JP2005086626A - Wide area monitoring device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a monitoring device displaying various pieces of information regarding a mobile object, selecting and switching information to be displayed according to situations, enhancing the monitoring efficiency of a watcher, and reducing mistakes. <P>SOLUTION: A video image picked up with a camera 10 is transmitted to a video processing part 12 and a position detection part 13 via a network 11. The video processing part 12 detects the information regarding the mobile object, the position detection part 13 detects a position of the mobile object, and these pieces of information are transmitted to a mobile object detection tracing part 14 and a map data part 15. The mobile object detection tracing part 14 selects image information required for processing and transmits the selected image information to a state judgment part 17 with the information of the map data part 15. The state judgment part 17 collates the image information with image information preliminarily registered in a data base 16 and attribute information and judges the tracing state of the mobile object, etc. A display switching part 18 selects information matching the situations on the basis of its result, and displays the selected information on a monitor 19. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a wide area monitoring apparatus that monitors a moving object such as a person over a wide range.

  2. Description of the Related Art Conventionally, there is known an automatic monitoring system that recognizes a suspicious vehicle or a suspicious person by detecting and tracking a moving object such as a vehicle or a person using a plurality of monitoring cameras. However, when these automatic monitoring systems are applied and a surveillance camera is installed outdoors and monitoring is performed over a wide area, there is a problem that is significantly different from that in a room with well-lit lighting conditions.

  For example, because the light exposure varies depending on the weather, season, and time, it is extremely difficult for the camera to capture the object to be monitored and automatically and accurately detect and track the suspicious person or object. Difficult, false detection and tracking mistakes become a big problem.

  A method of recognizing a specific person and a moving object by holding an ID tag or the like for each target object in order to recognize a moving object such as a vehicle or a person regardless of lighting conditions such as weather. However, since the monitoring target does not always hold an ID tag, it cannot be applied to a system for monitoring purposes.

  In view of this, when a plurality of cameras are installed in a monitoring system, automatic monitoring is difficult. Therefore, an image captured by a camera is configured to be displayed on a monitor corresponding to each camera. The monitor is often monitored. In such a case, the number of monitors will increase as the number of cameras installed increases. Therefore, it is necessary for the monitor to pay attention to each monitor, which increases the burden on the monitor and increases the monitoring. There was a possibility that mistakes also increased.

  As a conventional technology aimed at reducing the burden on observers, when an object such as an intruder enters the blind spot of the camera, the target person is displayed by displaying an image immediately before the object enters the blind spot. A video processing apparatus that displays that the camera has entered the blind spot is known (see, for example, Patent Document 1).

In addition, when a plurality of monitors are installed corresponding to a plurality of monitoring areas as other monitoring devices, the burden on the monitoring staff increases, and there is a moving body in one monitor in order to solve this problem. A device that simultaneously displays a position and an image of a moving object is also known. According to this, the monitoring person can know the moving object and its existence position at a glance by monitoring the map displayed on one monitor and the moving object on the map (see, for example, Patent Document 2). ).
Japanese Patent Laid-Open No. 2001-94968 (page 16, FIG. 6) JP 2000-295599 A (page 8, FIG. 5)

  However, in the conventional wide area monitoring device described in Patent Document 1, the method for determining whether the moving object has entered the blind spot is not specified as the method for determining whether the blind spot has been entered. Therefore, if there are a large number of unspecified persons or if the lighting conditions change, etc., which person in the large number of persons has entered the blind spot, and is not actually in the blind spot. Regardless, there was a possibility that it was determined that the blind spot was accidentally entered.

Further, in the conventional monitoring device described in Patent Document 2, when the monitoring area is wide and the number of moving objects is large, if the map of the entire monitoring area and all the moving objects are displayed simultaneously, the screen becomes complicated and difficult to understand. Therefore, it is necessary to take measures such as preparing a plurality of monitors by dividing the monitoring area into small scales, and there is a problem that the number of observers increases or the burden on the observers increases.

  The present invention has been made in view of the conventional problems as described above. In a monitoring area where an unspecified number of moving objects are present, the present invention is comprehensively considered based on various information related to moving objects and operations of monitoring personnel. It is an object of the present invention to provide a wide-area monitoring apparatus that makes it possible to appropriately select and switch information to be displayed in accordance with the situation, improve the monitoring efficiency of monitoring personnel, and reduce errors.

