or more.

  As an outline, in the crime prevention system 10 according to the present embodiment, the robot 1 always shoots an image in the bookstore, and the robot 1 moves autonomously to approach the store visitor and shoot the store visitor image at a short distance. In this way, the image captured by the robot 1 is transmitted to the server 2. The server 2 is an information processing device such as a server device managed by an operator of the crime prevention system 10, for example, and determines whether or not the store visitor is a person to be alerted based on the video received from the robot 1. Here, the attention target person is a target person who needs attention, and includes shoplifting addicts, suspicious persons, dangerous persons, and the like. The server 2 transmits an alarm to the communication terminal 3 when it is determined that the store visitor is an attention target person. The communication terminal 3 is an information processing apparatus such as a personal computer or a tablet terminal used by a store employee (store clerk), and receives an alarm from the server 2. In this way, the robot 1, the server 2, and the communication terminal 3 cooperate to realize the crime prevention system 10. Next, each component of the crime prevention system 10 will be specifically described.

(Configuration of robot 1)
First, the robot 1 will be described. The robot 1 is preferably a humanoid robot and includes a head including a face, a torso, and an arm. FIG. 2 shows a block diagram of the robot 1. The robot 1 includes a communication unit 11, a moving unit 12, a photographing unit 13, a voice interaction function unit 14, a display unit 15, a storage unit 16, and a control unit 17. The communication unit 11, the moving unit 12, the photographing unit 13, the voice interaction function unit 14, the display unit 15, the storage unit 16, and the control unit 17 are connected by a bus 18.

  The communication unit 11 is an interface that communicates with an external device wirelessly or by wire and transmits and receives information. In the present embodiment, the communication unit 11 can transmit and receive information to and from the server 2 via the network N. The communication unit 11 can transmit and receive information to and from the communication terminal 3 via the network N. The communication unit 11 appropriately performs solid authentication processing (login processing or the like) of the robot 1 when communicating with the server 2. Thus, when there are a plurality of robots 1, the server 2 grasps which robot 1 is accessing from. If the login is not regular, access to the server 2 is appropriately blocked.

  The moving means 12 moves and rotates the main body of the robot 1 in all directions (front and rear, left and right), and is, for example, an omni wheel. The omni wheel is a tire that can move in all directions, and has a plurality of rotatable rollers. By controlling the rotation of the roller, it is possible to move in any direction, and it is also possible to rotate. Since the robot 1 is used in a store, it is preferable to move slowly so as not to collide with a store visitor or the like. For example, the maximum value of the moving speed by the moving means 12 is 2 km / h. The moving means 12 is not limited to the omni wheel, and any moving means can be adopted.

  The photographing unit 13 includes at least one 2D camera for photographing a store visitor, and photographs a video with a predetermined screen resolution and frame rate. The screen resolution is, for example, full high vision (1920 × 1080 pixels). The frame rate is 30 fps, for example. Preferably, the number of 2D cameras is two, and they are provided at different positions (for example, the position of the forehead and the mouth) of the face of the robot 1. When 2D cameras are placed at different positions and one of the 2D cameras cannot capture the store visitor (for example, there is a shield in front of the forehead, and part or all of the shooting range is blocked by the shield) The customer can be photographed with the other 2D camera. The robot 1 detects the presence of the store visitor based on the video imaged by the 2D camera. Specifically, the in-store video is constantly taken by the 2D camera, and the taken video is transmitted to the server 2 by the communication unit 11. The video may be a still image or a moving image. When the server 2 detects that the store visitor is included in the captured video, the server 2 transmits a movement instruction to the robot 1. When the robot 1 receives a movement instruction via the communication unit 11, the robot 1 moves autonomously so as to approach the store visitor by the moving unit 12. Preferably, the photographing unit 13 transmits the compressed video to the server 2. For example, the vertical and horizontal directions of the video are compressed to ¼, and the output video is 480 × 270 pixels. Also, the frame rate is compressed to ½ to 15 fps. As a result, it is possible to improve the efficiency of communication resources related to video transmission and the storage capacity of the server 2.

  Furthermore, the imaging unit 13 includes at least one 3D camera for distance measurement. The 3D camera can acquire the distance to the object by irradiating the infrared line of sight and detecting the reflected light. The 3D camera acquires 3D video (video including distance information) at a predetermined screen resolution (for example, 320 × 240 pixels) and a predetermined frame rate (for example, 20 fps). Preferably, the number of 3D cameras is two, and they are provided at different positions (for example, positions of both eyes) of the face of the robot 1. By placing 3D cameras at different positions, if one 3D camera cannot capture a visitor (for example, if there is a shield in front of the right eye and the shooting range is blocked by the shield), the other Visitors can be photographed by a 3D camera (a 3D camera provided at the position of the left eye). The robot 1 detects the distance to the object using a 3D camera, and moves autonomously by the moving unit 12 based on the detected distance.

  The voice dialogue function unit 14 has a function of voice dialogue with the store visitor. The microphone collects the voice of the visitor's conversation, the voice analysis unit analyzes the collected voice content, and the analyzed voice content. A response control unit that controls appropriate response content based on the response content and a speaker for speaking the response content and interacting with the store visitor are included. The voice dialogue function unit 14 can perform voice dialogue based on a voice dialogue knowledge database built in the robot 1 or stored in the server 2 and can learn dialogue. The voice interaction function unit 14 enables the robot 1 to serve customers in stores. The robot 1 can make a call to the store visitor by using the voice dialogue function unit 14 such as “Come to me” or “Is looking for something?”. Further, according to the voice of the store visitor's conversation, for example, a response regarding a product (book) that the store visitor is looking for can be made.

