JP5151451B2 - Person identification system, person identification device, person identification method, and person identification program - Google Patents

Description

  The present invention relates to processing of images taken by a plurality of cameras, and more particularly to a technique for identifying a person from the taken images.

  With the face recognition technology, it is already possible to identify the person from the face photographed by the camera. Security systems applying such technology have been introduced into airports, stations, banks, and downtowns, for example. Face recognition systems are also used in commercial digital cameras and mobile phones, and are used for user authentication and focus position detection.

  As a technique for tracking or specifying a person photographed by a camera, for example, there are the following documents. Patent Document 1 discloses a technique of detecting the same person from images taken by a plurality of cameras and collecting flow line information of the person. Patent Document 2 discloses a method for tracking a specific person by detecting feature quantities indicating the same person from images taken by a plurality of autonomously operating cameras.

  Patent Document 3 discloses a method in which a plurality of autonomously operating cameras exchange information on the moving speed and direction of a moving body to track a specific moving body. Patent Document 4 discloses a technique for exchanging information on the position and position accuracy of moving objects from images taken by a plurality of cameras, and integrating the information when they are the same moving object. . Patent Document 5 discloses a technique for recording trajectory information of a user who has a card or the like.

JP 2000-200377 A JP 2004-072628 A JP 2006-086591 A Japanese Patent Laid-Open No. 2006-245795 Japanese Patent Laid-Open No. 11-027751

  Even if a video of a person is captured by the surveillance camera, it may be difficult to identify the person depending on conditions such as the installation location and angle of the surveillance camera. Although it is possible to extract an image of a face at a level that can identify a person using face recognition technology, it can be extracted only when the target person faces the front of the camera. When the target person faces in the other direction, the video cannot be extracted.

  To make it easier to identify a person, it is desirable to increase the number of surveillance cameras installed so that the target person is photographed from many angles for as long as possible. However, the larger the number of cameras installed, the larger the number of images that are shot and recorded. In order to find a video of a face that can identify a person from there, it is necessary for a human to view all of those videos, which requires enormous effort.

  In addition, the more cameras are installed, the higher the cost, and there is a problem that a large amount of video data is constantly transmitted and a load is imposed on the network. Incidentally, in Patent Documents 1 to 5, although it is possible to detect and track a person's movement from an image taken by a camera, the person is not specified therefrom.

  An object of the present invention is to provide a person identification system, a person identification device, and a person identification device that can easily extract a face image at a level that can identify a person from among a large amount of images in which it is difficult to identify a person with a practically possible number of cameras. The object is to provide a person identification method and a person identification program.

In order to achieve the above object, a person identification system according to the present invention is a person identification system including a plurality of cameras and a center processing device that receives video data captured by each camera. A video storage unit for storing video data received from a camera, a map processing unit for associating the stored video data with coordinate data on a map, and a plurality of video data and coordinate data on the map. The moving object on the image where the shooting area of the camera overlaps is recognized as the same moving object, and the moving locus detection engine for detecting the locus of the moving object, and the person from the moving object image on the locus detecting an image that reflected the face of the level of human identifiable, in that it has a face detection engine to determine whether the face is a face of a specific person previously registered And butterflies.

In order to achieve the above object, a person identification device according to the present invention is a person identification device that receives video data captured by a plurality of cameras, a video storage unit that stores video data received from the cameras, and a storage A moving body on the image of the portion where the imaging areas of the plurality of cameras overlap from the map processing unit for associating the captured image data with the coordinate data on the map, and the accumulated image data and the coordinate data on the map While recognizing as the same moving object, a moving locus detection engine for detecting the locus of the moving object, and detecting an image of a human face at a level that can identify a person from the moving object image on the locus , the face and having a face detection engine to determine whether the face of a specific person previously registered.

