Monitoring device

The monitoring device efficiently tracks and analyzes suspicious individuals in crowded areas by identifying a predetermined person's actions and selecting targets based on positional relationships, reducing processing load and false alarms.

JP7954107B2Active Publication Date: 2026-11-19SECOM CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
JP2023049715
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2023-03-27
Publication Date
2026-11-19
Estimated Expiration
2043-03-27

AI Technical Summary

Technical Problem

Existing systems face excessive processing load and high false alarm rates when detecting suspicious individuals in crowded areas due to the large number of persons being analyzed.

Method used

A monitoring device that identifies a predetermined person, determines their actions, and selects other persons as targets based on their positional relationship, using a reference direction and field of view to prioritize analysis on suspicious individuals.

Benefits of technology

Facilitates efficient tracking and analysis of individuals in crowded areas by focusing on perceived suspicious persons, reducing the number of individuals to be tracked and improving the accuracy of suspicious person detection.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007954107000001
    Figure 0007954107000001
  • Figure 0007954107000002
    Figure 0007954107000002
  • Figure 0007954107000003
    Figure 0007954107000003
Patent Text Reader

Abstract

To facilitate processing to track and analyze persons in a monitoring area when there are a large number of persons in the monitoring area.SOLUTION: A monitoring device 10 includes: a predetermined person identifying unit 32 which identifies a predetermined person 2 located in a monitoring area 1; an action determination unit 34 which determines whether the predetermined person 2 performed a predetermined action; behavioral analysis units (36, 39) configured to perform processing for behavioral analysis of a person located in the monitoring area 1; and a target person selection unit 35 which selects another person, as a target person who is to be subjected to the above processing, on the basis of a relative positional relationship between the predetermined person and the other persons who are located in the monitoring area when the predetermined person performs the predetermined action.SELECTED DRAWING: Figure 2
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to a monitoring apparatus.

Background Art

[0002] Currently, systems have been proposed that track detailed behaviors (such as postures, movements, behavior histories, etc.) of persons detected by sensors to determine whether the persons are suspicious individuals. For example, the following Patent Document 1 discloses a technique of tracking a person whose face authentication result does not match a pre-registered registered person, and determining that the person is a suspicious individual when the behavior pattern of the person matches a predetermined pattern.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] When there is a large number of persons present in the monitoring area, performing analysis for suspicious person detection on all of these persons leads to an excessive load that makes the analysis difficult. In addition, the larger the number of persons targeted for analysis, the more likely it is to cause false alarms. The present invention has been made in view of the above problems, and an object of the present invention is to facilitate the processing of tracking and analyzing persons in the monitoring area when the number of persons in the monitoring area is large.

Means for Solving the Problems

[0005] A monitoring device according to one embodiment of the present invention comprises: a predetermined person identification unit that identifies a predetermined person present in a monitoring area; an action determination unit that determines whether or not the predetermined person has performed a predetermined action; an action analysis unit that performs processing for analyzing the actions of persons present in the monitoring area; and a target person selection unit that selects other persons as target persons for the above processing based on the relative positional relationship between the predetermined person and other persons present in the monitoring area when the predetermined person has performed a predetermined action. [Effects of the Invention]

[0006] According to the present invention, when there are many people in the surveillance area, the process of tracking and analyzing people in the surveillance area can be easily performed. [Brief explanation of the drawing]

[0007] [Figure 1] This is a schematic diagram of an example of the overall configuration of the monitoring system according to the embodiment. [Figure 2] This is a block diagram showing an example of the functional configuration of the monitoring device according to the first embodiment. [Figure 3] This is a schematic diagram of the first example of the reference direction for a given person. [Figure 4] This is a schematic diagram of the second example of the reference direction for the subject. [Figure 5] This is a schematic diagram illustrating another example of a method for selecting target individuals. [Figure 6] This is a schematic diagram of an example of a surveillance image from which the image region of the target person has been extracted. [Figure 7] This is a flowchart of an example of the monitoring method in the first embodiment. [Figure 8] This is a block diagram showing an example of the functional configuration of the monitoring device according to the second embodiment. [Figure 9] This is a flowchart of an example of a monitoring method according to the second embodiment. [Modes for carrying out the invention]

[0008] Embodiments of the present invention will be described below with reference to the drawings. The embodiments of the present invention described below are illustrative examples of devices and methods for realizing the technical concept of the present invention, and the technical concept of the present invention is not limited to the structure, arrangement, etc., of the components described below. Various modifications can be made to the technical concept of the present invention within the technical scope defined by the claims described in the patent claims.