  The display unit 15 is a display device such as a liquid crystal display or an organic EL display, and can display various screens. Preferably, the display unit 15 is a touch panel and accepts an input operation of a store visitor. Hereinafter, the display unit 15 will be described as a touch panel. When the voice interaction function unit 14 detects that the book that the store visitor is searching for is detected, the display unit 15 displays a menu screen of a book search application. Based on a touch operation (input operation) by a visitor, the display unit 15 displays a list of corresponding books by book name, author name, genre, keyword, and the like. When a touch operation (selection operation) on any book in the list is detected, the display unit 15 displays information related to the book (book name, author name, book shelf position, inventory number, etc.). The robot 1 executes the search application by communicating with the server 2 (or a book database server) as appropriate. In addition, you may make it detect the book which the store visitor is searching for by the voice interaction function part 14, In this case, the touch operation by the store visitor can be omitted.

  The storage unit 16 includes, for example, a primary storage device and a secondary storage device, and stores various information provided from the server 2 and programs necessary for information processing. For example, the storage unit 16 stores a robot application (hereinafter referred to as a robot application) of the security system 10. The robot application can be acquired from a predetermined distribution server via the network N, for example. The operation according to this embodiment of the robot 1 is realized in a state where the robot application is executed (activated).

  The control unit 17 includes a dedicated or general-purpose processor. The control unit 17 controls the operation of the entire robot 1 using the robot application. For example, the control unit 17 transmits and receives information through the communication unit 11. Specifically, for example, the control unit 17 transmits the video imaged by the imaging unit 13 to the server 2 by the communication unit 11. Further, the control unit 17 transmits the voice of the store visitor collected by the voice interaction function unit 14 to the server 2. Further, when the store 17 is detected (when a movement instruction is received from the server 2), the control unit 17 autonomously moves so as to approach the store by the moving unit 12. Further, the control unit 17 approaches the store visitor by the moving unit 12 and photographs the store visitor by the image capturing unit 13. That is, the robot 1 photographs the store visitor at a closer distance. By doing so, it is possible to photograph the face and the hand of the visitor more clearly.

  Further, when receiving a tracking instruction from the server 2 or the communication terminal 3, the control unit 17 tracks the store visitor by the moving unit 12, and captures an image of the store visitor by the imaging unit 13 (continues to capture the image). The tracking instruction is an instruction to perform tracking for a specific store visitor, and includes information related to the store target to be tracked (for example, a face image of the store visitor and other information for identifying the store visitor). Moreover, the control part 17 stops the tracking to a store visitor, when it is a case where the tracking unnecessary instruction | indication is received from the server 2 or the communication terminal 3, and it is tracking. When the tracking unnecessary instruction is received from the server 2 or the communication terminal 3 and the robot 1 is not tracking the customer, the customer is not tracked.

(Configuration of server 2)
FIG. 3 is a block diagram of a server according to an embodiment of the present invention. The server 2 includes a server communication unit 21, a server storage unit 22, and a server control unit 23. The server communication unit 21, the server storage unit 22, and the server control unit 23 are connected to each other by a bus 24.

  The server communication unit 21 is an interface that communicates with an external device wirelessly or by wire and transmits and receives information. In the present embodiment, the server communication unit 21 can transmit and receive information to and from the robot 1 via the network N. The server communication unit 21 can transmit and receive information to and from the communication terminal 3 via the network N.

  The server storage unit 22 includes, for example, a primary storage device and a secondary storage device, and stores various information and programs necessary for providing and controlling the crime prevention system 10.

  The server control unit 23 includes a dedicated or general-purpose processor and performs various processes. The server control unit 23 controls the operation of the entire server 2. For example, the server control unit 23 transmits and receives information using the server communication unit 21. Specifically, for example, the server control unit 23 receives video and audio transmitted from the robot 1 by the server communication unit 21. The server control unit 23 stores the received information (video and audio) in the server storage unit 22. The stored information is stored for a certain period after being stored, and then deleted. The certain period is, for example, one month. The amount of data stored in one month is about 1 TB when storing images from one robot 1 (10 hours / day, 30 days operation). The server storage unit 22 has a necessary storage capacity.

  Further, the server control unit 23 determines whether or not the store visitor is an attention target person based on at least one of the received video or audio. In the following, the attention target person determination process will be described in detail.

  When determining whether or not the store visitor is a person to be alerted based on the video received by the server control unit 23, the store visitor's face image included in the video is extracted and the store visitor is subjected to face authentication based on the person authentication database 221. It is determined whether or not is an attention target person. Artificial intelligence technology is used for face authentication as appropriate, and machine learning is used to improve the accuracy of authentication. The person authentication database is a database that stores information related to the attention target person, and is stored in the server storage unit 22. FIG. 4 shows an example of the person authentication database 221. The person authentication database 221 illustrated in FIG. 4 includes an attention target person ID, a name, a face image, sex, age, physical characteristics, and the like. The attention target person ID is an identifier for uniquely identifying the attention target person. The person authentication database 221 stores the name, face image, gender, age, physical characteristics, and the like of the attention target person associated with the ID in association with the attention target person ID. The server control unit 23 accesses the server storage unit 22 and searches for whether or not there is information matching or similar to the face image of the store visitor included in the video received from the robot 1 to perform face authentication. If information matching or similar to the face image of the visitor is stored in the person authentication database 221, it is determined that the visitor is the person to be alerted. In the face authentication, in addition to the face image, at least one information of gender, age, and physical characteristics may be appropriately combined to improve the accuracy of face authentication.