In order to achieve the above object, a person identifying method according to the present invention is a method for identifying a person photographed by a camera network comprising a plurality of cameras and a center processing device that receives video data photographed by the cameras. A video storage process for storing video data received from a camera, a map processing process for associating the stored video data with coordinate data on a map, and a plurality of stored video data and coordinate data on the map. The moving object on the image where the shooting area of the camera overlaps is recognized as the same moving object, and the moving locus detecting step for detecting the locus of the moving object, and the person from the moving object image on the locus detecting an image that reflected the face of the level of human identifiable, and a face detection process to determine whether the face is a face of a specific person previously registered And wherein the door.

In order to achieve the above object, a person identification program according to the present invention is a program for identifying a person photographed by a camera network comprising a plurality of cameras and a center processing device that receives video data photographed by the cameras. , Processing for storing video data received from the camera in a computer constituting the center processing apparatus, processing for associating the stored video data with coordinate data on the map, and stored video data and coordinates on the map From the data, the moving body on the image where the shooting areas of the plurality of cameras overlap is recognized as the same moving body, and the process of detecting the locus of the moving body and the moving object on the locus detecting an image that reflected the face of the level of human identifiable person from, determine whether the face is a face of a specific person previously registered Characterized in that for the process and to be executed.

  Since the present invention is configured to detect a human face at a level that can identify a person from the trajectory detected from the video data received from the camera as described above, a human views all of a huge amount of video. There is no need. This makes it possible to easily extract a face image of a level that can identify a person from a large amount of images where it is difficult to identify a person with a camera that can be installed in a realistic manner. A person identification device, a person identification method, and a person identification program can be provided.

  FIG. 1 is a block diagram showing a configuration of a camera network 1 according to an embodiment of the present invention. The camera network 1 is configured by connecting a stand-alone camera 100 (hereinafter referred to as camera 100), a dedicated terminal 200, and a center processing device 300 to each other. Although only two cameras 100 are shown in FIG. 1 due to space limitations, in reality, a larger number of cameras can exist.

  The cameras 100 perform short-range wireless communication between the cameras 100 and with the dedicated terminal 200. As the short-range wireless communication here, a known communication method such as IEEE 802.11 series, Bluetooth (registered trademark), or wireless USB can be applied. The connection between the dedicated terminal 200 and the center processing apparatus 300 is a connection by a wired network such as Ethernet (registered trademark) or a connection by short-range wireless communication as described above.

  FIG. 2 is a block diagram showing the configuration of the camera 100 shown in FIG. The camera 100 is referred to as a stand-alone type in the sense that it does not need to be supplied with electric power from another, and can autonomously shoot and perform operations associated therewith. It is not standalone in the sense that it does not connect to the network.

  Many cameras 100 are arranged in the monitoring target area. At this time, the camera 100 and the other camera 100 are arranged so that an overlapping area always exists between the shooting area of one camera 100 and the shooting area of another camera 100.

  The camera 100 includes a power generation mechanism unit 101, a video shooting unit 102, a video storage unit 103, a wireless transmission unit 104, a wireless reception unit 105, and a terminal authentication unit 106.

  The power generation mechanism unit 101 is a mechanism that generates power for operating the camera 100. As a power generation method, for example, solar power generation, fluorescent lamp power generation, wind power generation, or the like can be used. By providing the power generation mechanism 101, the camera 100 does not need to receive power supply from others, and there is an advantage that restrictions on the installation location are reduced. On the other hand, the camera 100 may not be provided with the power generation mechanism 101 if it is an installation location where power can be easily supplied from other sources, and may operate by receiving power supplied from others.

  The video photographing unit 102 is a mechanism for photographing a video and converting it into video data, which includes optical components such as a lens and a shutter, and a CCD element that converts an image into an electronic signal. The video storage unit 103 is a mechanism for storing the video shot by the video shooting unit 102 as video data, and is mainly configured by a storage device such as a flash memory or a hard disk.

  The wireless transmission unit 104 is a mechanism configured to transmit data to the outside by wireless communication, which includes a wireless module and software that drives the wireless module. When the camera 100 is in a normal operation state, the wireless transmission unit 104 transmits unique ID information 107 transmitted from a terminal authentication unit 106 described later to another camera 100 in the vicinity.