[0009] (First Embodiment) Figure 1 is a schematic diagram of an example of the overall configuration of the monitoring system of the embodiment. The monitoring system of the embodiment is a system that monitors people present in the monitoring area 1, and comprises a monitoring device 10 installed in or near the monitoring area 1, and a central device 20 connected to the monitoring device 10 via a communication network such as a local area network (LAN) or the internet.

[0010] For example, surveillance area 1 may be a retail store such as a convenience store or supermarket, or a video rental store, and the person being monitored may be a user of these stores. As an example, let's assume that monitoring area 1 contains designated person 2, other persons 3a-3c, and structures 4a-4c. In the following explanation, persons 3a-3c will sometimes be collectively referred to as "person 3".

[0011] Designated Person 2 is a person who cooperates with the monitoring operations in monitoring area 1. Designated Person 2 may be, for example, an employee working in monitoring area 1 (e.g., a store employee), or a monitor (security guard) involved in monitoring operations in monitoring area 1. Person 3 is a person who is the subject of surveillance in surveillance area 1, and is, for example, an unspecified person who uses surveillance area 1.

[0012] Structures 4a and 4b are objects that can act as obstructions in the field of view for observing a person 3 from a predetermined person 2 or a camera 11, and are, for example, fixtures having a height that obstructs the field of view of the predetermined person 2 or the camera 11 (such as product shelves) or structural components of the building in the monitoring area 1 (such as pillars). The structure 4c is an object (such as a checkout counter) that is unlikely to act as an obstruction in the field of view for observing the person 3 from the predetermined person 2 or the camera 11.

[0013] One or more cameras 11a to 11d that capture images of the inside of the monitoring area 1 are connected to the monitoring device 10. In the following description, the cameras 11a to 11d may be collectively referred to as "camera 11". The monitoring device 10 includes an input unit 12, a communication unit 13, a storage unit 14, a control unit 15, and an output unit 16. Among these, the storage unit 14 and the control unit 15 can be implemented by a so-called computer, and the input unit 12, the communication unit 13, and the output unit 16 can be implemented as peripheral devices of the computer.

[0014] The input unit 12 includes user interfaces such as a keyboard and a mouse that are operated by a user and used for data input and the like. The input unit 12 is connected to the control unit 15, converts a user operation into an operation signal, and outputs the operation signal to the control unit 15. Further, the input unit 12 may include a DVD (Digital Versatile Disc) drive and a USB (Universal Serial Bus) interface. The input unit 12 inputs data to the control unit 15 as a file, and outputs data from the control unit 15 as a file.

[0015] The communication unit 13 includes a network interface or the like that transmits and receives data between the monitoring device 10 and the center device 20 via wired communication or wireless communication. Further, the communication unit 13 may include a network interface or the like that transmits and receives data between the monitoring device 10 and a mobile terminal device such as a smartphone or a tablet terminal carried by the predetermined person 2 via wireless communication. The storage unit 14 is a memory device such as a ROM (Read Only Memory) or a RAM (Random Access Memory), and stores various programs and various data. For example, the storage unit 14 records captured images generated by photographing the inside of the monitoring area 1 with the camera 11 (may be referred to as "monitoring images" in the following description). The storage unit 14 is an example of the "recording medium" described in the claims.

[0016] The control unit 15 is configured by an arithmetic device (processor) such as a CPU (Central Processing Unit), a DSP (Digital Signal Processor), or an MCU (Micro Control Unit). The control unit 15 is connected to the storage unit 14, operates as various processing units by reading and executing computer programs from the storage unit 14, stores various data in the storage unit 14, and reads out these data. The functions of the monitoring device 10 described below are realized by the control unit 15 executing the computer program stored in the storage unit 14.