  Preferably, the person authentication database 221 is encrypted. By encrypting, even if information leaks to the outside, information such as face images and physical characteristics can be substantially unrecoverable.

In addition, the server control unit 23 may extract a visitor's behavior pattern from the received video, and determine whether or not the visitor is an attention target person based on the extracted behavior pattern. Examples of the behavior pattern include the following.
・ I have a large bag (eco bag, tote bag, etc.).
・ Squatting.
・ Wears a hat, glasses and mask.
・ It is solidified as a group.
・ Continues possession of the product for a predetermined time.
・ Gaze moves frequently.
When the visitor has taken these behavior patterns, the server control unit 23 determines that the visitor is a person to be alerted.

  Further, the server control unit 23 may determine whether or not the store visitor is an attention target person from the received voice. Specifically, the server control unit 23 determines the degree of tension of the store visitor from the received voice, and determines that the store visitor is an attention target person when the degree of tension is equal to or greater than a predetermined value.

  When the store visitor determines that the customer is a person to be alerted by the method described above, the server control unit 23 transmits an alarm to the communication terminal 3 by the server communication unit 21.

  Here, the server control unit 23 may determine a warning level for the attention target person. The alert level is an index that represents the degree of attention to the attention target person, and is determined by, for example, three levels (low, medium, and high). For example, when the visitor is taking an action corresponding to any one of the above action patterns, the alert level is set to “low”. Further, for example, when a visitor takes an action corresponding to any one of the above action patterns and the degree of tension based on the voice is equal to or higher than a predetermined value, the alert level is set to “medium”. For example, when information that matches or is similar to the face image of the visitor is stored in the person authentication database 221, the alert level is set to “high”. The method for setting the alert level is not limited to this. For example, points may be set for the degree of coincidence of face authentication and the action pattern, and the alert level may be set based on the total value of points. The server control unit 23 includes information related to the alert level in the alarm, and transmits the alarm to the communication terminal 3 by the server communication unit 21.

  Further, the server control unit 23 may transmit a tracking instruction to the robot 1 via the server communication unit 21 according to the alert level. For example, when the alert level is “high”, a tracking instruction for the visitor may be transmitted to the robot 1 via the server communication unit 21. As a result, when the alert level is equal to or higher than the predetermined level, the robot 1 can be immediately tracked.

(Configuration of communication terminal 3)
FIG. 5 is a block diagram of the communication terminal 3 according to an embodiment of the present invention. The communication terminal 3 includes a terminal communication unit 31, a terminal storage unit 32, a terminal display unit 33, a terminal input unit 34, and a terminal control unit 35. The terminal communication unit 31, the terminal storage unit 32, the terminal display unit 33, the terminal input unit 34, and the terminal control unit 35 are connected to each other by a bus 36.

  The terminal communication unit 31 is an interface that communicates with an external device wirelessly or by wire and transmits and receives information. In the present embodiment, the terminal communication unit 31 can transmit and receive information to and from the server 2 via the network N.

  The terminal storage unit 32 includes, for example, a primary storage device and a secondary storage device, and stores various information provided from the server 2 and programs necessary for information processing. For example, the terminal storage unit 32 stores an application of the security system 10 (hereinafter referred to as a security application). The security application can be acquired from a predetermined distribution server via the network N, for example. The operation according to the present embodiment of the communication terminal 3 is realized in a state where the security application is executed (activated). Note that the communication terminal 3 may not include a crime prevention application, and the operation according to the present embodiment may be realized based on standard software such as an OS (Operating System) of the communication terminal 3.

  The terminal display unit 33 is a display device such as a liquid crystal display or an organic EL display, and can display various screens.

  The terminal input unit 34 is an input interface such as a keyboard, a pen tablet, a touch pad, or a mouse, and can accept an operation of an operator (a store clerk in the present embodiment) on the communication terminal 3. The terminal input unit 34 may be a touch panel provided integrally with the terminal display unit 33. Hereinafter, in the present embodiment, the terminal input unit 34 is described as a touch panel provided integrally with the terminal display unit 33.

  The terminal control unit 35 includes a dedicated or general-purpose processor. The terminal control unit 35 controls the operation of the entire communication terminal 3 using a security application. For example, the terminal control unit 35 transmits and receives information using the terminal communication unit 31. Specifically, for example, the terminal control unit 35 receives an alarm from the server 2. Further, the terminal control unit 35 displays an alarm screen on the terminal display unit 33 based on the received alarm.

  FIG. 6 shows an example of an alarm screen displayed by the terminal display unit 33. The alarm screen illustrated in FIG. 6 is an example of an alarm screen when the alert level is “high”, and includes that the person to be alerted has been detected and the alert level. The alarm screen illustrated in FIG. 6 includes information related to the attention target person and real-time information. The real-time information is real-time shooting information shot by the robot 1. The shooting information may be a moving image or a still image. The real-time information includes information related to the shooting date and time of the video and the shooting location. In FIG. 6, the book shelf name (book shelf A) is shown as information relating to the shooting location, but the present invention is not limited to this, and arbitrary information indicating the position can be used. The alarm screen includes an input button 341 and a home button 342 that can be touch-operated. The input button 341 is a button for transmitting a tracking unnecessary instruction. When an operation on the input button 341 is detected, the terminal control unit 35 transmits a tracking unnecessary instruction to the robot 1 via the server 2 by the terminal communication unit 31. The home button 342 is a button for returning to the home screen. When an operation on the home button 342 is detected, the screen returns to the home screen. When the alert level is “high”, even if an operation on the home button 342 is detected, the home screen may not be returned.