  When there is an instruction from the dedicated terminal 200, the wireless transmission unit 104 encrypts the video data stored in the video storage unit 103 with the unique ID information 107 described later and transmits the encrypted video data to the dedicated terminal 200. When the video data has been transmitted to the dedicated terminal 200, the video data stored in the video storage unit 103 is deleted.

  The wireless reception unit 105 is a mechanism configured to receive data from the outside through wireless communication, which includes a wireless module and software that drives the wireless module. The wireless module and software constituting the wireless reception unit 105 are usually the same as those in the wireless transmission unit 104, but here, for convenience of explanation, the wireless transmission unit 104 and the wireless reception unit 105 in FIG. Are described separately.

  When the camera 100 is in a normal operation state, the wireless reception unit 105 receives unique ID information 107 sent from another camera 100 in the vicinity. In addition, the instruction information from the dedicated terminal 200 is received after being decoded by the unique ID information 107. In any case, the wireless reception unit 105 transmits the received data to the terminal authentication unit 106 described later.

  The terminal authentication unit 106 includes a processor that controls the operation of each part of the camera 100 described above, software that operates on the processor, and the like. The terminal authentication unit 106 is attached with unique ID information 107, a timer 108 that issues an interrupt signal at a constant period, and an ID list 109 that is a list of cameras 100 that are present in the vicinity.

  When the unique ID information 107 is sent from the nearby camera 100, the terminal authenticating unit 106 records the unique ID information 107 and its reception time in the ID list 109. A newly installed camera 100 can participate in the camera network 1 at any time by adding its own unique ID information 107 to the existing camera 100 and adding to the ID list 109 on the existing camera 100. .

  In addition, the terminal authentication unit 106 sends the unique ID information 107 to the wireless transmission unit 104 every cycle when the timer 108 issues an interrupt signal. Then, for each cycle of the timer 108, it is checked whether there is any camera on the ID list 109 that has not received the unique ID information 107 within one cycle of the timer 108. If there is nothing, error information is generated and sent to the wireless transmission unit 104.

  When the wireless reception unit 105 receives instruction information from the dedicated terminal 200, the wireless reception unit 105 authenticates by decoding the instruction information. When the wireless reception unit 105 receives instruction information from the authorized dedicated terminal 200, the wireless reception unit 105 stores the instruction information stored in the video storage unit 103. The terminal authentication unit 106 is instructed to send the video data to the wireless transmission unit 104. If the instruction information is not from the regular dedicated terminal 200, the instruction information is discarded.

  The unique ID information 107 is an identifier that does not overlap among the individual cameras 100. The unique ID information 107 is transmitted to the camera 100 in the vicinity every time the timer 108 issues an interrupt signal. The unique ID information 107 is also used for encryption and decryption of data exchanged with the dedicated terminal 200 when the video data stored in the video storage unit 103 is transmitted to the dedicated terminal 200.

  The timer 108 issues the interrupt signal described above to the terminal authentication unit 106, and at the same time holds the time synchronized among all the cameras 100. Time synchronization is possible by acquiring time data from a GPS satellite or an NTP server, for example. The unique ID information 107 and the time held by the timer 108 are stored in the video storage unit 103 together with the video shot by the video shooting unit 102.

  Note that the reason why wireless communication is performed by the wireless transmission unit 104 and the wireless reception unit 105 is to reduce restrictions on the installation location of the camera 100 as in the power generation mechanism unit 101 described above. When the camera 100 is installed in a place where the camera 100 can be connected to the center processing apparatus 300 via a wired network, the wireless transmission unit 104 and the wireless reception unit 105 may be replaced with a transmission unit and a wireless reception unit that use wired network communication, respectively.

  Returning to FIG. 1, the dedicated terminal 200 performs short-range wireless communication with the camera 100, and performs communication between the dedicated terminal 200 and the center processing device 300 through a wired network or short-range wireless communication connection. . Data received from the camera 100 is sent to the center processing device 300 as it is, and conversely, data transmitted from the center processing device 300 is sent to each camera 100 as it is. The dedicated terminal 200 can be connected simultaneously with the plurality of cameras 100. A plurality of dedicated terminals 200 may exist.