[0017] The control unit 15 is also connected to the input unit 12 and the output unit 16, and starts or ends the monitoring processing of the monitoring area 1 by the monitoring device 10 in response to the user's operation on the input unit 12. The control unit 15 also outputs the monitoring image captured by the camera 11 to the output unit 16. In the second embodiment described later, the control unit 15 executes suspicious person determination processing for determining whether or not the person 3 present in the monitoring area 1 is a suspicious person, and outputs the determination result to the output unit 16. The control unit 15 is also connected to the communication unit 13, and transmits the monitoring image and the determination result of the suspicious person determination processing to the center device 20 via the communication unit 13.

[0018] The output unit 16 may include a display device such as a liquid crystal display or a CRT (Cathode Ray Tube) display. The output unit 16 is connected to the control unit 15 and may display surveillance images or the results of the suspicious person detection process performed by the control unit 15. The output unit 16 may also be equipped with an audio signal output device such as a speaker or buzzer. The output unit 16 may output audio information of the results of the suspicious person detection process performed by the control unit 15.

[0019] The central device 20 is, for example, a server and includes an input unit 22, a communication unit 23, a storage unit 24, a control unit 25, and an output unit 26. The input unit 22 is, for example, a touch panel, mouse, keyboard, etc., and receives operations from a monitor of the monitoring device equipped with a center device 20, and sends a signal corresponding to the received operation to the control unit 25.

[0020] The communication unit 23 includes a network interface, etc., for transmitting and receiving data between the center device 20 and the monitoring device 10 via wired or wireless communication. The storage unit 24 is a memory device such as ROM or RAM, and stores various programs and data. For example, the storage unit 24 records monitoring images received from the center device 20. The control unit 25 has a calculation unit similar to the control unit 15 of the monitoring device 10 and performs various signal processing for the center device 20.

[0021] For example, the control unit 25 receives surveillance images and the results of the suspicious person detection process transmitted from the monitoring device 10 via the communication unit 23. The control unit 25 also outputs the surveillance images and the results of the suspicious person detection process received from the monitoring device 10 to the output unit 26. The output unit 26 may include a display device such as a liquid crystal display or a CRT display. The output unit 26 may display surveillance images or the results of the suspicious person detection process. The output unit 26 may also include an audio signal output device such as a speaker or buzzer for outputting audio information of the results of the suspicious person detection process.

[0022] Figure 2 is a block diagram showing an example of the functional configuration of the monitoring device 10 of the first embodiment. The monitoring device 10 includes an image storage means 30, a person detection means 31, a predetermined person identification means 32, a reference direction estimation means 33, an action determination means 34, a target person selection means 35, an image extraction means 36, an image output means 37, and an image transmission means 38. The control unit 15 in Figure 1 functions as a person detection means 31, a predetermined person identification means 32, a reference direction estimation means 33, an action determination means 34, a target person selection means 35, and an image extraction means 36. The storage unit 14 functions as an image storage means 30. The output unit 16 functions as an image output means 37. The communication unit 13 functions as an image transmission means 38.

[0023] The image storage means 30 stores the surveillance images generated by the camera 11. The person detection means 31 detects a person present within the monitoring area 1 based on the monitoring image generated by the camera 11 and estimates the position of that person. The predetermined person identification means 32 identifies a predetermined person 2 from among the people detected by the person detection means 31.

[0024] For example, the predetermined person identification means 32 may detect a predetermined person 2 whose face image has been previously registered in the monitoring device 10 by performing image recognition processing on the face image of the person detected by the person detection means 31. Alternatively, for example, the predetermined person 2 may be detected by performing image recognition processing on the full-body image of the person detected by the person detection means 31, such as detecting a person wearing a predetermined outfit (e.g., a store uniform) or a predetermined accessory (e.g., an identification card, insignia, badge, etc.). Furthermore, for example, a person who has performed a predetermined behavior or gesture may be detected as a predetermined person 2.