  The wide area monitoring apparatus of the present invention relates to a tracking target object included in an image, using the obtained information, situation determination means for judging the situation, and display switching means for switching necessary information based on the judgment of the situation judgment means And a monitor for displaying necessary information.

  With this configuration, it is possible to selectively display information necessary for the monitor according to the situation, and it is possible to improve monitoring efficiency and reduce monitoring errors.

  In the wide area monitoring apparatus according to the present invention, the obtained information is information included in a report from the outside or reliability information obtained by collating a tracking target object with a previously accumulated image.

  With this configuration, it is possible to selectively display necessary information to the monitoring staff according to the situation, including the notification of an emergency situation, and it is possible to improve monitoring efficiency and reduce monitoring errors can do.

  In the wide area monitoring device of the present invention, the display switching means switches the information included in the external notification to the information synthesized in advance when the external notification is input. is there.

  With this configuration, it is possible to selectively display information necessary for the monitor according to the situation using information included in the notification from the outside, and it is possible to improve monitoring efficiency and reduce monitoring errors.

  Moreover, the wide area monitoring apparatus of this invention is a display switching means, and switches to the information which asks a user confirmation, when reliability information is more than a predetermined value.

  With this configuration, since the user can confirm the reliability by the verification, it is possible to improve monitoring efficiency and reduce monitoring errors.

  Further, the wide area monitoring apparatus of the present invention is a display switching means for switching necessary information when the reliability information is not less than a predetermined value and the user approves.

  With this configuration, confirmation by the user is possible, so that monitoring efficiency can be improved and monitoring errors can be reduced.

  Further, the wide area monitoring device of the present invention provides the necessary information when the reliability information obtained by comparing the tracking target object approved by the user with the image stored in advance is less than a predetermined value by the display switching unit. Is to switch.

  With this configuration, it is possible to select information necessary to detect the tracking target object again, selectively display information necessary for monitoring to the monitoring staff, assist the monitoring staff, and improve monitoring efficiency. And monitoring errors can be reduced.

  Further, the wide area monitoring apparatus of the present invention includes map information and image information as necessary information, and reliability information obtained by comparing a tracking target object approved by a user with a previously accumulated image is included in the map information in advance. An escape prediction route is displayed in consideration of traffic conditions and transportation means in the vicinity of a position that is less than a predetermined value.

  With this configuration, it is possible to display in detail the place where the movement of the tracking target is predicted after the loss, assisting the operation of the observer, and reducing monitoring errors.

  Further, the wide area monitoring device of the present invention is configured such that the reliability information obtained by comparing the tracking target object approved by the user with the image accumulated in advance in the map information is reduced in scale according to the elapsed time when it is less than a predetermined value. The escape prediction route is changed and displayed.

  With this configuration, based on the time elapsed since the tracking object was lost, the mobility area can be expanded, and the map scale and escape prediction route can be changed accordingly. This makes it possible to reduce monitoring errors.

  In the wide area monitoring device of the present invention, the display switching means switches to information for displaying a plurality of images of the tracking target object in accordance with the reliability information obtained by comparing the tracking target object with the previously stored image.

  With this configuration, when collating with the database, the number of images to be displayed can be changed based on the reliability information of the collation processing performed by the system, and an apparatus collation error can be prevented. The monitoring error can be reduced by checking with the monitoring person whether or not it is the same person as the database registration image.

  The wide area monitoring apparatus according to the present invention detects and tracks a moving object from an image captured by at least one camera and an image captured by the imaging unit, and compares the detected target object with a pre-registered database. In addition, there is a situation determination means for associating the location of the target object using map information and determining the situation from a plurality of information including information on the tracking target object. It is equipped with a means to selectively display necessary information, and it can selectively display necessary information to the monitor according to the situation, enabling improvement of monitoring efficiency and reduction of monitoring errors. be able to.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(Embodiment 1)
FIG. 1 is a schematic configuration diagram of a wide area monitoring apparatus according to Embodiment 1 of the present invention.

  In the wide area monitoring apparatus according to the present embodiment, in order to set a wide area such as a building or factory premises or a street as a monitoring area, respective cameras are installed at a plurality of points and exist in the area. A moving object such as a person is detected and tracked and monitored.