  The short-range wireless communication of the camera 100 has restrictions on the reach of radio waves. Therefore, in order to enable the camera network 1 to monitor a wider range than the reach of radio waves when the plurality of cameras 100 and the center processing device 300 communicate directly, the camera 100 and the center are connected via the dedicated terminal 200. The processing device 300 communicates. If a range narrower than the reach of short-range wireless communication radio waves is set as the monitoring target range, the plurality of cameras 100 and the center processing apparatus 300 may directly communicate with each other.

  The center processing device 300 is a computer device configured by a server or a personal computer, and may be physically one computer or a plurality of computers connected to each other. The center processing device 300 includes a video storage unit 301, a video playback unit 302, a map processing unit 303, a face detection engine 304, a movement locus detection engine 305, a control unit 306, and an encryption / decryption unit 308. The control unit 306 is a function that executes software, which includes a CPU, a memory, and an OS. The control unit 306 is attached with a timer 307 that issues an interrupt signal at a constant cycle.

  The video storage unit 301 is a mechanism for storing video collected from each camera 100, and is mainly configured by a large-capacity storage device such as a hard disk. The video playback unit 302 is a mechanism for playing back video stored in the video storage unit 301, and includes a display, hardware for driving the display, and software.

  The map processing unit 303 is a function executed by the control unit 306, converts the unique ID information 107 sent from each camera 100 into position information on the map, and converts the video of the camera to a unique map on the map. Position to the place. The map processing unit 303 holds a coordinate table 311 that associates the unique ID information 107 with the coordinates on each map.

  The face detection engine 304 is a function executed by the control unit 306 and detects a human face from the video stored in the video storage unit 301. It is also possible to register a plurality of human faces in the face detection engine 304 in advance, and when the detected faces match the registered human faces, information about the persons can be presented to the user. is there.

  The movement trajectory detection engine 305 is a function executed by the control unit 306, and recognizes a moving body reflected in a portion where the imaging areas of the images of the plurality of cameras 100 overlap as the same moving body, and the same movement trajectory. Track as a moving object. Note that the processing used when the face detection engine 304 detects a human face and the processing used by the movement trajectory detection engine 305 to detect the trajectory of a moving object are already known from the above-mentioned patent documents. I will not mention it in detail.

  The encryption / decryption unit 308 encrypts the instruction information to be transmitted to the camera 100 with the unique ID information 107 of each camera 100 that is a transmission destination. Also, the video data sent from the camera 100 is decoded by the unique ID information 107 of each camera 100 that is the transmission source. The instruction information transmitted by the center processing apparatus 300 and the received video data are all encrypted / decrypted through the encryption / decryption unit 308.

  FIG. 3 is a flowchart showing processing executed in the camera 100 shown in FIGS. 1 and 2. When the process is started, the terminal authentication unit 106 waits for an interrupt signal from the timer 108 (step S401). If there is an interrupt signal, the process proceeds to step S406 and later. If there is no interrupt signal, the terminal authentication unit 106 checks whether or not the unique ID information 107 is sent from another camera 100 in the vicinity (step S402). If it is sent, the unique ID information 107 is added to the ID list 109. (Step S403), the video data stored in the video storage unit 103 is deleted, and the process proceeds to step S404. If not, the process proceeds to step S404 as it is.

  Subsequently, the terminal authentication unit 106 checks whether or not the instruction information from the dedicated terminal 200 has been received via the wireless reception unit 105 (step S404), and if received, the terminal authentication unit 106 stores it in the video storage unit 103. The wireless transmission unit 104 is instructed to transmit the recorded video data to the dedicated terminal 200 (step S405), and the process returns to step S401. If the instruction information has not been received, the process directly returns to step S401.