[0025] The reference direction estimation means 33 estimates a reference direction, which is the direction specified by the gaze, head, body, etc., of a predetermined person 2, based on the image of the predetermined person 2 included in the surveillance image. Figure 3 is a schematic diagram of a first example of the reference direction of a predetermined person 2. For example, the reference direction estimation means 33 may estimate the gaze direction of the predetermined person 2 as the reference direction D. For example, the reference direction estimation means 33 may detect the orientation of the predetermined person 2's head from the monitoring image and detect the forward direction of the predetermined person 2's head as the gaze direction of the predetermined person 2. Alternatively, for example, the gaze direction of the predetermined person 2 may be detected by recognizing the pupil of the predetermined person 2's face image.

[0026] For example, the reference direction estimation means 33 may estimate the side or rear of the head of the predetermined person 2 as the reference direction D. The reference direction estimation means 33 may also estimate the front, side, or rear of the body (torso) of the predetermined person 2 as the reference direction D. The reference direction estimation means 33 may estimate the reference direction D based on the behavior of a predetermined person 2. For example, the direction specified by a gesture such as pointing with a finger or pointing with the chin may be estimated as the reference direction D. The reference direction estimation means 33 may estimate the orientation of the mobile terminal device held by a predetermined person 2 (for example, the vertical direction of the mobile terminal device) as the reference direction D. For example, the reference direction estimation means 33 may determine the orientation of the mobile terminal device by receiving detection signals from the mobile terminal device's electronic compass or gyro sensor via the communication unit 13.

[0027] Figure 4 is a schematic diagram of a second example of the reference direction of a given person 2. For example, the reference direction estimation means 33 may estimate the field of view range Rv of the given person 2 as the reference direction of the given person 2. That is, the reference direction of the given person 2 does not only mean one direction represented by a line segment starting from the position of the given person 2, but may also include an angular range with a finite width that has the position of the given person 2 as the origin. For example, the reference direction estimation means 33 may estimate the field of view range Rv as an angular range with a predetermined width that has the line of sight direction of the given person 2 as the center.

[0028] Furthermore, the reference direction estimation means 33 may calculate the field of view Rv as the area excluding the area occluded by structures 4a and 4b from the field of view geometrically calculated from the position and line of sight direction of a predetermined person 2, based on structural information regarding the arrangement position, size, and shape of structures 4a to 4c stored in the memory unit 14.

[0029] Refer to Figure 2. The action determination means 34 determines whether a predetermined person 2 has performed a predetermined action based on the image of the predetermined person 2 included in the surveillance image. For example, the predetermined action may be an action such as winking, raising a hand, scratching one's head, touching a specific part of one's face (e.g., nose or mouth), or making eye contact with the camera 11. Furthermore, the predetermined action may be a predetermined operation on a mobile terminal device held by the predetermined person 2, such as a button press or a graphical user interface (GUI) operation. For example, the reference direction estimation means 33 may receive from the mobile terminal device via the communication unit 13 whether or not these predetermined operations have been performed. The target person selection means 35 selects one of the persons 3a to 3c as the target person based on the relative positional relationship between the predetermined person 2 and persons 3a to 3c at the time the action determination means 34 determines that the predetermined person 2 has performed a predetermined action.

[0030] Refer to Figure 3. For example, the target person selection means 35 may select individuals 3a and 3b located in the reference direction D from the predetermined person 2 as target people when the predetermined person 2 performs a predetermined action. If there are multiple individuals located in the reference direction D, the target person selection means 35 may select the individual 3b that is closest to the predetermined person 2 among the individuals 3a and 3b located in the reference direction D as the target person.

[0031] Refer to Figure 4. For example, the target person selection means 35 may select as target people persons 3b and 3c that are within the field of view Rv of a predetermined person 2 when that predetermined person 2 performs a predetermined action. If there are multiple people within the field of view Rv, the target person selection means 35 may select as the target person person 3c that is closest to the predetermined person 2 among the people 3b and 3c within the field of view Rv.

