JP2011128911A - Target detection system, target detection method, target detecting apparatus and mobile information acquisition device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To detect the location and behavior of a target not only when the target is located in a detection range of a fixed camera but also when the target is located outside the detection range of the fixed camera. <P>SOLUTION: A management device uses a fixed camera A to detect appearance information of the target, and refines the target on the basis of the detected appearance information. The management device also uses detection results of appearance information by the fixed camera A and a fixed camera B to detect an abnormal condition of the target. Then, the management device, upon detecting the abnormal condition of the target, moves a mobile robot to non-detection ranges of the fixed camera A and the fixed camera B, and identifies the target on the basis of face information detected by using the mobile robot. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a target person detection system, a target person detection method, a target person detection apparatus, and a mobile information acquisition apparatus that detect a target person located in a predetermined area.

  Conventionally, various methods for managing the location of a target person in facilities such as welfare facilities and offices have been proposed.

  For example, in Patent Document 1, when an IC tag is worn by a target person and the identification information of the IC tag is read by a reading device installed in a facility, A location management system that displays an image of an installed surveillance camera on a display of a management device is disclosed.

  Further, in Patent Document 2, a face authentication device is installed at the entrance of each room in the facility, and the subject person who enters the room is subjected to an authentication procedure using the face authentication device. A location management system for managing the location is disclosed.

JP 2007-272852 A JP 2007-80149 A

  However, a method for detecting a target person using an information acquisition device (hereinafter referred to as a fixed information acquisition device) that is fixedly installed such as an IC tag reading device, a monitoring camera, or a face authentication device includes a fixed information acquisition device. There is a problem that it is impossible to detect a subject who is outside the acquisition range (for example, blind spot of a surveillance camera).

  Here, when it is desired to reduce the places outside the acquisition range, it is conceivable to install a large number of fixed information acquisition devices in the facility. However, adding a fixed information acquisition device requires a large amount of money. Even if it is added, the fixed information acquisition device that is fixed to a wall or ceiling is used. Because there is a limit to the cover, it is not realistic.

  Further, if the number of surveillance cameras is increased, the burden on the supervisor who monitors the video of the surveillance camera increases, and as a result, there is a possibility that a serious crisis that the target person has fallen over may be missed.

  Therefore, not only when the subject is within the acquisition range of the fixed information acquisition device, but also when the subject is located outside the acquisition range of the fixed information acquisition device. How to realize a target person detection system or a target person detection method that can detect the problem is a big problem.

  For example, when the fixed information acquisition device is a face authentication device, even if the target person is shown on the face authentication device, if the target person is not shown in a state where no face is shown such as sideways or behind, the target person It can be said that it is located outside the acquisition range (outside the detection range) of the authentication device. That is, “when the subject is located outside the acquisition range of the fixed information acquisition device” indicates a case where the information acquisition device cannot detect the subject. The same applies to other biometric authentication devices such as IC tag reading devices and fingerprint authentication devices.

  The present invention has been made to solve the above-described problems caused by the prior art, and not only when the target person is located within the acquisition range of the fixed information acquisition apparatus but also when the target person acquires the fixed information. An object of the present invention is to provide a target person detection system, a target person detection method, a target person detection apparatus, and a mobile information acquisition apparatus capable of detecting the location and behavior of a target person including a case where the target person is located outside the acquisition range of the apparatus. And

  In order to solve the above-described problems and achieve the object, the present invention includes a storage unit that stores information for specifying a subject as registration information, a movable mobile information acquisition device that acquires information about the subject, A fixed information acquisition device that is fixedly installed and acquires information about the subject, and a collation unit that detects the subject by collating information acquired by the fixed information acquisition device with the registration information; And an abnormality detection unit that detects an abnormality of the subject based on the detection result of the subject by the collation unit. For this reason, according to the present invention, not only when the subject is located within the acquisition range of the fixed information acquisition device, but also when the subject is located outside the acquisition range of the fixed information acquisition device. There is an effect that it is possible to detect the whereabouts and behavior.

  Further, the present invention provides an instruction to move the mobile information acquisition device to a non-acquisition range of the fixed information acquisition device when the abnormality detection unit detects an abnormality of the subject in the above invention. It is further characterized by further including a movement instruction unit for performing the operation. For this reason, according to this invention, when abnormality of a subject person is detected, the information of the subject person who exists in the non-acquisition range of a fixed information acquisition device is acquirable.

  Further, the present invention is the above invention, wherein the fixed information acquisition device includes a first fixed information acquisition device and a second fixed information acquisition device installed along a predetermined path, and the abnormality The detection unit is configured to detect the target person based on an elapsed time from when the target person is detected by the first fixed information acquisition apparatus to when the target person is detected by the second fixed information acquisition apparatus. When the abnormality is detected by the abnormality detection unit, the movement instructing unit detects a non-operation between the first fixed information acquisition device and the second fixed information acquisition device. The mobile information acquisition apparatus is instructed to move to a range determined based on the elapsed time in the acquisition range. For this reason, according to the present invention, for example, the mobile information acquisition device can be directed to the subject who seems to have experienced an abnormality such as fainting.

  Further, in the present invention according to the above-described invention, the collation unit may be the fixed information acquisition device or the mobile information acquisition device when the movement instruction unit is not instructed to move the mobile information acquisition device. The target person is detected based on the information acquired by the above. Therefore, according to the present invention, as a result of the mobile information acquisition device also serving as a fixed information acquisition device, the number of imaging units that are fixedly installed can be reduced.

  Further, the present invention is the above invention, wherein the mobile information acquisition device includes an information acquisition unit that acquires information about the target person, and travels a predetermined route when there is no instruction from the movement instruction unit. Then, the information on the subject is acquired, and the movement instruction unit is based on the detection result of the fixed information acquisition device and / or the information on the subject acquired by the mobile information acquisition device. The mobile information acquisition device is instructed to move to a non-acquisition range of the information acquisition device. For this reason, according to the present invention, the mobile information acquisition device circulates a predetermined route, whereby the number of imaging units that are fixedly installed can be further reduced.

  Further, the present invention is the above invention, wherein the fixed information acquisition device or the mobile information acquisition device includes an imaging unit that acquires an image of the subject as information about the subject, and the storage unit includes: The appearance information of the subject person is stored as the registration information, and the collation unit collates the image of the subject person acquired by the imaging unit with the appearance information stored in the storage unit. A target person associated with appearance information having a high degree of coincidence is narrowed down as a detection target. For this reason, according to the present invention, it is possible to narrow down who the target person is even when the target person is at a position away from the imaging unit or when the face of the target person is hidden.

  Further, the present invention is the above invention, wherein the storage unit further stores face information of the subject as the registration information, and the mobile information acquisition device includes the image of the subject as information about the subject. The collation unit identifies the target person by collating the face image of the target person narrowed down as the detection target with the image captured by the imaging unit. For this reason, according to the present invention, it is possible to more reliably identify the target person narrowed down by the narrowing-down means.