  If there is an interrupt signal in step S401, the terminal authentication unit 106 transmits the unique ID information 107 to the other camera 100 in the vicinity (step S406), and the unique authentication information from the other camera 100 stored in the ID list 109 is obtained. The ID information 107 is checked (step S407). If all the cameras 100 in the ID list 109 have received and recorded the unique ID information 107 within one cycle of the timer 108, it is determined that there is no abnormality, and the process returns to step S401.

  In step S407, if there is even one camera 100 for which the reception record of the unique ID information 107 does not remain, it is determined that an abnormality has occurred, and an alert is issued to all the cameras 100 and the dedicated terminal 200 belonging to the camera network 1. Transmit (step S408), and the process ends. As a result, it is possible to detect an anomaly that has occurred due to a failure or vandalism.

  FIG. 4 is a flowchart showing the process executed in the center processing apparatus 300 shown in FIG. When the process is started, the control unit 306 waits for an interrupt signal from the timer 307 (step S501). When there is an interrupt signal, the control unit 306 transmits instruction information for transmitting the captured video to all the cameras 100 belonging to the camera network 1 (step S502).

  Upon receiving the instruction information, the camera 100 transmits the video data to the center processing device 300 together with the unique ID information 107 and the time held by the timer 108. The control unit 306 records the received video data together with the unique ID information 107 and the time information in the video storage unit 301 (step S503), and proceeds to step S504. At that time, the map processing unit 303 updates the coordinate table 311 to associate the unique ID information 107 of the camera 100 with the coordinates on the map. If there is no interrupt signal, the process proceeds to step S504.

  Note that the period of the timer 307 is not particularly related to the period of the timer 108 in the camera 100. Depending on the capacity of the video storage unit 103 of the camera 100, any period may be used as long as video data can be collected and recorded before the video storage unit 103 becomes full.

  The communication with each camera 100 is encrypted based on each unique ID information 107. Therefore, even if communication between the camera 100 and the center processing device 300 is intercepted by another device, a device other than the center processing device 300 causes the camera 100 to transmit video data, or the center processing device 300 The transmitted video data cannot be played back. Note that a publicly known method can be applied as an encryption method, and data other than the unique ID information 107 can be applied as an encryption key.

  If there is no interrupt signal from the timer 307 in step S501, the control unit 306 waits for an operation from the user (step S504). The user can operate the center processing device 300 to play back video at an arbitrary point and time in the monitoring target area and time of the camera network 1 using the video playback unit 302.

  In the center processing device 300, the coordinate table 311 associates the unique ID information 107 of the camera 100 with the coordinates on the map. In addition, since the time information of the videos from the respective cameras 100 is synchronized, the videos at a specific time can be reproduced in synchronization with the plurality of cameras 100. These reproductions are not particularly described in FIG.

  Therefore, the user who operates the center processing apparatus 300 can see how a person at a specific time and point is captured by the plurality of cameras 100. At the same time, by synchronizing the plurality of cameras 100 and performing playback operations such as rewind, fast forward, slow motion and frame advance, the movement of a person at a specific time and point is tracked by the images of the plurality of cameras. It becomes possible.

  FIG. 5 is a conceptual diagram showing detection of a person's movement locus and identification of a person in the camera network 1 shown in FIG. A user operates the center processing device 300 to watch videos taken by a plurality of cameras 100 installed in the monitoring target area. Now, the user pays attention to the person image 701 photographed by the specific camera 100b.

  However, since the information such as the face, the body shape, and the clothes is insufficient with only the person image 701 at a specific time and point, the user cannot specify the person. In addition, it is difficult for the user to view all the images taken by the large number of cameras 100. Therefore, the user designates the person image 701 as a target person and issues a person specifying command for causing the center processing apparatus 300 to specify a person.

  Upon receiving the person specifying command, the control unit 306 uses the movement trajectory detection engine 305 to link the images captured by the plurality of cameras 100 using the coordinate table 311 and detect the movement trajectory 702 of the person corresponding to the person image 701. The video necessary for specifying the person is extracted (step S505).