[0032] Figure 5 is a schematic diagram of another example of a method for selecting target individuals. The target individual selection means 35 may select individuals 3a and 3b that are within a predetermined distance range Rd from the designated individual 2 when the designated individual 2 performs a predetermined action as target individuals. For example, the target person selection means 35 may select a person 3 that is located within a predetermined distance range Rd, regardless of the reference direction D or field of view range Rv of a predetermined person 2, as the target person. Alternatively, for example, if the target person selection means 35 determines that there is no person 3 located in the reference direction D or field of view range Rv, it may select a person 3 that is located within a predetermined distance range Rd as the target person. Furthermore, if there are multiple persons within a predetermined distance range Rd, the target person selection means 35 may select person 3b, who is closest to the predetermined person 2, as the target person from among persons 3a and 3b within the predetermined distance range Rd. Alternatively, the target person selection means 35 may estimate the orientation of persons 3b and 3c within the field of view range Rv and select a person who is not facing the predetermined person 2 or a person with their back to the predetermined person 2 (for example, 3c) as the target person. In this case, the orientation of the head may be detected from the surveillance image, and the orientation of persons within the field of view range Rv may be determined from the direction of the head's front.

[0033] The image extraction means 36 extracts images from the surveillance images stored in the image storage means 30 that capture the target person selected by the target person selection means 35. For example, the image extraction means 36 may extract surveillance images generated by camera 11, which captured the target person, from among the multiple cameras 11a to 11d. Alternatively, the image extraction means 36 may extract images of the target person from among the surveillance images generated by the camera 11 at multiple different times. For example, it may extract only the portion of the video that shows the target person, or frames that show the target person, from the video generated by the camera 11 as surveillance images.

[0034] Alternatively, the image extraction means 36 may extract a partial image region in which the target person is photographed from the entire area of ​​the surveillance image generated by the camera 11, and superimpose an identification display (for example, a figure indicating the partial image region in which the target person is photographed) on the surveillance image to identify the partial image region in which the target person is photographed. Figure 6 is a schematic diagram of an example of a surveillance image from which a partial image region of a target person has been extracted. The example in Figure 6 assumes that the surveillance image 40 contains people 3a and 3b and a structure 4a. The image extraction means 36 may superimpose a figure indicating the partial image region in which the target person 3a is photographed onto the surveillance image 40 as an identification display 41 to identify the target person 3a selected by the target person selection means 35. For example, the image extraction means 36 may superimpose a rectangle surrounding the partial image region of the target person 3a onto the surveillance image 40 as an identification display 41. Note that the image extraction means 36 is an example of the "behavior analysis unit" in the present invention.

[0035] Refer to Figure 2. The image output means 37 displays the image extracted by the image extraction means 36 on the display device of the output unit 16. The image transmission means 38 transmits the image extracted by the image extraction means 36 to the center device 20. Refer to Figure 1. The control unit 25 of the center device 20 receives the image (extracted image of the target person 3a) transmitted from the monitoring device 10 via the communication unit 23, and displays the received image on the display device of the output unit 26.

[0036] (operation) Figure 7 is a flowchart of an example of the monitoring method of the first embodiment. In step S1, the person detection means 31 detects a person present in the monitoring area 1 based on the monitoring image generated by the camera 11. In step S2, the predetermined person identification means 32 identifies a predetermined person 2 that is located within the monitoring area 1. The reference direction estimation means 33 estimates the reference direction of the predetermined person 2.

[0037] In step S3, the action determination means 34 determines whether or not a predetermined person 2 has performed a predetermined action based on the image of the predetermined person 2 included in the monitoring image. If the predetermined person 2 has not performed the predetermined action (step S3:N), the process proceeds to step S7. If the predetermined person 2 has performed the predetermined action (step S3:Y), the process proceeds to step S4. In step S4, the target person selection means 35 selects one of the persons 3a to 3c as the target person based on the relative positional relationship between the predetermined person 2 and persons 3a to 3c when the predetermined person 2 performs the predetermined action.