  Further, the present invention is the above invention, wherein the mobile information acquisition device is a mobile terminal device, and when the abnormality of the subject is detected by the abnormality detection unit, based on the detection mode of the abnormality, The information processing apparatus further includes a notification unit that notifies the mobile terminal device that the information acquired by the fixed information acquisition device and the abnormality are detected. Therefore, according to the present invention, compared with the case where the mobile information acquisition device is a self-propelled mobile robot, the cost required for system introduction can be suppressed.

  According to the present invention, not only when the subject is located within the acquisition range of the fixed information acquisition device, but also when the subject is located outside the acquisition range of the fixed information acquisition device, There is an effect that an action can be detected.

FIG. 1 is a diagram showing an outline of a subject detection method according to the present invention. FIG. 2 is a diagram illustrating a network environment to which the management apparatus according to the present embodiment is connected. FIG. 3 is a block diagram illustrating the configuration of the management device and the mobile robot according to the present embodiment. FIG. 4 is a diagram illustrating an example of registration information. FIG. 5 is a diagram for explaining an operation example of the abnormality detection unit and the movement instruction unit. FIG. 6 is a diagram for explaining an operation example of the abnormality detection unit and the movement instruction unit when a target person for which appearance information is not registered is detected. FIG. 7 is a diagram illustrating an example of an external configuration of a mobile robot. FIG. 8 is a sequence diagram illustrating a processing procedure executed by the management apparatus and the mobile robot according to the present embodiment. FIG. 9 is a diagram illustrating a display example of the display. FIG. 10 is a diagram for explaining a case where the mobile robot is also used as a fixed camera. FIG. 11 is a diagram for explaining a case where the subject detection method according to the present invention is applied outdoors. FIG. 12 is a diagram for explaining a case where the mobile robot covers outside the reading range of the fixed IC tag reader.

  Embodiments of a subject detection system and subject detection method according to the present invention will be described below in detail with reference to the accompanying drawings.

  First, prior to detailed description of the embodiment, an outline of the subject detection method according to the present invention will be described with reference to FIG. FIG. 1 is a diagram showing an outline of a subject detection method according to the present invention. As shown in the figure, in the target person detection method according to the present invention, the target person is narrowed down based on the appearance information detected using the fixed information acquisition apparatus, and the face detected using the mobile information acquisition apparatus It is characterized in that the target person is specified based on the information. The subject detection method according to the present invention is also characterized in that the non-acquisition range (dead angle) of the fixed information acquisition device is supplemented by the mobile information acquisition device.

  Hereinafter, as an example of the information acquisition apparatus, a description will be given using a camera that captures an image of a target person. Specifically, a fixed camera that is fixedly installed in a place such as a wall or a ceiling in a facility will be used as the fixed information acquisition device, and a mobile robot equipped with the camera will be used as the mobile information acquisition device. .

  As shown to (A) of the figure, the image imaged with the fixed camera A and the fixed camera B is sent to the object detection apparatus (henceforth "management apparatus") which is not shown in figure. Then, the management apparatus detects the appearance information of the target person from the image captured by the fixed camera A (see (A-1) in the figure), and narrows down the target person based on the detected appearance information ((A-2 in the figure)). )reference).

  Here, the appearance information is information (appearance information) regarding the appearance of the subject. Specifically, the appearance information includes static information such as a silhouette of the subject's clothes and hairstyle or clothes color and dynamic information such as a gait (how to walk) as a feature amount. In addition, the appearance information may include face information such as a profile of the subject or a front face as a feature amount.

  Subsequently, the management apparatus detects that the target person is not detected by the fixed camera B that should next detect the target person even though a predetermined time has passed since the target person was detected by the fixed camera A (see FIG. (See (A-3)), the abnormality of the subject is detected (see (A-4) in the figure). As such an abnormality, for example, a case where the subject falls down on the route between the fixed camera A and the fixed camera B, or a case where the subject is tricking the route between the fixed camera A and the fixed camera B can be considered.

  When the abnormality of the subject person is detected in this way, the management device passes the appearance information and face information of the subject person registered in advance to the mobile robot. The management apparatus is a path between the fixed camera A and the fixed camera B, and moves the mobile robot using the non-detection range (dead angle) of the fixed camera A and the fixed camera B as a search range ((B -1)).

  Here, appearance information other than face information can be detected even when the subject is located at a relatively distant position from the fixed camera. On the other hand, it is said that the matching accuracy is lower than that using face information. Has properties. Therefore, when the mobile robot finds a target person within the search range based on the appearance information passed from the management apparatus, the mobile robot approaches the target person and detects face information (see (B-2) in the figure). The mobile robot collates the detected face information with the face information passed from the management device, thereby reliably identifying the target person narrowed down by the appearance information ((B- 3)).

  As described above, in the subject detection method according to the present invention, the mobile information acquisition device is moved to the non-acquisition range of the fixed information acquisition device based on the detection result by the fixed information acquisition device. The location and behavior of the subject can be detected not only when the subject is located within the detection range of the camera but also when the subject is located outside the detection range of the fixed camera.

  In the subject detection method according to the present invention, the appearance information of the subject is detected by a fixed camera, and the face information of the subject is detected by a mobile robot. For this reason, according to the subject detection method according to the present invention, it is possible to accurately identify the subject by face matching using the mobile robot after narrowing down the subject in which the abnormality has occurred with the appearance information.

  If a high-resolution camera is used as the fixed camera, the target person's face information can be detected even when the target person is located at a relatively distant position. Can also be used. That is, the appearance information may include face information.

  Further, here, face information is used as information for identifying the person, but the information is not limited to this. For example, biometric information such as fingerprints, veins, or irises is used as information for identifying the person. Also good.

  By the way, the appearance information includes clothes, clothes color, and the like as feature quantities, and thus can be said to be information based on short-term characteristics of the subject. That is, when the subject changes clothes, there is a problem that the appearance information registered in advance and the appearance information detected after changing clothes do not match, so that the subject cannot be collated. Therefore, in the subject detection method according to the present invention, a registration camera is installed in the entrance / exit of the facility, in the elevator, or in the hall in front of the elevator, etc., and the appearance information is obtained every time the facility enters / exits or the floor is moved. I am going to register again.

  Below, the Example about the subject detection system to which the subject detection method demonstrated using FIG. 1 is applied is described in detail. In addition, the Example shown below demonstrates the case where behavior of care recipients, such as elderly people, are managed in welfare facilities.

  FIG. 2 is a diagram illustrating a network environment to which the management apparatus according to the present embodiment is connected. Note that (A) in the figure shows the connection relationship of each device including the management device.