  From the video corresponding to the movement locus 702 specified in step S505 (the person is shown), the face detection engine 304 detects whether or not there is a video showing a face that can specify the person. If it exists, the person image 701b photographed by the specific camera 100c is detected and displayed as an image showing the face (step S507).

  Furthermore, when the face shown in the detected person image 701b is a face of a specific person registered in the face detection engine 304, information about the person is presented to the user (steps S508 to 509), and step Return to S501. For example, if faces such as wanted criminals and missing persons are registered in the face detection engine 304, the person is detected and the movement trajectory 702 is recorded, which is extremely useful information in the investigation.

  If there is no face image at a level that can identify a person in step S506, or if it is not a registered person face in step S508, the process returns to step S501.

  Note that the processing by the movement trajectory detection engine 305 and the face detection engine 304 may be terminated when a person image 701b that can identify a person is detected from the movement trajectory 702 in steps S506 to S507. Moreover, you may make it continue the process by them and detect a several person image with respect to the same object person.

  This makes it possible to detect a person image at a level that can identify a person without viewing all the images taken by a large number of cameras 100. For example, when a crime occurs in a monitoring target area of the camera network 1, etc. This makes it easier to identify people. Furthermore, since the movement trajectory of a person can be detected and recorded, it is also useful for investigating wanted persons and missing persons.

  Further, since it is only necessary that the image capturing areas of the plurality of cameras 100 be arranged so that there is an overlapping portion somewhere, it is not necessary to install more cameras 100 than that. Therefore, it is possible to reduce the cost of installing the camera 100 more than necessary and the load on the network associated with transmitting a huge amount of video data.

  Since the camera 100 is a stand-alone type, there are few restrictions on the installation location, and the camera 100 can be installed in a place where connection with a power source and a wired network is impossible. In addition, even when there are cameras whose functions are stopped due to disasters or vandalism, they are detected and an alert is issued, so that the user can respond immediately.

  Although the present invention has been described with the specific embodiments shown in the drawings, the present invention is not limited to the embodiments shown in the drawings, and is known so far as long as the effects of the present invention are achieved. It goes without saying that any configuration can be adopted.

  It can be used in security systems with surveillance cameras installed in airports, stations, banks, downtown areas, stores, etc.

It is a block diagram which shows the structure of the camera network which concerns on embodiment of this invention. It is a block diagram which shows the structure of the camera shown in FIG. FIG. 3 is a flowchart showing processing executed in the camera shown in FIGS. 1 and 2. FIG. It is a flowchart showing the process performed in the center processing apparatus shown by FIG. It is a conceptual diagram which shows the detection of the movement locus | trajectory of a person in the camera network shown in FIG. 1, and identification of a person.

Explanation of symbols

1 Camera network (person identification system)
DESCRIPTION OF SYMBOLS 100 Camera 101 Power generation mechanism part 102 Image | video imaging | photography part 103 Image | video storage part 104 Wireless transmission part 105 Wireless reception part 106 Terminal authentication part 107 Unique ID information 108 Timer 109 ID list 200 Dedicated terminal 300 Center processing apparatus (person identification apparatus)
DESCRIPTION OF SYMBOLS 301 Image | video storage part 302 Image | video reproduction | regeneration part 303 Map processing part 304 Face detection engine 305 Movement locus detection engine 306 Control part 307 Timer 308 Encryption / decryption part 701 Person image 702 Movement locus

Claims (18)