In FIG. 1, cameras 10-1 to 10-n are installed at various points in order to monitor a wide area, and the cameras 10-1 to 10-n each hold position information installed. Each camera 10-
The images captured by 1 to n and the position information of the cameras 10-1 to 10-n are transmitted to the image processing unit 12 and the position detection unit 13 via the network 11. At this time, a server may be provided for each of the cameras 10-1 to 10-n, but a plurality of cameras 10-1 to 10-n may be collectively controlled by one server.

  The video processing unit 12 performs background difference or inter-frame difference on the transmitted image, and limits an area where moving objects such as a person and a vehicle exist. Therefore, for a limited area, M.I. Oren, C.I. Pageagegiou, P.A. Sinha, E .; Osuna and T. Poggio, “Pedestrian Detection using Wavelet Templates”, Proc. of CVPR97, pp. A moving object is detected using a person detection method such as 193-199, 1997 (reference document 3). Thereafter, the face area is detected and cut out based on the person detection result.

  Further, the position detection unit 13 assigns position information to the face image and the person image cut out by the video processing unit 12 based on the position information transmitted from the cameras 10-1 to 10-n.

  Information processed by the video processing unit 12 and the position detection unit 13 is transmitted to the moving object detection tracking unit 14.

  The moving object detection and tracking unit 14 assigns an index to the same person from the information obtained by the video processing unit 12 and the position detection unit 13 and the map data of the map data unit 15. The judgment of the same person is C.I. Wren, A.M. Azerbayejani, T.A. Darrel, A.M. Pentland, “Pfinder: Real Time Tracking of Human Body” IEEE Trans. PAMI, Vol. 19, NO. Using tracking methods such as 7, July, 1997, etc., all moving objects detected by the video processing unit 12 are tracked using the texture and color feature values of the clothing, and detected while tracking. Comprehensively performed by sequentially matching the face images.

  However, this does not limit the feature amount or tracking method for tracking. At this time, it is possible to match persons who are close to each other in a short time by using the relationship between the positional information of the captured cameras 10-1 to 10-n and time, thereby reducing tracking errors. It becomes possible to make it.

  Next, the situation determination unit 17 will be described with reference to FIG.

  The situation determination unit 17 includes a lost determination unit 171, a route search unit 172, a database processing unit 173, a notification processing unit 174, and an attribute processing unit 175.

  The lost determination unit 171 always operates while the moving object detection tracking unit 14 automatically tracks the moving object, and determines that the object to be tracked has been lost. When the moving object is a person, The lost determination is performed when tracking using the texture or color of clothes becomes impossible, or when matching of face images cannot be performed.

  In the present embodiment, the degree of similarity is determined by comparing the feature amount such as texture and color related to the tracking target for the past N frames with the feature amount such as texture and color of the tracking target in the current frame. If all frames are below the threshold, it is determined that tracking has failed.

Also, with regard to face image matching, a similarity is calculated by matching the face image to be tracked for the past N frames with the face image to be tracked in the current frame, and based on the tracking result and the face image matching result. Comprehensive lost judgment.

  For example, it is possible to make the lost determination by weighting the tracking result and the face image matching result and determining a threshold value. In addition, when it is determined that the tracking result is lost, the face image matching result can be used as an auxiliary.

  By the way, when such a lost determination is performed, in order to allow temporary occlusion, a determination process is made comprehensively from a plurality of frame images without performing a determination process for each frame.

  If the lost determination unit 171 determines that the object to be tracked has been lost, the route search unit 172 automatically determines the movable route of the tracked object based on the map data in the map data unit 15. The search result is output to the display switching unit 18. When the tracking target object is a person, a movable route including the nearest station is searched. At this time, the search is performed on the assumption that the moving speed of the person is an average moving speed, for example, 10 km / h from the time when the lost determination is performed. The search route is output to the display switching unit 18 and the moving object detection / tracking unit 14, and the moving object detection / tracking unit 14 assigns an index again based on the search route.

  Next, the database processing unit 173 operates when a face image is detected by the moving object tracking detection unit 14, and the face image of the person registered in the database 16 in advance and the face image of the person detected during monitoring. If the collation result is equal to or greater than the threshold, it is determined that the database 16 has been collated.