  As shown in FIG. 3A, in this embodiment, the management apparatus 10 is connected to a registration camera and fixed cameras A to D, which are fixed information acquisition units, via a network such as a LAN (Local Area Network). Has been. The management apparatus 10 is connected to the mobile robot 20 via a wireless network such as a wireless LAN. In addition, although (A) of the figure illustrates the case where four fixed cameras are used, any number of fixed cameras may be used.

  The management device 10 is a device that manages the behavior of the target person using images captured by the fixed cameras A to D and instructs the mobile robot 20 to search for the target person. In addition, the mobile robot 20 is equipped with a registration camera and a camera (information acquisition unit) similar to the fixed cameras A to D, and moves to a target location in accordance with an instruction from the management device 10 to search for a target person. It is a mobile robot. Note that the resolution of the camera of the mobile robot 20 may be higher than that of the fixed camera, or the focal length of the camera of the mobile robot 20 may be shorter than that of the fixed camera.

  FIG. 5B shows the arrangement of each device in the welfare facility. As shown to (B) of the figure, the management apparatus 10 and the mobile robot 20 are installed in the management room where the manager of a welfare facility stands by. Here, when receiving an instruction from the management device 10, the mobile robot 20 leaves the management room and moves to a target location in the welfare facility. The mobile robot 20 may be circulated in the welfare facility instead of waiting in the management room.

  The registration camera is installed at the entrance of the welfare facility, captures an image of a person who visits the welfare facility, and sends the captured image to the management apparatus 10 via the network. A subject who visits a welfare facility enters the facility after registering appearance information and face information using the registration camera. However, when face information has already been registered, only appearance information, which is short-term information, may be registered in association with the registered face information.

  Here, the entrance of the welfare facility is a registration gate in which one registration camera and a gate device (not shown) are installed. Specifically, when the registration procedure for the subject person by the registration camera is completed, the management device 10 opens the gate of the gate device to allow the subject person to enter the welfare facility. Note that a plurality of registration cameras may be installed at the entrance of the welfare facility so that the target person can be imaged from a plurality of angles. Also, a predetermined number of cameras for acquiring face information and cameras for acquiring appearance information may be provided. If the welfare facility is a multi-storey building, a registration camera may be installed in the elevator or on the stairs.

  The fixed cameras A to D are fixedly installed in a passage or a room in the welfare facility, pick up an image of a target person located in the passage or the room, and send the picked up image to the management apparatus 10 via the network. That is, the fixed cameras A to D acquire the image of the subject as information about the subject. Here, in (B) of the figure, the range in which the appearance information of the subject can be detected among the imaging ranges of the fixed cameras A to D is indicated by hatching. That is, the part not hatched in (B) in the figure is the non-detection range (dead angle) of the fixed cameras A to D.

  Here, the registered camera has been described as a dedicated camera for the registration procedure, but the registered camera may also be used as a fixed camera.

  Next, the configuration of the management apparatus 10 and the mobile robot 20 according to the present embodiment will be described with reference to FIG. FIG. 3 is a block diagram illustrating the configuration of the management device 10 and the mobile robot 20 according to the present embodiment. In the figure, only components necessary for explaining the features of the management apparatus 10 and the mobile robot 20 are shown, and descriptions of general components are omitted.

  As shown in the figure, the management apparatus 10 includes a wireless communication unit 11, a display 12, a speaker 13, a control unit 14, and a storage unit 15. The control unit 14 includes a registration processing unit 14a, an appearance collation unit 14b, an abnormality detection unit 14c, a movement instruction unit 14d, and a collation result acquisition unit 14e. The storage unit 15 stores registration information 15a, verification history information 15b, and in-facility map information 15c.

  On the other hand, the mobile robot 20 includes a wireless communication unit 21, a drive unit 22, a camera 23, a speaker 24, a control unit 25, and a storage unit 26. The control unit 25 includes a movement execution unit 25a, an appearance collation unit 25b, and a face collation unit 25c. The storage unit 26 stores in-facility map information 26a.

  Below, each component of the management apparatus 10 is demonstrated first. The wireless communication unit 11 is a wireless communication device that transmits and receives various data to and from the mobile robot 20. The display 12 is a display device that displays various types of information, and the speaker 13 is an audio output device that notifies various types of information by voice.

  The control unit 14 is a processing unit that performs processing such as subject registration processing, appearance verification, detection of an abnormality of the subject based on the verification history of appearance verification, movement instructions for the mobile robot 20, and acquisition of verification results from the mobile robot 20. It is.

  The registration processing unit 14a detects the appearance information and face information of the target person from the image captured by the registration camera, and stores the registration information 15a in the storage unit 15 in association with the target person ID for identifying the target person. It is. Here, the registration information 15a stored in the storage unit 15 will be described with reference to FIG. FIG. 4 is a diagram illustrating an example of the registration information 15a.

  As shown in FIG. 4, the registration information 15a is information in which face information and appearance information are associated with each target person. Specifically, as shown in FIG. 5A, the registration information 15a includes a “subject ID” item, a “name” item, a “face information ID” item, and an “appearance information ID” item. This information includes

  The “target person ID” item is an item in which a target person ID for identifying the target person is stored. The “name” item is an item in which the name of the target person is stored. The “face information ID” item is an item in which a face information ID for identifying face information is stored. The “appearance information ID” item is an item in which appearance identification information for identifying appearance information is stored. For example, in the registration information 15a, the name “Eiko Taro”, the face information ID “FA001”, and the appearance information ID “AP0841” are stored in association with the subject ID “001”.

  Further, as shown in FIG. 5B, the registration information 15a stores face information in association with the face information ID. Similarly, as shown in (C) of the figure, the registration information 15a stores appearance information in association with the appearance information ID. Here, face information and appearance information are stored as image data. However, a feature amount extracted from image data may be stored as face information and appearance information.

  Returning to FIG. 3, the appearance verification unit 14b will be described. The appearance collation unit 14b collates the images captured by the fixed cameras A to D with the appearance information stored as the registration information 15a, thereby identifying the target person associated with the appearance information having a high degree of coincidence with the image. It is a processing unit that narrows down as a detection target.

  For example, the appearance matching unit 14b creates a template from appearance information (image data) stored as the registration information 15a, and matches images captured by the fixed cameras A to D with the template. Subsequently, the appearance collation unit 14b calculates the degree of coincidence between the images captured by the fixed cameras A to D and the template, and when there is an area in the captured image where the degree of coincidence is equal to or greater than the threshold, the appearance information Detect as information. Thus, appearance collation part 14b specifies appearance information with high coincidence with appearance information detected from a picked-up image.

  Then, the appearance collation unit 14b extracts the target person ID associated with the identified appearance information from the registration information 15a, and associates it with the camera ID of the camera that detected the target person and the current date and time as collation history information 15b to the storage unit 15. Remember.

  In some cases, the fixed cameras A to D detect appearance information that does not match any of the appearance information stored as the registration information 15a or has a low matching degree. This may occur, for example, when a target person (suspicious person) who has not performed the registration procedure at the entrance of the welfare facility is detected by the fixed cameras A to D.