[0038] In step S5, the image extraction means 36 extracts images from the surveillance images stored in the image storage means 30 that capture the target person selected by the target person selection means 35. In step S6, the image output means 37 displays the image extracted by the image extraction means 36 on the display device of the output unit 16. The image transmission means 38 transmits the image extracted by the image extraction means 36 to the center device 20. The control unit 25 of the center device 20 displays the received image on the display device of the output unit 26. In step S7, the control unit 15 of the monitoring device 10 determines whether or not the operation to terminate the monitoring process has been performed at the input unit 12. If the termination operation has not been performed (step S7:N), the process returns to step S1. If the termination operation has been performed (step S7:Y), the process terminates.

[0039] (Second Embodiment) The monitoring device 10 of the second embodiment determines whether or not a person 3 present in the monitoring area 1 is a suspicious person. Figure 8 is a block diagram showing an example of the functional configuration of the monitoring device 10 of the second embodiment. The monitoring device 10 of the second embodiment has a configuration similar to the monitoring device 10 of the first embodiment, and the same or similar components are assigned the same reference numerals.

[0040] The monitoring device 10 of the second embodiment includes a suspicious person determination means 39 that determines whether a person 3 present in the monitoring area 1 is a suspicious person or not, based on the monitoring image generated by the camera 11. The control unit 15 in Figure 1 functions as the suspicious person determination means 39. The image extraction means 36 and the suspicious person determination means 39 are examples of the "behavior analysis unit" described in the claims. For example, the suspicious person detection means 39 may analyze the surveillance images generated by the camera 11 to calculate the degree of suspiciousness of person 3, and if the degree of suspiciousness exceeds the determination threshold Th, it may determine that person 3 is a suspicious person. For example, the suspicious person detection means 39 may calculate the likelihood that person 3's actions are suspicious as the degree of suspiciousness of person 3. For example, the suspicious person detection means 39 may estimate whether person 3 is making movements to put items into a bag or pocket, making movements to put multiple items into a bag or pocket, repeatedly touching a bag or pocket, or making preparatory movements for shoplifting, and calculate the likelihood that person 3's actions are suspicious. In the following description, the process of determining whether person 3, who is present in the monitoring area 1, is a suspicious person may be referred to as the "suspicious person detection process". Note that the suspicious person detection means 39 is an example of the "behavior analysis unit" in the present invention.

[0041] The suspicious person detection means 39 prioritizes determining whether a person selected by the target person selection unit 35 is a suspicious person from among the multiple persons 3a to 3c present in the monitoring area 1. In other words, the suspicious person detection process for the target person takes precedence over the suspicious person detection process for other persons. For example, the suspicious person detection means 39 may prioritize the suspicious person detection process for the target person by making the frequency of the suspicious person detection process for the target person higher than the frequency of the suspicious person detection process for other people. For example, the interval between the suspicious person detection process for the target person among the multiple people 3a to 3c may be shorter than the interval between the suspicious person detection processes for other people.

[0042] Furthermore, the suspicious person detection means 39 may prioritize the suspicious person detection process for a target person by making the accuracy of the suspicious person detection process for that target person higher than the accuracy of the suspicious person detection process for other people. For example, the suspicious person detection means 39 may make the accuracy of the suspicious person detection process for a target person higher than the accuracy of the suspicious person detection process for other people by changing the resolution of the image used for the suspicious person detection process or the algorithm of the suspicious person detection process. For example, the suspicious person detection means 39 may prioritize the suspicious person detection process for the target person by performing the suspicious person detection process for the target person before the suspicious person detection process for other people.

[0043] Furthermore, the suspicious person detection means 39 may make it easier to determine that the target person is suspicious compared to other persons among the multiple persons 3a to 3c present in the monitoring area 1. For example, the suspicious person detection means 39 may make it easier to determine that the target person is suspicious by setting the determination threshold Th, which is used to compare the suspiciousness of the target person, lower than the determination threshold Th, which is used to compare the suspiciousness of other persons.

[0044] The suspicious person detection means 39 adds the results of the suspicious person detection process to the surveillance image generated by the camera 11. For example, the suspicious person detection means 39 may superimpose symbols, characters, or figures indicating the results of the suspicious person detection process for each person 3 in the surveillance image onto the position of person 3 in the surveillance image or in its vicinity. The image output means 37 displays the surveillance image with the result of the suspicious person detection process added on the display device of the output unit 16. The image transmission means 38 transmits the surveillance image with the result of the suspicious person detection process added to the center device 20.