  FIG. 7 shows another example of the alarm screen displayed by the terminal display unit 33. The alarm screen illustrated in FIG. 7 is an example of an alarm screen in a case where the alert level is “medium”, and includes that the target person has been detected and the alert level. The alarm screen shown in FIG. 7 includes real-time information. The alarm screen shown in FIG. 7 includes an input button 343 in addition to the input button 341. The input button 343 is a button for transmitting a tracking instruction. When detecting an operation on the input button 343, the terminal control unit 35 transmits a tracking instruction to the robot 1 via the server 2 by the terminal communication unit 31.

  Information on the shooting location in FIGS. 6 and 7 is determined based on the video shot by the robot 1. Hereinafter, determination of information related to the shooting location will be described. First, the robot 1 is in principle located at a predetermined position in the bookstore (hereinafter referred to as a home position). 8 and 9 respectively show a case where the robot 1 is located at the home position and a case where the robot 1 has moved from the home position. As shown in FIG. 8, a cash register 4 and a plurality of book shelves 5 a to 5 h are arranged in the bookstore, and the robot 1 is usually located at a home position near the entrance 6. In FIG. 8, two shop assistants 7 a and 7 b are arranged at the cash register, and the robot 1 photographs the store visitor 8 who enters the store from the entrance 6. By setting the home position to the position shown in FIG. 8, it is possible to reliably photograph a customer coming from the entrance 6. In general, popular books and comics are arranged in a bookcase 5a near the entrance 6 where the store clerk can reach. That is, the vicinity of the bookshelf 5a is an area to be monitored with priority (hereinafter also referred to as a priority area). By placing the home position in the vicinity of the priority area, criminal acts such as shoplifting can be efficiently suppressed.

  In FIG. 9, the robot 1 tracks the customer 8 and moves away from the home position. When the robot 1 moves from the home position, the current location of the robot 1 is grasped based on a marker provided in the bookstore. The marker is, for example, a symbol such as an arrow or a two-dimensional barcode, and is printed or pasted on a bookcase or a floor. As shown in FIG. 9, a marker 9 is also provided at the home position. The robot 1 captures the marker by the imaging unit 13 and transmits an image including the marker to the server 2. The server control unit 23 of the server 2 determines the position (imaging location) of the robot 1 based on the marker included in the video. The server 2 includes information on the shooting location in the alarm. For example, when the robot 1 images the store visitor 8 at the position shown in FIG. 9 and the server 2 determines that the store visitor 8 is a subject of attention based on the video, the server 2 transmits an alarm to the communication terminal 3. The alarm screen displayed by the communication terminal 3 includes information on the bookshelf 5f as a shooting location. Since the information on the shooting location is included in the alarm, the store clerk can immediately check the site.

  Further, the server 2 transmits information on the detected shooting location to the robot 1. The robot 1 can return from the current position to the home position based on the received shooting location information. Thus, the marker provided in the store can provide information on the shooting location even in a place where it is difficult to use a GPS (Global Positioning System) function such as indoors.

  Next, the operation of the crime prevention system 10 according to the first embodiment of the present invention will be described with reference to the flowchart shown in FIG.

  First, the robot 1 detects a store visitor by the server 2 (step S10). Specifically, the robot 1 always shoots a video in the store by the shooting unit 13 and transmits the shot video to the server 2 by the communication unit 11. The server 2 detects the presence / absence of a visitor from the video. If a visitor is detected, the process proceeds to step S20.

  When a store visitor is detected in step S10, the robot 1 autonomously moves so as to approach the store visitor (step S20). Specifically, when the server 2 detects that a store visitor is included in the video, the server 2 transmits a movement instruction to the robot 1. When the robot 1 receives the movement instruction, the robot 1 autonomously moves so as to approach the store visitor by the moving means 12.

  Subsequently, the robot 1 photographs the store visitor by the photographing unit 13 (step S30). That is, the robot 1 photographs the store visitor at a closer distance. The robot 1 transmits the captured video to the server 2 through the communication unit 11 (step S40).

  Subsequently, the robot 1 serves a customer by the voice interaction function unit 14 (step S50). The robot 1 transmits the customer's voice collected by the voice interaction function unit 14 to the server 2 (step S60).

  Subsequently, the server 2 determines whether or not the store visitor is an attention target person based on at least one of the received video or audio (step S70). When it is determined that the visitor is the attention target person, the process proceeds to step S80. If it is determined that the visitor is not the attention target person, the process returns to step S10.

  When it is determined in step S70 that the store visitor is a person to be alerted, the server communication unit 21 of the server 2 transmits an alarm to the communication terminal 3 (step S80).

  Subsequently, when there is a tracking instruction (step S90: Yes), the robot 1 tracks the store visitor based on the tracking instruction (step S100). Specifically, when receiving a tracking instruction from the server 2 or the communication terminal 3, the control unit 17 of the robot 1 tracks the store visitor by the moving unit 12. Then, the robot 1 continues to photograph the store visitor by the photographing unit 13 (step S110). Preferably, the shooting is continued until the visitor leaves the store. Then, the process ends.