  In such a case, the appearance collation unit 14b has a subject ID (for example, “unknown 01” or the like) indicating that it is an unknown subject with respect to the camera ID of the fixed camera that detected the subject (suspicious person). The current date and time are associated with each other and stored in the storage unit 15 as collation history information 15b. In this way, by associating an ID for identifying a suspicious person for each appearance information of the suspicious person, even if there are multiple suspicious persons in the welfare facility, the location and behavior of each suspicious person can be individually determined. Can be managed.

  The abnormality detection unit 14c is a processing unit that detects the abnormality of the subject using the detection results of the subject using the fixed cameras A to D by the appearance matching unit 14b. For example, when the time when the target person is detected by the fixed camera A is set as the reference time, the abnormality detection unit 14c uses the fixed camera B to detect the target person within a predetermined time from the reference time. If no is detected, the abnormality of the subject is detected. Further, after the target person is detected by the fixed camera A, when the target person is detected earlier than usual by the fixed camera B, the abnormality of the target person is detected.

  As described above, the abnormality detection unit 14c detects the abnormality of the subject based on the elapsed time from when the subject is detected by the first fixed camera to when the subject is detected by the second fixed camera. To do. Details of this point will be described later with reference to FIG.

  The movement instruction unit 14d is a processing unit that instructs the mobile robot 20 to go to the subject when the abnormality detection unit 14c detects the abnormality of the subject. First, the movement instruction unit 14d refers to the in-facility map information 15c, and specifies a place where the target person is supposed to be located, that is, a search range of the mobile robot 20. Here, the in-facility map information 15c is information including a map in the welfare facility, an installation location of the fixed cameras A to D in the facility, an imaging direction and a detection range of the fixed cameras A to D, and the like.

  Subsequently, the movement instructing unit 14d refers to the registration information 15a and specifies the face information and appearance information of the target person to be searched. Then, the movement instruction unit 14 d transmits a movement instruction including the search range, face information, and appearance information to the mobile robot 20. As a result, the mobile robot 20 moves to the designated search range, searches for the target person, and performs face matching.

  The matching result acquisition unit 14e is a processing unit that receives the matching result of the face matching from the mobile robot 20 and notifies the information corresponding to the received matching result using the display 12 or the speaker 13.

  Here, an operation example of the abnormality detection unit 14c and the movement instruction unit 14d will be described with reference to FIG. FIG. 5 is a diagram for explaining an operation example of the abnormality detection unit 14c and the movement instruction unit 14d. FIG. 5A shows the contents of the collation history information 15b that is the premise of the description. In the following, as shown in FIG. 6A, a case where “fixed camera A” detects a target person with the target person ID “001” at “November 1, 21:30:30” will be described. To do.

  Further, (B) in the figure shows an abnormality detection condition corresponding to the matching history information 15b shown in (A) in the figure. As shown in FIG. 5B, the abnormality detection unit 14c detects that the detection time of the target person “001” by the fixed camera B is the time “21:30” when the target person “001” is detected by the fixed camera A. If it is after “21:40:00” 10 minutes after “minute 00 seconds” (reference time), an abnormality of the subject “001” is detected (abnormality 1).

  Here, as the occurrence state of the abnormality 1, for example, the subject “001” is a passage between the fixed camera A and the fixed camera B, and is outside the detection range of the fixed camera A and the fixed camera B (dead angle). ), Or when falling within such a range.

  When the abnormality detection unit 14c detects “abnormality 1” of the subject “001”, the movement instruction unit 14d gives a movement instruction including the face information, appearance information, and search range of the subject “001” to the mobile robot 20. Send to. Here, the search range of the subject “001” is a path between the fixed camera A and the fixed camera B and outside the detection range (dead angle) of the fixed camera A and the fixed camera B. As described above, the movement instruction unit 14d includes the first fixed camera in the non-detection range between the first fixed camera and the second fixed camera when the abnormality detection unit 14c detects an abnormality of the subject. The mobile robot 20 is instructed to move to a range determined based on the elapsed time from when the subject is detected by the second fixed camera until the subject is detected by the second fixed camera.

  As a result, the mobile robot 20 is a path between the fixed camera A and the fixed camera B and is outside the detection range of the fixed camera A and the fixed camera B as shown in FIG. Move to the blind spot). Then, the target person is searched while the appearance information is acquired by the camera 23, and it is specified by face matching that the target person included in the movement instruction is indeed the target person “001”.

  On the other hand, as shown in FIG. 5B, the abnormality detection unit 14c detects that the target person “001” is detected by the fixed camera B, and the time “21” when the target person “001” is detected by the fixed camera A. Even when it is earlier than “21:31:00” that is one minute after “30:00” (reference time), the abnormality of the subject “001” is detected (abnormality 2).

  This indicates that the subject “001” has moved through the path between the fixed camera A and the fixed camera B at an abnormal speed. That is, as a situation of occurrence of abnormality 2, there may be mentioned a case where the subject “001” is in an unstable state.

  When the abnormality detection unit 14c detects “abnormality 2” of the subject “001”, the movement instruction unit 14d is a path between the fixed cameras B to D and the detection range of the fixed cameras B to D. A movement instruction specifying the outside (dead angle) as the search range of the subject “001” is transmitted to the mobile robot 20.

  Thereby, as shown in (C-2) of the figure, the mobile robot 20 is a path between the fixed cameras B to D and a place outside the detection range (dead angle) of the fixed cameras B to D. Move up. Then, the target person is searched while the appearance information is acquired by the camera 23, and it is specified by face matching that the target person included in the movement instruction is indeed the target person “001”.

  As described above, in this embodiment, when the location of the target person in the welfare facility is roughly grasped by the fixed cameras A to D, and the abnormality of the target person outside the detection range of the fixed cameras A to D is detected. In this case, the mobile robot 20 is dispatched, and the target person whose abnormality is detected is accurately specified by face matching. For this reason, the target person in a welfare facility can be managed more efficiently, and even when only a few administrators can be prepared, the risk of overlooking the occurrence of an abnormality in the entire welfare facility can be reduced. .

  Further, for example, even when “abnormality 1” is detected, the target person whose abnormality is detected is actually only looking out from the window, or is talking with another target person. There is also the possibility of just doing. In this way, if the administrator is sent to the site in a situation where it cannot be determined that an abnormality has occurred reliably, a great burden is placed on the administrator. Therefore, as in the present embodiment, when an abnormality is detected, the burden on the administrator can be reduced by moving the mobile robot 20 to the site prior to the dispatch of the administrator.

  By the way, although FIG. 5 demonstrated the case where the abnormality of the registered target person was detected at the entrance of a welfare facility, the target person (suspicious person) who did not perform registration procedure at the entrance lurked in the welfare facility. In some cases. Accordingly, an example of the operation of the abnormality detection unit 14c and the movement instruction unit 14d when a target person whose appearance information is not registered is detected will be described with reference to FIG.