  Next, when an emergency such as an incident occurs and the notification processing unit 174 receives a notification such as the occurrence of an emergency via the network 21, the notification processing unit 174 displays the notification contents, the display switching unit 18, the database 16, and the moving object detection. Output to the tracking unit 14.

  If the tracking target object is, for example, a person, the attribute processing unit 175 receives tracking target object information such as a person's height, weight, name, clothes, and other characteristics. Each tracking target object information is stored in a database. If the information is registered in advance, the registered information can be received. If the information is not registered, information input by the operator via the I / F unit 20 can be received.

  Next, the operation of monitoring a moving object will be described with reference to FIG.

  Here, a case where four types of display modes are prepared in advance and the operation is performed in the display mode will be described. However, the display modes are not necessarily four types, and display modes for displaying other information are provided. Needless to say, it may be provided.

  The four display modes consist of a notification mode (mode 1), a verification mode (mode 2), an incident mode (mode 3), and a lost mode (mode 4). Is a mode that displays the content of the notification when there is a report, and the mode 2 verification mode is a mode that assists the operator in making an accurate tracking target determination when a tracking target person is found. The incident mode of mode 3 is a mode that assists the operator to quickly track and find the tracking target when the incident occurs, and the lost mode of mode 4 is lost to the operator when the tracking target is lost. In this mode, an operation is requested so that the detected object can be detected and tracked.

First, in S301, when the status determination unit 17 receives a report that the tracking target person has been found via the network 21, the process proceeds to S302, and when the report is not received, the process proceeds to S312.

  Next, in S302, the display switching unit 18 causes the monitor 19 to display the notification mode.

  Next, in S <b> 303, the notification processing unit 174 adds the notification information 42 to the database 16.

  Next, in S <b> 304, the moving object detection tracking unit 14 detects a tracking target person based on the notification information 42.

  Next, in S305, the database processing unit 173 compares the detected image 52 with the database image. If they match, the process proceeds to S306, and if they do not match, the process proceeds to S304.

  Next, in S306, the display switching unit 18 displays the collation mode on the monitor 19.

  Next, in S307, the supervisor confirms the detection image 52 and inputs a confirmation result from the I / F unit 20. If the detection image 52 and the tracking target person match, the process proceeds to S308, where the detection is performed. If the image 52 and the tracking target person do not match, the process proceeds to S304.

  Next, in S <b> 308, the display switching unit 18 displays the incident mode on the monitor 19.

  Next, in S309, when the situation determination unit 17 has lost the tracking target person, the process proceeds to S310, and during tracking, the process proceeds to S309.

  Next, in S310, the display switching unit 18 displays the lost mode on the monitor 19.

  Next, in S311, the moving object detection tracking unit 14 detects and tracks the tracking target person in consideration of the search route, and proceeds to S304.

  Next, in a case where the notification is not received, in S312, the moving object detection and tracking unit 14 detects the tracking target person, and when the tracking target person is detected, the process proceeds to S313. Migrate to

  Next, in S313, the database processing unit 173 collates the detected image 52 with the database image. If they match, the process proceeds to S314, and if they do not match, the process proceeds to S301.

  Next, in S314, the display switching unit 18 displays the collation mode on the monitor 19.

  Next, in S315, the supervisor confirms the detection image 52, inputs the confirmation result from the I / F unit 20, and if the detection image 52 and the tracking target person match, the process proceeds to S316, and the detection is performed. If the image 52 and the tracking target person do not match, the process proceeds to S301.

  Next, in S316, the display switching unit 18 displays the incident mode on the monitor 19.

  Next, in S317, when the situation determination unit 17 has lost the tracking target person, the process proceeds to S318, and during tracking, the process proceeds to S301.

  Next, in S318, the display switching unit 18 displays the lost mode on the monitor 19.

  Next, in S319, the moving object detection tracking unit 14 detects and tracks the tracking target person in consideration of the search route, and proceeds to S301.

  The above is an operation flow for monitoring a moving object.

  The database processing unit 173 of the situation determination unit 17 determines that the matching between the detected image 52 and the image stored in the database 16 is a collation if the matching result is equal to or greater than a predetermined threshold. .