  As described above, according to the crime prevention system 10 according to the first embodiment, since the robot 1 moves autonomously and shoots a video of the customer, the customer can be photographed regardless of the location of the customer. Can do. In addition, the photographed video is transmitted to the server 2, and it is determined based on the video whether or not the store visitor is an attention target person. When it is determined that the store visitor is the person to be alerted, the server 2 transmits an alarm to the communication terminal 3. Therefore, the person to be alerted can be immediately identified, and a criminal act such as shoplifting can be suppressed by approaching and speaking with a clerk or the like.

  In particular, according to the crime prevention system 10 according to the first embodiment, the server 2 determines whether or not the store visitor is a person to be alerted based on the video imaged by the robot 1, and thus it is not always necessary to deploy security guards. . Guards to deter shoplifting require advanced skills including experience and intuition, and labor costs are high. Therefore, deploying security guards puts pressure on store management. In addition, the security guards are at risk of being attacked by criminals such as shoplifting addicts, and the actual situation is that there is a shortage of shoplifting security personnel. If the crime can be suppressed by using the robot 1 as in the crime prevention system 10 according to the first embodiment, the labor cost can be reduced and the human damage caused by the assault can be reduced, thereby eliminating the shortage of human resources. Can also contribute.

  In addition, the crime prevention system 10 according to the first embodiment includes the person authentication database 221 that stores the face image related to the attention target person, extracts the face image of the store visitor from the video taken by the robot 1, and the person authentication. It is determined whether or not the store visitor is an attention target person by face authentication based on the database 221. Therefore, shoplifting addicts and the like can be more reliably determined, and an alarm is transmitted to the communication terminal 3, so that criminal acts such as shoplifting can be suppressed.

  Further, according to the crime prevention system 10 according to the first embodiment, the person authentication database 221 is stored in the server 2. Therefore, it is not necessary to provide the robot 1 with a large-capacity storage unit. Further, addition, update, and deletion of records in the person authentication database may be performed by the server 2, and there is no need to change the setting of each robot 1. Also, even if the robot 1 is destroyed or stolen, there is no possibility that the person authentication database 221 will leak. Furthermore, since the person authentication database 221 is encrypted and stored in the server 2, there is no possibility that information in the person authentication database 221 is leaked from the server 2 even in the unlikely event.

  In this embodiment, information related to the attention target person is stored in the person authentication database 221, but a person other than the attention target person (for example, a store clerk or a normal visitor who purchases a regular product) is stored in the person authentication database 221. May be stored. In this case, when the person authentication database 221 stores information that matches or resembles the face image of the visitor included in the video captured by the robot 1, and the information is information on a person other than the person to be noted. , It is determined that the visitor is not a person to be noted. By doing so, it is possible to efficiently detect a person who is not the attention target person. In other words, when it is detected that the person is not a target person, the customer can be excluded from the target to be monitored. Note that a database (white list) relating to a person who is excluded from the attention target person may be stored in the server storage unit 22 as a database different from the person authentication database 221. Further, although not a subject of attention, a database (a list between the white list and the black list) related to the store to be noted may be stored in the server storage unit 22 to some extent. In this case, in addition to the information in the database, it is determined whether the visitor is a person to be alerted based on other information such as the visitor's behavior pattern and voice.

  In addition, according to the crime prevention system 10 according to the first embodiment, the behavior pattern of the store visitor is extracted from the video, and it is determined whether or not the store visitor is an attention target person based on the behavior pattern. Therefore, even if the store visitor is not registered in advance as the attention target person (for example, registered in the person authentication database 221), the attention target person can be appropriately detected. Thereby, the deterrent effect of criminal acts such as shoplifting can be enhanced.

  In the crime prevention system according to the first embodiment, the robot 1 includes the voice interaction function unit 14 so that the robot 1 can serve customers. Since it becomes difficult to conduct criminal acts such as shoplifting while being on the customer service, the deterrent effect of criminal acts such as shoplifting can be enhanced. Since security guards usually monitor shoplifting etc. while acting like a visitor in plain clothes, they usually do not speak to the store visitor. On the other hand, since the store clerk originally has a business, it is difficult to reside in a place where shoplifting is likely to occur. On the other hand, according to the robot 1 of Embodiment 1, both customer service and security can be performed simultaneously. Since customer service by robots has not yet become widespread, there is also a rarity, and it is possible to increase the effect of attracting customers to the store and increase sales by visitors who enjoy dialogue with the robot.

  Furthermore, since the voice of the store visitor collected by the voice interaction function unit 14 is also used to determine the target person to be watched, the store visitor is not registered in advance (for example, registered in the person authentication database 221) as the target target person. However, it is possible to detect the target person appropriately. Therefore, the deterrent effect of criminal acts such as shoplifting can be enhanced.

  In addition, according to the crime prevention system 10 according to the first embodiment, a warning level for a person to be alerted is determined based on an image captured by the robot 1, and the warning includes information related to the warning level. Therefore, the urgency level and importance level of the alarm displayed by the communication terminal 3 can be easily grasped, and appropriate measures can be taken without delay.