  FIG. 6 is a diagram for describing an operation example of the abnormality detection unit 14c and the movement instruction unit 14d when a target person whose appearance information is not registered is detected. Here, FIG. 6A shows the contents of the collation history information 15b as a premise of the description.

  As shown in (A) of the figure, a target person (suspicious person) with the target person ID “unknown 01” was detected by “fixed camera B” at “November 1, 23:55:30”. In this case, the abnormality detection unit 14c detects the abnormality of the target person, that is, the intrusion of the suspicious person. When such an abnormality is detected by the abnormality detection unit 14c, the movement instruction unit 14d is a path between the fixed cameras B to D as shown in FIG. A movement instruction in which the outside of the detection range (dead angle) of the cameras B to D is designated as the search range of the subject “001” is transmitted to the mobile robot 20.

  In FIG. 5B, the mobile robot 20 is directed to the path between the fixed cameras B to D and out of the detection range of the fixed cameras B to D. The mobile robot 20 may be directed to the path between the cameras B and out of the detection range of the fixed cameras A and B. In this way, when there are a plurality of search range candidates, the movement instruction unit 14d determines the traveling direction of the subject from the images captured by the fixed cameras A to D, and determines the search range according to the calculated traveling direction. That's fine.

  Moreover, the movement instruction | indication part 14d is good also as moving the mobile robot 20 within the detection range of the fixed camera which detected this suspicious person, when a suspicious person is detected.

  Returning to FIG. 3, the configuration of the mobile robot 20 will be described. The wireless communication unit 21 is a wireless communication device that transmits and receives movement instructions, image data, and the like to and from the management apparatus 10. The drive unit 22 is a mechanism unit for driving each part of the mobile robot 20. For example, the drive unit 22 detects a motor, a wheel, or an obstacle for moving the mobile robot 20, and measures a distance to the subject. It consists of sensors for the purpose. The camera 23 is an image capturing unit for capturing an image of the target person, like the registration camera and the fixed cameras A to D, and the speaker 24 is an audio output device that notifies the target person and the like of predetermined audio information. It is.

  Here, the external configuration of the mobile robot 20 shown in FIG. 3 will be described with reference to FIG. FIG. 7 is a diagram illustrating an example of an external configuration of the mobile robot 20. Note that the appearance configuration of the mobile robot 20 shown in FIG. 7 is merely an example, and other appearances may be used.

  As shown to (A) of the figure, the mobile robot 20 has the head 20a and the main-body part 20b. The head 20a is formed to imitate a human head, and a camera 23 is built in a portion corresponding to the human eye, and a speaker 24 is built in a portion corresponding to the human mouth.

  Moreover, the drive part 22 is incorporated in the main-body part 20b. For example, the drive part 22 rotates the wheel provided in the bottom part of the main-body part 20b, and moves the mobile robot 20. FIG. Here, the drive unit 22 has an obstacle detection function and a self-positioning function. As the obstacle detection function, for example, a laser omnidirectional distance sensor can be used. This detects an obstacle by detecting the reflected light of the laser beam emitted from its own device, or measures the distance to the subject.

  Moreover, as a self-localization function, a laser self-localization system can be used, for example. In this method, a marker for confirming the self position is attached in advance on the route in the facility, and the self position is identified by detecting the laser reflected light from the marker. As another self-localization function, a method of identifying the self-position by measuring the rotation speed of the wheel can be used.

  By using these functions, the mobile robot 20 detects an obstacle ahead and stops, identifies the current position of its own device, or measures the distance to the subject to determine the location of the subject. Can be determined.

  Further, as shown in FIG. 5B, the drive unit 22 also changes the height and orientation of the head 20a using a motor. Thus, the face image of the subject can be captured at an appropriate position by changing the height and orientation of the head 20a. The drive unit 22 also has a shooting direction specifying function for determining the shooting direction of the camera 23 from the direction of the main body 20b and the direction of the head 20a.

  An arm 27 is provided on one side surface of the main body 20b. Further, a camera 27 a that captures the vicinity of the foot of the mobile robot 20 is attached to the tip of the arm 27. As shown in FIG. 5B, the drive unit 22 rotates the attachment portion (shoulder portion) of the arm 27 and the main body portion 20b, the attachment portion of the camera 27a, or extends and contracts the arm 27. A subject who is difficult to capture with the camera 23 (for example, a subject who has fallen on the ground) can be captured with the camera 27a.

  Returning to FIG. 3, the controller 25 will be described. The control unit 25 is a processing unit that executes processing such as drive control of the drive unit 22, appearance matching using a captured image by the camera 23, and face matching.

  The movement execution unit 25a is a processing unit that moves the own apparatus to the search range in accordance with the movement instruction received from the movement instruction unit 14d of the management apparatus 10. Specifically, the movement execution unit 25a uses the in-facility map information 26a stored in the storage unit 26 to specify the search range instructed from the movement instruction unit 14d. Here, the facility map information 26 a is the same information as the facility map information 15 c stored in the management device 10. And the movement execution part 25a drives the drive part 22, and moves an own apparatus to the specified search range.

  Further, when the own apparatus arrives at the search range, the movement execution unit 25a moves within the search range or changes the direction of the head 20a, and images the target person located within the search range with the camera 23. .

  The appearance collation unit 25 b is a processing unit that collates the image of the subject imaged within the search range with the appearance information received from the management apparatus 10. Thereby, even if it is a case where a some target person exists in the search range, the target person used as search object can be narrowed down from these some target persons.

  The face matching unit 25c detects the target person's face from the captured image when the appearance matching unit 25b detects the target person having appearance information that matches the appearance information received from the management apparatus 10 (highly coincident). A processing unit that detects information and collates with face information received from the management apparatus 10.

  Here, in order to accurately detect the target person's face information, the face matching unit 25c estimates the target person's face position from the appearance information of the target person detected by the camera 23, and approaches the subject apparatus to the target person. The position and orientation of the head 20a are adjusted according to the estimated face position.

  Further, the face matching unit 25c outputs a voice such as “Hello” from the speaker 24, draws the attention of the subject, and performs face matching of the subject. Thus, by performing a voice call to the subject before performing face matching, for example, even when the subject is turning his back against his / her device, the subject is turned around to face An image can be taken. Note that the mobile robot 20 may capture the face image by turning around in front of the target person instead of turning the target person around to capture the face image.

  In addition, the face matching unit 25 c transmits the matching result of face matching to the management apparatus 10 via the wireless communication unit 21. Here, the collation result transmitted to the management apparatus 10 includes collation contents indicating whether or not the face information of the subject matches the face information received from the management apparatus 10, the current position of the own apparatus, and the imaging of the camera 23. Directions and captured images are included.