  Here, the display of the notification mode (mode 1) is shown in FIG. When the captured image, the information indicating the characteristics of the person to be tracked, and the presence position are notified as the notification content, the notification image 41, the notification information 42, and the notification position map 43 indicating the presence position are displayed and displayed to the monitor. , Indicate the presence of the person being tracked. Here, the position of the tracking target person is indicated on the notification position map 43 using a star-shaped icon.

  Further, in the collation mode (mode 2), as shown in FIG. 5, the tracking target person image 51 registered in the database 16 and a plurality of actually detected detection images 52 are simultaneously displayed.

  Further, when tracking target person information 53 such as the height and weight of the tracking target person is registered in the database 16 in advance, the tracking target person information 53 can be displayed at the same time.

  Furthermore, it is possible to actually change the number of detected images 52 based on the reliability of the result of collation by the database processing unit 173. As the reliability, for example, the similarity between the detected image 52 and the tracking target person image registered in the database 16 can be used. Accordingly, when the reliability is low, it is possible to display more detection images 52, and the monitoring person determines whether or not the collation result is valid.

  Here, for the purpose of preventing misrecognition, the monitor uses the confirmation button 54 to determine whether or not the plurality of detected images 52 and the tracking target person image 51 in the database 16 coincide with each other. When NO of 54 is selected, a signal for continuing detection and tracking is output from the I / F unit 20 to the moving object detection tracking unit 14, and YES of the confirmation button 54 is selected again, or The detection image is displayed as long as the tracking target person can be tracked.

  If YES is selected on the confirmation button 54, it is determined that the person is a tracking target, and the fact is output from the I / F unit 20 to the attribute processing unit 175 to recognize that an incident has occurred, and the incident mode (mode) Process 3) is performed.

  Note that when the notification processing unit 174 receives a notification of occurrence of an emergency from the network 21, the incident mode of mode 3 can be set.

  In the case of an incident, the place where the incident occurred becomes an important issue. Therefore, in the incident mode (mode 3), as shown in FIG. 6, a tracking position map 61 and an address near the place where the incident occurred are displayed. And an icon indicating the position of the tracking target person is displayed on the map. At the same time, a representative image 62 is selected from the tracking camera images 63 taken during tracking and displayed. The representative image 62 may be selected by a monitor, or an image with high reliability of the collation result collated by the database processing unit 173 may be automatically selected.

It is also possible to display tracking target person information 64 such as height and weight held by the attribute processing unit 175. The tracking position map 61 can also display a movement locus of the tracking target. Here, it is indicated by a dotted line.

  Note that the tracking process is automatically performed, and the representative image 62 can be automatically updated when an image with higher reliability is detected. Thereby, the monitor can easily perform the operation for arranging the guard. In addition, by clicking on an icon on the map, the monitor can display the tracking camera image 63 that captures the vicinity, and can further select the cameras 10-1 to 10-n being tracked. Since this has a correspondence relationship between the position detection unit 13 and the map data unit 15, the cameras 10-1 to 10-n can be selected from the information of the position detection unit 13.

  Next, the display of the lost mode will be described with reference to FIG.

  The lost mode is determined to be lost from a plurality of images in order to allow the lost due to temporary occlusion or the like, and the display switching unit 18 displays the mode 4 lost mode.

  The display of the lost mode displays the lost position map 71 including information on buildings and houses in the vicinity of the lost area, and at the same time searches for an escapeable route based on the elapsed time from the time of the lost, and the lost position map 71 Display above. Here, the escapeable route is indicated by a dotted arrow.

  Further, since the scale of the lost position map 71 is determined based on the elapsed time and the escapeable route, the wide area is displayed as time passes after the lost. In addition, by simultaneously displaying the camera image 73 near the lost, the tracked person information 74 that has been lost, and the representative image 72 that was taken at the time of tracking, the person to be tracked is lost so as to find the object that the monitor has lost. It is possible to give accurate instructions to searchers nearby. At this time, if there is a nearby station, it is also displayed.

  Similarly, when a person who is being tracked automatically goes out of the monitoring area, the feature amount such as texture or color related to the track target for the past N frames and the feature amount such as texture or color of the track target in the current frame Since the degree of similarity is lowered, the lost determination unit 171 determines that it has been lost. At this time, the lost mode is displayed in the same manner. In addition, the nearest monitoring area is searched using the data of the map data unit 15 and displayed in different colors.