  Although FIG. 6 and FIG. 7 are shown as an example of the alarm, the display form of the alarm is not limited to this. For example, the server 2 may send an alarm to the communication terminal 3 in the form of an email. For example, an image of the person to be watched or a video taken by the robot 1 may be attached to the e-mail.

  In addition, according to the crime prevention system 10 according to the first embodiment, the robot 1 tracks the attention target person based on the tracking instruction and photographs the attention target person from a short distance. Can be increased. In addition, even if the subject of caution performs a criminal act such as shoplifting, an evidence video can be taken. The evidence video can be transmitted to the server 2 and the evidence can be recorded by the server 2.

  When the tracking instruction is received, at least one of the screen resolution and the frame rate of the video to be captured may be higher than before (when not received) the tracking instruction is received. By doing in this way, it is possible to clearly shoot the image of the attention target person to be noted. Conversely, in a state before receiving the tracking instruction, at least one of the screen resolution and the frame rate of the video to be taken may be lowered. By doing in this way, the amount of communication data from the robot 1 to the server 2 can be reduced, and communication resources can be used efficiently. In addition, the amount of data stored in the server 2 can be reduced.

  In the present embodiment, when a store visitor is detected in step S10 in FIG. 10, the robot 1 is controlled to move autonomously and approach the store visitor in step S20. However, the present invention is not limited to this. For example, only when the store visitor is detected and the possibility that the store visitor is the attention target is equal to or greater than the threshold value, the process proceeds to step S20, and the robot 1 may move autonomously and approach the store visitor. . The server 2 determines the possibility of being an attention target based on the video received from the robot 1. Specifically, for example, the server 2 may determine the possibility based on face authentication based on the person authentication database 221, or may determine the possibility based on a visitor's behavior pattern. As a result of the determination, for example, when the alert level is equal to or higher than a predetermined level, the store visitor determines that the possibility of the attention target person is equal to or higher than the threshold. In this way, for example, when a squatting person is found, the robot 1 moves autonomously to approach the customer 1 so as to approach, and shoots at a short distance. In addition, for example, when a person wearing a hat, glasses, and a mask is found, the robot 1 moves autonomously to approach the customer so as to approach, and photographs at a short distance. Or, for example, when customers are gathered together as a group, the robot 1 moves autonomously so as to approach the customer 1 to deal with customers, and shoots at a short distance. In this way, by approaching the store visitor when the possibility of the target audience is greater than or equal to the threshold, the video of the store visitor who is likely to be the target audience can be efficiently captured and criminal acts such as shoplifting are suppressed. can do.

  In addition, in a predetermined time zone (for example, at the time of opening and closing a store), the frequency at which the robot 1 approaches the customer and serves customers may be higher than that other than the predetermined time zone. For example, in a predetermined time period, the frequency of approaching customers and approaching customers is increased by lowering the threshold value. Alternatively, for example, when a visitor is detected during a predetermined time period, the customer may approach all of the store visitors detected by the robot 1 regardless of the possibility that the visitor is an attention target person. By doing so, criminal acts such as shoplifting can be suppressed even during times when the number of visitors is small and the store clerk is busy, such as at the time of opening and closing. In addition, according to time, you may change the frequency that the robot 1 approaches a store visitor. For example, the frequency may be increased around the time of shoplifting (November to January). On the other hand, the frequency may be lowered around the time when shoplifting is low (February, August).

  In the present embodiment, the degree of tension of the store visitor is determined from the received voice, and when the degree of tension is equal to or greater than a predetermined value, the store visitor is determined to be an attention target person, but is extracted from the received video. The mental state such as the degree of tension may be determined based on the movement of the shop visitor.

  The server 2 may store the visitor history in the server storage unit 22. In this case, the server 2 may use the history information for the determination of the attention target person. For example, the server 2 records the number of visits to the store and the store visit date and time as a visitor database based on the face image of the visitor received from the robot 1. FIG. 11 shows an example of the store visitor database. Shoplift addicts usually visit the store for a preview. Therefore, the number of visits may be recorded, and if the number of visits is greater than or equal to a predetermined number (for example, 2 times), it may be determined that there is a possibility of a person being warned. Alternatively, in addition to the visitor's history information, it may be determined that the person is an attention target person based on at least one of an action pattern and voice. By doing in this way, the attention object person who has visited in preview can be determined appropriately, and the effect of deterring crime can be enhanced. Further, the store visitor history information may be stored in the server storage unit 22 together with the purchase history. Then, the server control unit 23 of the server 2 may determine that the customer is not a person to be alerted when the customer has purchased the product properly. Further, for example, when the name of the store visitor can be acquired (for example, by the voice interaction function unit 14), the name of the store visitor may be stored as a history. Then, when the robot 1 receives a customer, the information is acquired from the server 2 and the name of the visitor (eg, “Taro”) is included in the call, and “Taro, what are you looking for?” You may call out. Or you may say, “You came to the store the other day. What are you looking for today?” Thereby, for example, if a visitor is an attention target person, the deterrent effect of criminal acts such as shoplifting can be enhanced. Further, even when the visitor is not a person to be alerted, the robot 1 remembers the visitor and attachment is generated, so that the effect of attracting customers to the store and an increase in sales can be achieved.

(Embodiment 2)
The second embodiment of the present invention will be described below. FIG. 12 is a block diagram showing a state where the security system 10b according to the second embodiment of the present invention is arranged in a store. The crime prevention system 10b according to the second embodiment is different from the configuration according to the first embodiment in that a fixed camera 131 is provided. In addition, since each component is the same as the structure which concerns on Embodiment 1, it attaches | subjects the same code | symbol and abbreviate | omits description.