  Next, specific operations of the management apparatus 10 and the mobile robot 20 according to the present embodiment will be described with reference to FIG. FIG. 8 is a sequence diagram illustrating a processing procedure executed by the management apparatus 10 and the mobile robot 20 according to the present embodiment.

  As shown in the figure, when the abnormality detection unit 14c detects an abnormality of the subject (step S101), the movement instruction unit 14d gives a movement instruction including the search range, the appearance information of the subject, and face information. (Step S102).

  Subsequently, when the mobile robot 20 moves to the search range in accordance with the movement instruction from the management device 10 (step S103), the appearance verification unit 25b detects appearance information having a high degree of coincidence with the appearance information included in the movement instruction. It is determined whether or not (step S104). In such processing, when the appearance information having a high degree of coincidence with the appearance information included in the movement instruction has not been detected (No in step S104), the movement execution unit 25a moves within the search range and is the target person to be searched. Is continuously searched (step S105).

  On the other hand, when the appearance information having a high degree of coincidence with the appearance information included in the movement instruction is detected (Yes in step S104), the face matching unit 25c estimates the face position of the subject (step S106), and the estimated face The position and orientation of the head 20a are adjusted according to the position (step S107). Then, the face matching unit 25c speaks to the subject (step S108), images the subject's face (step S109), and is included in the face information and the movement instruction detected from the captured image. The face information is collated (step S110), and the collation result is transmitted to the management apparatus 10 (step S111).

  Subsequently, in the management device 10, the verification result acquisition unit 14e notifies the administrator of information according to the verification result received from the mobile robot 20 using the display 12 or the speaker 13 (step S112), and ends the processing. To do. Here, an example of information notified to the administrator will be described with reference to FIG. FIG. 9 is a diagram illustrating a display example of the display 12.

  As shown in the figure, the collation result acquisition unit 14e uses the information included in the collation result from the mobile robot 20, and includes a dispatch request including an instruction to go to the target person to be searched. A screen is generated and displayed on the display 12. On this dispatch request screen, as shown in the figure, information such as the current position and direction of the mobile robot 20 included in the collation result is the search target subject (here, “Eiko Taro”). It is displayed as a discovery place. In addition, on the dispatch request screen, the face image and name of the target person to be searched, the target person ID, and the like are also displayed.

  Furthermore, as shown in the figure, an image of the subject imaged by the mobile robot 20 using the camera 23 or the camera 27a is also displayed on the dispatch request screen. As a result, the administrator can know the current state of the target person, and can take an appropriate response to the target person. For example, the administrator rushes to the scene when an image showing the subject is falling on the call-out request screen, but there is no abnormality when an image that looks out from the window is displayed. It can be judged.

  As described above, in the present embodiment, the appearance verification unit of the management device detects the appearance information of the target person using a fixed camera fixedly installed, and the movement instruction unit of the management device detects the appearance information detection result. Based on the above, the mobile robot is moved to the non-detection range of the fixed camera, and the face matching unit of the mobile robot detects the face information of the subject who is located in the non-detection range of the fixed camera. Therefore, it is possible to detect a subject who is located outside the acquisition range of the fixed information acquisition device while suppressing the number of fixed information acquisition devices fixedly installed in the area.

  In the above-described embodiment, the target person is narrowed down by detecting appearance information using the fixed cameras A to D, and the target person is accurately specified by detecting face information using the mobile robot 20. I've explained, but it's not limited to this.

  For example, appearance information may be detected using the fixed cameras A to D, and the appearance information may also be detected by the mobile robot 20. This is effective, for example, when it is difficult to register face information (when an unspecified number of subjects enter and exit the area). In addition, when the detection accuracy of the fixed cameras A to D is improved in the future and the face information of the subject who is far away can be detected, the face information may be detected by the fixed cameras A to D. Good. As described above, when the face information can be detected by the fixed cameras A to D, the target person may be specified using only the face information. Note that the frequency of detection of face information by the fixed cameras A to D can be increased by estimating a profile or the like from the face image (front image) of the subject registered in the storage unit 15.

  In the above-described embodiment, the case where the mobile robot 20 moves within the welfare facility only when instructed by the management apparatus 10 has been described. However, the mobile robot 20 does not receive an instruction from the management apparatus 10. It may also be used as a fixed camera. Hereinafter, such a case will be described with reference to FIG. FIG. 10 is a diagram for explaining a case where the mobile robot 20 is also used as a fixed camera.

  For example, as shown to (A) of the figure, when there is no instruction | indication from the management apparatus 10, the mobile robot 20 is good also as going around the welfare institution. Specifically, the mobile robot 20 circulates a pre-stored prescribed route (see (1) in the figure), and periodically transmits the circulation information to the management apparatus 10 (see (2) in the figure). ). Here, the traveling information is information including the current time, the current position, the camera direction, and the captured image. Then, the management device 10 performs the appearance verification of the target person in consideration of the patrol information received from the mobile camera (see (3) in the figure).

  In this way, by causing the mobile robot 20 to go around the welfare facility, the mobile robot 20 also plays a role as the fixed cameras A to D, so that the number of fixed cameras can be further suppressed. it can. In addition, due to the nature of the fixed cameras A to D that are used by being fixed to a wall or a ceiling, the mobile robot 20 is to be circulated mainly in places where imaging with the fixed cameras A to D is difficult. Thus, it is possible to further expand the detection range of the target person in the welfare facility. The mobile robot 20 may transmit only the captured image to the management apparatus 10, and the management apparatus 10 may generate the tour information including the captured image. In such a case, it is assumed that the management device 10 always knows the current position of the mobile robot 20 and the camera direction.

  Further, as shown in FIG. 5B, a plurality of mobile robots 20 may be installed in a passage in a welfare facility, and the appearance information of the target person may be detected using the mobile robot 20. When the management apparatus 10 detects an abnormality of the subject, if the mobile robot 20 installed in the place closest to the search range of the subject is instructed to move, the subject is more quickly It becomes possible to discover.

  Moreover, the management apparatus 10 may move the mobile robot 20 not only when the abnormality detection unit 14c detects an abnormality of the target person but also when searching for a specific target person. In such a case, the movement instruction unit 14d of the management apparatus 10 takes out the face information and appearance information of the target person to be searched from the registration information 15a and passes it to the mobile robot 20. Then, the mobile robot 20 searches for a specific target person according to a predetermined route. In addition, when the target person who seems to be a specific target person is detected by the fixed cameras A to D, the management apparatus 10 notifies the mobile robot 20 of the detection place of the target person, and moves the mobile robot 20 to the detection place. It may be directed.

  Further, in the above-described embodiment, a case has been described in which the mobile robot 20 performs the appearance matching and the face matching after imaging the subject using the camera 23 or the camera 27a. However, the present invention is not limited to this, and the management device 10 may perform appearance matching and face matching on an image captured by the mobile robot 20.