A person identification system comprising a plurality of cameras and a center processing device that receives video data captured by each of the cameras,
The center processing device is
A video storage unit for storing video data received from the camera;
A map processing unit associating the accumulated video data with coordinate data on a map;
From the accumulated video data and the coordinate data on the map, the moving body on the video of the portion where the imaging areas of the plurality of cameras overlap is recognized as the same moving body, and the locus of the moving body is determined. A moving locus detection engine to detect,
Face said detected video images reflected the face of the level of human identifiable person from among the moving body of the image on the trajectory, to determine whether the face is a face of a specific person previously registered A person identifying system comprising: a detection engine.   The person specifying system according to claim 1, wherein the center processing device includes a timer that generates an interrupt signal for causing the plurality of cameras to transmit instruction information for transmitting the captured video data. . Each of the plurality of cameras has unique ID information corresponding to itself.
The person identifying system according to claim 2, wherein the center processing apparatus includes an encryption unit that encrypts the instruction information based on the unique ID information and transmits the encrypted instruction information.   The said center processing apparatus has a decoding part which decodes the video data received from these several cameras based on the said specific ID information, and preserve | saves it in the said video storage part, The characterized by the above-mentioned. Person identification system.   The person specifying system according to claim 2, wherein the center processing apparatus includes a video playback unit that plays back the stored video data.   3. The face detection engine according to claim 2, wherein the face detection engine can determine whether or not a face included in the detected video matches a face of a specific person registered in advance. Person identification system.   The person identification system according to claim 2, wherein the plurality of cameras and the center processing apparatus are connected by wireless communication.   The person identification system according to claim 2, wherein the plurality of cameras and a dedicated terminal are connected by wireless communication, and the dedicated terminal and the center processing apparatus are connected by a network.   The person identification system according to claim 2, wherein each of the plurality of cameras has a power generation mechanism unit that generates power for operating the camera.   3. The plurality of cameras each exchange unique ID information corresponding to each other, and send an alert when the unique ID information is not received from the cameras in the vicinity. Person identification system. A person identification device that receives video data captured by a plurality of cameras,
A video storage unit for storing video data received from the camera;
A map processing unit associating the accumulated video data with coordinate data on a map;
From the accumulated video data and the coordinate data on the map, the moving body on the video of the portion where the imaging areas of the plurality of cameras overlap is recognized as the same moving body, and the locus of the moving body is determined. A moving locus detection engine to detect,
Face said detected video images reflected the face of the level of human identifiable person from among the moving body of the image on the trajectory, to determine whether the face is a face of a specific person previously registered A person identification device comprising: a detection engine.   The person identifying apparatus according to claim 11, further comprising a timer that generates an interrupt signal for causing the plurality of cameras to transmit instruction information for transmitting the captured video data.   The person identifying apparatus according to claim 12, further comprising: an encryption unit configured to encrypt the instruction information based on unique ID information corresponding to each of the plurality of cameras and then transmit the encrypted instruction information.   14. The person identifying apparatus according to claim 13, further comprising: a decoding unit that decodes video data received from the plurality of cameras based on the unique ID information and stores the decoded data in the video storage unit.   The person identifying apparatus according to claim 12, further comprising a video playback unit that plays back the stored video data.   13. The face detection engine according to claim 12, wherein the face detection engine can determine whether a face included in the detected video matches a face of a specific person registered in advance. Person identification device. A method for identifying a person photographed by a camera network comprising a plurality of cameras and a center processing device that receives video data photographed by the cameras,
A video storage step of storing video data received from the camera;
A map processing step for associating the accumulated video data with coordinate data on a map;
From the accumulated video data and the coordinate data on the map, the moving body on the video of the portion where the imaging areas of the plurality of cameras overlap is recognized as the same moving body, and the locus of the moving body is determined. A movement trajectory detection step to detect,
Face said detected video images reflected the face of the level of human identifiable person from among the moving body of the image on the trajectory, to determine whether the face is a face of a specific person previously registered And a detection step. A program for identifying a person photographed by a camera network composed of a plurality of cameras and a center processing device that receives video data photographed by the camera, the computer constituting the center processing device,
Processing to store video data received from the camera;
A process of associating the accumulated video data with coordinate data on a map;
From the accumulated video data and the coordinate data on the map, the moving body on the video of the portion where the imaging areas of the plurality of cameras overlap is recognized as the same moving body, and the locus of the moving body is determined. Processing to detect,
Processing said detected image that reflected the face of the level of human identifiable person from among the moving body of the image on the trajectory, to determine whether the face is a face of a specific person previously registered A person identification program characterized by causing

Images