[0045] Refer to Figure 1. The control unit 25 of the center device 20 receives an image (a monitoring image with the result of the suspicious person detection process added) transmitted from the monitoring device 10 via the communication unit 23, and displays the received image on the display device of the output unit 26. Furthermore, the control unit 25 of the center device 20 may also include a suspicious person determination means 39. For example, the monitoring device 10 may transmit a monitoring image and target person information that identifies the target person selected by the target person selection means 35 to the center device 20. The control unit 25 of the center device 20 may perform the above-described suspicious person determination process to determine whether the person 3 present in the monitoring area 1 is a suspicious person or not. For example, the control unit 25 may prioritize determining whether the target person selected by the target person selection unit 35 is a suspicious person or not among a plurality of people 3a to 3c present in the monitoring area 1, making it easier to determine that the target person is a suspicious person compared to other people.

[0046] Figure 9 is a flowchart of an example of the monitoring method of the second embodiment. The processing in steps S11 to S14 is the same as the processing in steps S1 to S4 described with reference to Figure 7. In step S15, the suspicious person determination means 39 determines whether or not person 3 present in the monitoring area 1 is a suspicious person. For example, the suspicious person determination means 39 may prioritize determining whether or not a target person selected by the target person selection unit 35 is a suspicious person among a plurality of people 3a to 3c present in the monitoring area 1, and may make it easier to determine that the target person is a suspicious person compared to other people.

[0047] In step S16, the suspicious person detection means 39 adds the result of the suspicious person detection process to the surveillance image generated by the camera 11. The image output means 37 displays the surveillance image with the suspicious person detection result added on the display device of the output unit 16. The image transmission means 38 transmits the surveillance image with the suspicious person detection result added to the center device 20. The control unit 25 of the center device 20 displays the received image on the display device of the output unit 26. The process in step S17 is the same as the process in step S7 described with reference to Figure 7.

[0048] (Effects of the embodiment) (1) The monitoring device 10 includes a predetermined person identification means 32 for identifying a predetermined person present in the monitoring area, an action determination means 34 for determining whether or not the predetermined person has performed a predetermined action, an image extraction means 36 or a suspicious person determination means 39 which functions as an action analysis means for performing actions of persons present in the monitoring area, and a target person selection means 35 which selects other persons as target persons for the above processing based on the relative positional relationship between the predetermined person and other persons present in the monitoring area when the predetermined person has performed a predetermined action.

[0049] This improves the efficiency of monitoring operations within the monitoring area. In other words, in monitoring operations, not only is the system's determination of suspicious individuals important, but also the "suspiciousness" perceived by humans (such as store employees) on-site is an important factor in determining whether or not a person is suspicious. According to the present invention, the monitoring device 10 can select target individuals that humans cooperating in monitoring operations within monitoring area 1 perceive as suspicious. As a result, when there are many people in the monitoring area, more attention can be paid to and monitoring of target individuals that appear suspicious. Therefore, when there are many people in the monitoring area, the number of individuals to be tracked and analyzed by the monitoring device 10 and the central device 20 can be limited, making the process of tracking and analyzing individuals within the monitoring area easier.

[0050] (2) The monitoring device 10 may further include a reference direction estimation means 33 for estimating the reference direction of a predetermined person. The target person selection means 35 may select another person as the target person when the predetermined person performs a predetermined action and is located in the reference direction from the predetermined person. This allows for the identification of a target person based on their gaze direction, or a reference direction that can be represented by their head, body, etc.

[0051] (3) The reference direction estimation means 33 estimates the field of view of a predetermined person, and the target person selection means 35 may select another person who is within the field of view of the predetermined person when the predetermined person performs a predetermined action as the target person. This allows for the identification of a target person based on their field of vision.

[0052] (4) The target person selection means 35 may estimate the orientation of other people that are within the field of vision of the designated person when the designated person performs the designated action, and select a person who is not facing the designated person as the target person. This allows for the identification of a target person based on their positional relationship and body orientation.