  In such a case, the mobile robot 20 does not include the appearance verification unit 25b and the face verification unit 25c, and the management device 10 may further include a face verification unit. Specifically, the mobile robot 20 transmits an image captured by the camera 23 or the camera 27a to the management device 10 via the wireless communication unit 21, and the management device 10 uses the appearance matching unit 14b for the captured image. Appearance matching is performed, and face matching is performed using a face matching unit.

  Heretofore, the case where the mobile robot 20 is an autonomous mobile robot has been described. However, the management device 10 or an administrator may remotely control the mobile robot 20. In such a case, the in-facility map information 26a may not be stored in the mobile robot 20. A camera 23 mounted on the mobile robot 20 may be provided to a security guard or the like. In such a case, the guard with the camera 23 plays a role as the mobile robot 20.

  In the past, the mobile robot 20 was used as an example of the mobile information acquisition device. However, the mobile information acquisition device is a portable terminal device such as a PDA (Personal Digital Assistant) held by a security guard or the like. There may be. In such a case, the movement instructing unit 14d of the management device 10 carries the image captured by the fixed camera and the fact that the abnormality has been detected based on the abnormality detection mode when the abnormality detection unit 14c detects the abnormality of the subject. Notify the terminal device.

  For example, as shown in FIG. 5B, when the abnormality detection unit 14c detects “abnormality 1” of the subject, that is, the subject is fixed within a predetermined time after the subject is detected by the fixed camera A. When the target person is not detected by the camera B, the movement instructing unit 14d notifies the PDA possessed by the security guard that the captured image by the fixed camera A and the abnormality of the target person have been detected.

  Further, as shown in FIG. 5C, when the abnormality detection unit 14c detects the “abnormality 2” of the target person, that is, after the target person is detected by the fixed camera A, the target is detected by the fixed camera B. When the time until the person is detected is earlier than usual, the movement instructing unit 14d notifies the PDA possessed by the security guard that the image captured by the fixed camera B and the abnormality of the subject have been detected.

  As a result, the security guard can check the captured image sent to the PDA and then go to the site where the abnormality occurred. The movement instructing unit 14d may notify the PDA of the content of the abnormality such as “abnormality 1” and “abnormality 2” as well as the image captured by the fixed camera and the fact that the abnormality of the subject has been detected. . In this way, by notifying the content of the abnormality, for example, even when a plurality of notifications that the abnormality has been detected are sent, the guard goes to the site where the abnormality occurs by confirming the content of the abnormality. The priority of the case can be determined.

  In this way, if an abnormality of the subject is detected, the mobile information acquisition device is configured to notify the PDA possessed by the security guard that the abnormality has occurred and cause the security guard to confirm the abnormality. As compared with the case where the mobile robot 20 is used, the cost required for system introduction can be reduced.

  The captured image notified from the movement instruction unit 14d to the PDA may be a captured image of a fixed camera used for detecting an abnormality of the subject. That is, the movement instruction unit 14d may notify the PDA of an image captured by the fixed camera B when the abnormality detection unit 14c detects “abnormality 1” of the subject, or the abnormality detection unit 14c When “abnormality 2” is detected, an image captured by the fixed camera A may be notified to the PDA. In addition, the movement instruction unit 14d may notify the PDA of the captured images of both the fixed camera A and the fixed camera B.

  By the way, although the Example mentioned above has demonstrated the case where the target person located indoors, such as a welfare facility, was detected, the target person detection method which concerns on this invention is not only indoors but the target person who exists outdoors. It can also be applied to. Therefore, in the following, such a case will be described with reference to FIG. FIG. 11 is a diagram for explaining a case where the subject detection method according to the present invention is applied outdoors. In the figure, as an example, the case of managing the behavior of the child from the elementary school through the school nursery school until returning home is shown.

  As shown in (A) of the figure, the registration camera is installed at the entrance of an elementary school, school nursery school, or home, and sends an image of children entering and leaving these buildings to a management device (not shown). To do. The fixed cameras A to E are fixedly installed in various places on the route from the elementary school to the home via the school nursery school, and transmit the captured images to a management device (not shown). The registration camera may also be used as a fixed camera.

  In addition, the mobile robot patrolls a place outside the detection range of these fixed cameras A to E, and transmits the captured image to a management device (not shown). In addition, as shown in the figure, it is good also as installing a camera in moving bodies, such as a motor vehicle, and transmitting the image imaged with this camera to the management apparatus. Thereby, the number of fixed cameras A to E can be further suppressed.

  Here, for example, if a child is not detected by the fixed camera D that should detect the child within a predetermined time after the child who is the subject is detected by the fixed camera C, the management device Detect abnormalities. Then, the management device passes the appearance information and face information of the child to the mobile robot, and is a path between the fixed camera C and the fixed camera D, and is not detected by the fixed camera C and the fixed camera D. Designate the range as the search range and move it. Thereby, the child located in the non-detection range of the fixed cameras A to E can be found by the mobile robot.

  As shown in FIG. 5B, when there are a plurality of search range candidates (a range in the right direction and a range in the left direction with respect to the fixed camera C), the movement instruction unit 14d What is necessary is just to determine the traveling direction of the subject from the captured image and to determine the search range according to the determined traveling direction.

  In addition, if it is detected by a fixed camera that the child has left the school with a person whose appearance information is not registered (suspicious person), it is determined that there is a high possibility that the child has taken away, The mobile robot may be moved immediately. In this case, if the appearance information of the suspicious person is registered, the location information of the suspicious person can be specified by detecting the appearance information of the suspicious person again by the fixed camera.

  Moreover, although only the outdoor has been described as the target area in the figure, indoor and outdoor can also be the target area. That is, it is also possible to detect a subject who has left the welfare facility with a fixed camera installed outdoors, and move the mobile robot to the outdoors according to the detection results of these fixed cameras.

  The subject detection method according to the present invention can also be applied when searching for disaster evacuees when a disaster occurs. For example, at present, only the identity of a disaster evacuee who has evacuated to an evacuation center can be confirmed. It is also possible to specify the location of the person.

  Until now, a camera that captures an image of a subject has been described as an example of an information acquisition device. However, the information acquisition device is not limited to a camera, for example, an IC tag reader that reads ID information of an IC tag. May be. Therefore, such a case will be described below with reference to FIG. FIG. 12 is a diagram for explaining a case where the mobile robot covers outside the reading range of the fixed IC tag reader.

  As shown in FIG. 3A, IC tag reader A and IC tag reader B are fixedly installed on the route in the area. The subject wears an IC tag in advance. Here, as shown to (A) of the figure, suppose that the subject was detected by IC tag reader A (refer (1) of the figure). That is, it is assumed that the IC tag reader A reads ID information of the IC tag worn by the subject.