  In the crime prevention system 10b according to Embodiment 2, the fixed camera 131 is provided at a position where the vicinity of the priority area can be photographed (for example, the ceiling near the priority area). A broken line extending from the fixed camera 131 in FIG. 12 indicates a shooting range of the fixed camera 131. The fixed camera 131 has a communication function, and transmits a captured video (referred to as a second video) to the server 2.

  Based on the second video, the server control unit 23 determines whether or not the store visitor is an attention target person. When it is determined that the visitor is the person to be alerted, the server communication unit 21 transmits a tracking instruction to the robot 1.

  As described above, according to the crime prevention system 10b according to the second embodiment, it is determined based on the second video captured by the fixed camera 131 whether or not the store visitor is a person to be alerted. Therefore, even if the robot 1 is moving from the home position 9, it is possible to capture a visitor entering the store appropriately and capture the video of the person to be watched more reliably.

  In addition, when the robot 1 exists in the home position 9, you may provide the imaging range change means to change the imaging range of the fixed camera 131. In this case, the fixed camera 131 changes the shooting range by the shooting range changing means so as to take a picture outside the range shot by the robot 1. By doing in this way, it is possible to enlarge the area for always photographing.

  In the above embodiment, the fixed camera 131 is installed at a position where the vicinity of the priority area can be photographed. However, the present invention is not limited to this, and the fixed camera 131 may be installed at an arbitrary position. For example, an area that is not near the home position may be provided at a position where photographing can be performed. In this case, the monitored area can be enlarged.

(Embodiment 3)
The third embodiment of the present invention will be described below. FIG. 13 is a block diagram showing a state where the security system 10c according to the third embodiment of the present invention is arranged in a store. The crime prevention system 10c according to the third embodiment is different from the configuration according to the first embodiment in that a plurality of robots are arranged. In addition, since each component is the same as the structure which concerns on Embodiment 1, it attaches | subjects the same code | symbol and abbreviate | omits description.

  As shown in FIG. 13, the crime prevention system 10 c according to the third embodiment includes a robot 1 c in addition to the robot 1. The configuration of the robot 1c is the same as that of the robot 1. The robot 1 c can communicate with the robot 1 via the network N and the server 2.

  For example, by communication between the robots, the robot 1 and the robot 1c autonomously move so that the distance between the robots becomes a predetermined distance or more. By doing in this way, it can prevent that the imaging range of each robot overlaps, and can expand the area which always image | photographs.

  When there are two robots, two home positions (9 and 9c) are provided according to the number of robots as shown in FIG. A priority may be set for the plurality of home positions. In the example shown in FIG. 13, the priority of the home position 9 is higher than the priority of the home position 9c. Therefore, when one robot (robot 1 in FIG. 13) leaves the home position 9, the other robot (robot 1c in FIG. 13) autonomously moves to the home position 9 with higher priority. By doing in this way, the probability that any robot exists in a home position with high priority can be raised. Thereby, the store visitor who enters the store appropriately can be photographed, and the image of the attention target person can be captured more reliably. In addition, since the robot at the home position 9 is more likely to detect a store visitor and receive a customer more frequently, the processing load between the robots can be leveled. Note that the absence of a robot at a home position with high priority may be determined based on an image captured by the imaging unit 13 of the robot, or may be determined based on the current value of each robot.

(Embodiment 4)
The fourth embodiment of the present invention will be described below. FIG. 14 is a block diagram of a security system 10d according to the fourth embodiment of the present invention. The crime prevention system 10d according to the fourth embodiment is different from the configuration according to the first embodiment in that a remote monitoring terminal 3d is provided. In addition, since each component is the same as the structure which concerns on Embodiment 1, it attaches | subjects the same code | symbol and abbreviate | omits description.

  FIG. 15 is a block diagram of the remote monitoring terminal 3d. The remote monitoring terminal 3d includes a terminal communication unit 31d, a terminal storage unit 32d, a terminal display unit 33d, a terminal input unit 34d, and a terminal control unit 35d. The terminal communication unit 31d, the terminal storage unit 32d, the terminal display unit 33d, the terminal input unit 34d, and the terminal control unit 35d are connected by a bus 36d. Each functional block has the same function as the corresponding functional block of the communication terminal 3, and a description thereof will be omitted.

  The remote monitoring terminal 3d receives an alarm from the server 2 in the same manner as the communication terminal 3. Further, the remote monitoring terminal 3d receives the video transmitted from the robot 1 from the server 2. The remote monitoring terminal 3d is preferably a personal computer used by employees at home. The inside of the store can be monitored by the remote monitoring terminal 3d based on the video taken by the robot 1 at a location outside the store (such as a remote location).

  That is, by using the remote monitoring terminal 3d, it is possible to constantly monitor the attention target person from the image and alarm of the robot 1 when the employee is at home. Thereby, the effect which suppresses criminal acts, such as shoplifting, can be heightened.

  Although the present invention has been described based on the drawings and examples, it should be noted that those skilled in the art can easily make various modifications and corrections based on the present disclosure. Therefore, it should be noted that these variations and modifications are included in the scope of the present invention. For example, the functions included in each means, each step, etc. can be rearranged so that there is no logical contradiction, and a plurality of means, steps, etc. can be combined or divided into one. .