[0053] (5) The monitoring device 10 may include a suspicious person determination means 39 that performs behavioral analysis to determine whether each of the multiple persons present in the monitoring area is a suspicious person or not. The suspicious person determination means 39 may prioritize determining whether the target person is a suspicious person or not among the multiple persons. This allows for efficient determination of whether or not a person within the surveillance area is a suspicious individual.

[0054] (6) The monitoring device 10 may include a suspicious person determination means 39 that performs behavioral analysis to determine whether or not each of the multiple persons present in the monitoring area is a suspicious person. The suspicious person determination means 39 may be configured to make it easier to determine that the target person is a suspicious person compared to the other persons among the multiple persons. This can improve the accuracy of identifying suspicious individuals.

[0055] (7) The monitoring device 10 may also include an image extraction means 36 for extracting an image of the target person from a recording medium on which an image of the monitored area has been captured. This allows for efficient monitoring of individuals whom designated individuals perceive as suspicious. [Explanation of Symbols]

[0056] 1... Monitoring area, 2... Designated person, 3, 3a~3c... Person, 4a~4c... Structure, 10... Monitoring device, 11, 11a~11d... Camera, 12, 22... Input unit, 13, 23... Communication unit, 14, 24... Storage unit, 15, 25... Control unit, 16, 26... Output unit, 20... Center device, 30... Image storage means, 31... Person detection means, 32... Designated person identification means, 33... Reference direction estimation means, 34... Behavior determination means, 35... Target person selection means, 36... Image extraction means, 37... Image output means, 38... Image transmission means, 39... Suspicious person determination means

Claims

1. A designated person identification unit that identifies a designated person from among multiple people present within the monitoring area, An action determination unit that determines whether or not a predetermined person identified by the predetermined person identification unit has performed a predetermined action, When the action determination unit determines that the designated person has performed the designated action, the target person selection unit selects a target person from among a plurality of other persons present in the monitoring area based on the relative positional relationship between the designated person and other persons present in the monitoring area at the time the designated person performed the designated action. A behavioral analysis unit performs processing for analyzing the behavior of the person selected by the aforementioned person selection unit, A monitoring device characterized by being equipped with the following features.

2. A predetermined person identification unit that identifies a predetermined person present in the monitoring area, An action determination unit that determines whether or not the aforementioned designated person performed a designated action, A behavioral analysis unit that performs processing for analyzing the behavior of persons present within the aforementioned monitoring area, A target person selection unit selects the other person as a target person to be processed based on the relative positional relationship between the designated person and other persons present within the monitoring area when the designated person performs a predetermined action, It comprises a reference direction estimation unit that estimates the reference direction of the predetermined person, The aforementioned target person selection unit selects another person located in the reference direction from the predetermined person when the predetermined person performs a predetermined action as the target person. A monitoring device characterized by the following features.

3. The reference direction estimation unit estimates the field of view of the predetermined person, The aforementioned target person selection unit selects other persons who are within the field of vision of the predetermined person when the predetermined person performs a predetermined action as the target person. The monitoring device according to feature 2.

4. The monitoring device according to claim 3, wherein the target person selection unit estimates the orientation of other persons that are within the field of view of the predetermined person when the predetermined person performs a predetermined action, and selects as the target person a person among those other persons who is not facing the predetermined person.

5. The surveillance device according to any one of claims 1 to 3, characterized in that the behavioral analysis unit performs behavioral analysis to determine whether each of the multiple persons present in the surveillance area is a suspicious person or not, and prioritizes determining whether the target person is a suspicious person or not among the multiple persons.

6. The surveillance device according to any one of claims 1 to 3, characterized in that the behavioral analysis unit performs behavioral analysis to determine whether or not each of the multiple persons present in the surveillance area is a suspicious person, and makes it easier for the target person to be determined to be a suspicious person compared to the other persons among the multiple persons.

Citation Information

Patent Citations

  • Monitoring system and monitoring method

    JP2007213224A

  • Security device

    JP2008108186A

  • Intruder monitoring device

    JP2010211514A