  In such a case, the target person is not detected by the IC tag reader B that should detect the target person next, even though a predetermined time has elapsed since the target person was detected by the IC tag reader A ((2) in the figure). The management device detects the abnormality of the subject person (see (3) in the figure).

  When the abnormality of the subject is detected in this way, the management device passes the ID information of the IC tag worn by the subject to the mobile robot, and is a path between the IC tag reader A and the IC tag reader B, and The mobile robot is moved to the non-detection range of the IC tag reader A and the IC tag reader B. Then, as shown in (B) of the figure, the mobile robot discovers the target person within the search range based on the ID information passed from the management apparatus, and indicates that the target person to be searched has been found. It is transmitted to the management device together with the position.

  Further, an IC tag reader and a camera may be used together as an information acquisition device. For example, the management device stores the ID information of the IC tag in association with the face information and appearance information of the subject at the registration gate. When a fixed camera is installed on the route in the area, the management device detects an abnormality of the target person based on the detection result of the appearance information using the fixed camera, and the mobile robot detects the ID information of the IC tag. The target person may be specified based on the above.

  When an IC tag reader is installed on the route in the area, the management device detects an abnormality of the target person based on the ID information read result by the IC tag reader, and the mobile robot detects the target person based on the appearance information. What is necessary is just to narrow down or specify an object person based on face information.

  Up to now, the information for identifying an individual has been described using face information and IC tag ID information. However, the present invention is not limited to this. For example, biological information such as fingerprints, veins or irises can be used. Can be used.

  As described above, the target person detection system and the target person detection method according to the present invention are applicable not only when the target person is located within the acquisition range of the fixed information acquisition apparatus but also when the target person acquires the fixed information acquisition apparatus. This is useful when it is desired to detect the location and behavior of the subject, including when the subject is out of the range, and is particularly suitable for detecting a subject who is located indoors such as a welfare facility or hospital.

DESCRIPTION OF SYMBOLS 10 Management apparatus 11 Wireless communication part 12 Display 13 Speaker 14 Control part 14a Registration process part 14b Appearance collation part 14c Abnormality detection part 14d Movement instruction part 14e Collation result acquisition part 15 Storage part 15a Registration information 15b Collation history information 15c Facility map information DESCRIPTION OF SYMBOLS 20 Mobile robot 21 Wireless communication part 22 Drive part 23 Camera 24 Speaker 25 Control part 25a Movement execution part 25b Appearance collation part 25c Face collation part 26 Memory | storage part 26a In-facility map information

Claims (11)

A storage unit for storing information for identifying the target person as registration information;
A movable mobile information acquisition device for acquiring information about the subject;
A fixed information acquisition device that is fixedly installed and acquires information about the target person;
A collation unit that detects the target person by collating the information acquired by the fixed information acquisition device with the registration information;
An object detection system comprising: an abnormality detection unit that detects an abnormality of the object person based on a detection result of the object person by the collation unit. A movement instructing unit that gives an instruction to move the mobile information acquisition device to a non-acquisition range of the fixed information acquisition device when the abnormality of the subject is detected by the abnormality detection unit; The subject detection system according to claim 1. The fixed information acquisition device includes a first fixed information acquisition device and a second fixed information acquisition device installed along a predetermined route,
The abnormality detection unit
An abnormality of the subject is detected based on an elapsed time from the detection of the subject by the first fixed information acquisition device to the detection of the subject by the second fixed information acquisition device. ,
The movement instruction unit includes:
Based on the elapsed time out of the non-acquisition range between the first fixed information acquisition device and the second fixed information acquisition device when the abnormality of the subject is detected by the abnormality detection unit. The subject detection system according to claim 2, wherein an instruction to move the mobile information acquisition device to a range to be determined is given. The collation unit
Detecting the subject based on the information acquired by the fixed information acquisition device or the mobile information acquisition device when the movement instruction unit is not instructed to move the mobile information acquisition device. The subject detection system according to claim 2 or 3, characterized in that The mobile information acquisition device includes:
In addition to providing an information acquisition unit that acquires information about the target person, when there is no instruction by the movement instruction unit, to acquire information about the target person by traveling around a predetermined route,
The movement instruction unit includes:
The mobile information acquisition device is moved to a non-acquisition range of the fixed information acquisition device based on a detection result of the fixed information acquisition device and / or information on the subject acquired by the mobile information acquisition device. 5. The subject detection system according to claim 2, 3 or 4, wherein an instruction is given. The fixed information acquisition device or the mobile information acquisition device is:
An imaging unit that acquires an image of the subject as information about the subject,
The storage unit
Store the appearance information of the subject as the registration information,
The collation unit
The target person associated with the appearance information having a high degree of coincidence with the image is narrowed down as a detection target by collating the image of the subject acquired by the imaging unit with the appearance information stored in the storage unit. The target person detection system according to any one of claims 1 to 5. The storage unit
Further storing the subject's face information as the registration information,
The mobile information acquisition device includes:
An imaging unit that acquires an image of the subject as information about the subject,
The collation unit
The target person detection system according to claim 6, wherein the target person is specified by collating an image picked up by the image pickup unit with face information of the target person narrowed down as the detection target. The mobile information acquisition device is a mobile terminal device,
When the abnormality of the subject is detected by the abnormality detection unit, based on the detection mode of the abnormality, the information acquired by the fixed information acquisition device and the fact that the abnormality is detected to the portable terminal device The target person detection system according to claim 1, further comprising a notification unit for notifying. A storage step of storing information for identifying the target person as registration information;
A fixed information acquisition step of acquiring information about the subject using a fixed information acquisition device fixedly installed;
A collation step of detecting the target person by collating the information acquired in the fixed information acquisition step with the registration information;
An abnormality detection step for detecting an abnormality of the subject based on the detection result of the subject in the matching step;
A movement instruction for giving an instruction to move a movable mobile information acquisition device that acquires information about the subject to a non-acquisition range of the fixed information acquisition device when an abnormality of the subject is detected in the abnormality detection step And a mobile information acquisition step of acquiring information about the target using the mobile information acquisition device that has moved according to the instruction in the movement instruction step. A target person detection device used in a target person detection system for detecting the target person based on information about the target person acquired by a fixed information acquisition device installed in a fixed manner,
Based on the detection result of the target person, an abnormality detection unit that detects an abnormality of the target person,
An instruction to move the movable information acquisition device that acquires information about the subject to a non-acquisition range of the fixed information acquisition device when an abnormality of the subject is detected by the abnormality detection unit. An object detection device comprising: a movement instruction unit for performing the operation. A mobile information acquisition device used in a target person detection system that detects the target person based on information about the target person acquired by a fixed type information acquisition apparatus installed in a fixed manner,
A drive unit;
A movement execution unit that performs movement by driving the driving unit;
An information acquisition unit comprising: an information acquisition unit that acquires information about a target person who is located within a non-acquisition range of the fixed information acquisition device.

Images