  For example, although the example which uses the crime prevention system 10 in a bookstore was shown in this Embodiment, it is not restricted to this, It can use in arbitrary stores. In the present embodiment, the purpose is mainly to prevent shoplifting, but the present invention is not limited to this. It can also be used to suppress criminal acts such as annoying acts such as molesting and damage to equipment. Furthermore, a lost child, a lost person, or a wanted person may be registered in the person authentication database 221 as persons to be watched. When these persons are detected, an alarm is sent to detect the person to be watched. It is also applicable to.

DESCRIPTION OF SYMBOLS 1, 1c Robot 2 Server 3 Communication terminal 3d Remote monitoring terminal 4 Cash register 5a-5h Book shelf 6 Entrance 7a, 7b Clerk 8 Visitor 9 Marker 10, 10b, 10c, 10d Crime prevention system 11 Communication part 12 Moving means 13 Shooting part 14 Voice Interactive function unit 15 Display unit 16 Storage unit 17 Control unit 18, 24, 36, 36d Bus 21 Server communication unit 22 Server storage unit 23 Server control unit 31, 31d Terminal communication unit 32, 32d Terminal storage unit 33, 33d Terminal display unit 34, 34d Terminal input unit 35, 35d Terminal control unit 221 Person authentication database 341, 343 Input button 342 Home button 131 Fixed camera

Claims (15)

A security system comprising at least one robot that moves autonomously, a server, and a communication terminal,
The robot is
A vehicle that can move autonomously to approach the store visitor;
A voice interaction function unit for serving the customer;
A photographing unit for photographing the video of the customer during the service;
A communication unit for transmitting the video to the server,
The server
A server communication unit that receives the video from the robot;
A server control unit for determining whether the store visitor is a person to be alerted based on the video;
With
The server control unit is a crime prevention system that transmits an alarm to the communication terminal by the server communication unit when the store visitor is determined to be an attention target person ,
A security system in which the robot detects the visitor and approaches the visitor by the moving means when the possibility that the visitor is an attention target person is equal to or greater than a threshold value .   The security system according to claim 1, wherein the robot further includes a display unit that displays a search screen when the voice interaction function unit detects that the customer searches for a product. The crime prevention system according to claim 1 or 2 , wherein the threshold value is changed in a predetermined time zone. The crime prevention system according to any one of claims 1 to 3 , wherein the robot is capable of recognizing a home position, and the home position is located near a store entrance. The server
It has a person authentication database that stores face images related to the target person,
The server control unit extracts a face image of the visit's from the image, determines whether or not the visit who is watched person by the face authentication based on the person authentication database, of claims 1 to 4 The security system according to any one of the above. The crime prevention system according to claim 5 , wherein the person authentication database is encrypted. The server control unit extracts a behavior pattern of the visit's from the image, determines whether or not the visit person based on the behavior pattern is a note subject, any one of claims 1 to 6 Security system as described in The communication unit transmits the voice of the customer collected by the voice interaction function unit to the server;
The server communication unit receives the voice from the robot,
The crime prevention system according to any one of claims 1 to 7 , wherein the server control unit determines whether or not the store visitor is an attention target person based on the video and the audio. The server control unit determines a warning level for the attention target person based on the video,
The crime prevention system according to any one of claims 1 to 8 , wherein the warning includes information related to the warning level. When the robot receives the tracking instruction, the robot tracks the store visitor by the moving means,
The crime prevention system according to any one of claims 1 to 9 , wherein the photographing unit photographs at least one of a resolution and a frame rate of the store visitor before receiving the tracking instruction. . The crime prevention system further includes a fixed camera provided in the store,
The fixed camera transmits the captured second video to the server,
The server control unit determines whether or not the store visitor is an attention target person based on the second video. If the store control unit determines that the store visitor is an attention target person, the server communication unit The crime prevention system according to claim 10 , wherein the tracking instruction is transmitted to a robot. The security system comprises a plurality of robots,
A plurality of home positions corresponding to the plurality of robots are provided, a priority is set for the plurality of home positions, and the plurality of robots are set to one of the plurality of home positions based on the priority. The crime prevention system according to any one of claims 1 to 11 , which moves. The security system further comprises a remote monitoring terminal,
The crime prevention system according to any one of claims 1 to 12 , wherein the server transmits the video to the remote monitoring terminal. A security method by a security system comprising an autonomously moving robot, a server, and a communication terminal,
The robot
A moving step for autonomously moving to approach the store visitor according to the location of the store visitor;
A customer service step for serving the customer;
Photographing a video of the store visitor during the service by a photographing unit provided in the robot;
Transmitting the video to the server;
The server is
Receiving the video from the robot;
Determining whether the store visitor is a person to be alerted based on the video;
When it is determined that the store visitor is a person to be alerted, a step of transmitting an alarm to the communication terminal ,
The robot according to claim 1, wherein the robot detects the visitor, and when the possibility that the visitor is a person to be alerted is equal to or greater than a threshold, the robot approaches the visitor in the moving step . A vehicle that can move autonomously to approach the store visitor;
A voice interaction function unit for serving the customer;
A photographing unit for photographing the video of the customer during the service;
A communication unit for transmitting the video to a server;
With
When receiving a tracking instruction from the server, the robot tracking the visitor by the moving means ,
The robot detects the store visitor, and approaches the store visitor by the moving means when the possibility that the store visitor is an attention target person is equal to or greater than a threshold value .

Images