  Also, by displaying a monitoring area that can be moved from outside the monitoring area, it assists the operation of the monitoring person such as placing a person outside the monitoring area, and reduces monitoring errors.

  As described above, when the tracking target is a person, the situation determination unit 17 determines a situation such as an emergency call or lost from a plurality of data such as a person's face image, attribute data, and a position where the person exists, and the determination. By selecting and switching the necessary display according to the result, it is possible to realize efficient and accurate monitoring of the monitor and reduce monitoring errors. Also, it is possible to display information that is effective for operations such as the placement of guards.

(Embodiment 2)
FIG. 1 is a schematic configuration diagram of a wide area monitoring system according to the second embodiment.

  The wide area monitoring apparatus in the present embodiment is configured to install cameras at a plurality of points in order to set a wide area such as a street as a monitoring area, and detect, track and monitor vehicles existing in the area. is there.

  In FIG. 1, cameras 10-1 to n are installed at various points in order to monitor a wide area, and each camera 10-1 to 10-n holds position information. The images captured by the cameras 10-1 to n and the position information of the cameras 10-1 to 10-n are transmitted to the image processing unit 12 and the position detection unit 13 via the network 11. At this time, a server may be provided for each of the cameras 10-1 to 10-n, or a plurality of cameras 10-1 to 10-n may be collectively controlled by one server.

  The video processing unit 12 performs background difference or inter-frame difference on the transmitted image, and limits an area where moving objects such as a person and a vehicle exist. Therefore, a detection method such as Reference 3 described above is used for the limited region, or a plurality of vehicle templates are prepared in advance, and matching is performed using the templates. Thereafter, the number area is detected and cut out. The cut out number area is read here. Further, the position detection unit 13 adds position information to the vehicle image and number information cut out by the video processing unit 12 based on the position information transmitted from the cameras 10-1 to 10-n. Information processed by the video processing unit 12 and the position detection unit 13 is transmitted to the moving object detection tracking unit 14.

  The moving object detection and tracking unit 14 gives an index to the same vehicle from the information obtained by the video processing unit 12 and the position detection unit 13 and the map data of the map data unit 15. Judgment of the same vehicle is performed using the reading result of the number information. At this time, it is possible to match vehicles that are close to each other by using the relationship between the captured position information and time, and reduce tracking errors. Also, tracking errors are reduced by performing tracking processing using the color information of the detected vehicle and the type of vehicle template used at the time of detection, and further calculating the moving direction of the tracking target using the optical flow. Is also possible.

  Next, the situation determination unit 17 will be described with reference to FIG.

  The situation determination unit 17 includes a lost determination unit 171, a route search unit 172, a database processing unit 173, a notification processing unit 174, and an attribute processing unit 175.

  Similar to the first embodiment, the lost determination unit 171 and the database processing unit 173 operate when the moving object tracking detection unit 14 detects and tracks the target, and the lost determination of the tracking target and the detected license plate are performed. And the license plate held in the database 16 are collated.

  The lost determination unit 171 determines that the object to be tracked that is being automatically tracked by the moving object detection tracking unit 14 is lost, and the lost determination in the case of a vehicle basically cannot read and match the number. If you do.

  These lost determinations are not processed for each frame but are comprehensively determined from a plurality of frame images in order to allow lost due to temporary occlusion or the like. For example, in the past N frames, the read number and the current read number are collated, and if collation cannot be performed, it is determined that the lost number has been lost. If it is determined that the tracking target has been lost, the route search unit 172 searches for a movable route of the tracking target based on the map data in the map data unit 15. When the tracking target is a vehicle, a movable route is searched in consideration of the traffic situation in the vicinity of the lost point. At this time, the search range is expanded according to the time when the lost determination is performed.

The database processing unit 173 operates when the moving object detection unit 14 detects a license plate, and matches the vehicle number registered in the database 16 in advance with the vehicle number detected by the image processing unit during monitoring. Is to do. At this time, if the collation result is correct, it is determined that the database 16 has been collated.

  The notification processing unit 174 accepts operator input information via the I / F unit 20 or reports via the network 21 when an emergency such as an incident occurs and when the operator determines an emergency. Accept.

  When the tracking target is a vehicle, the attribute processing unit 175 receives tracking target object information such as the vehicle number, the vehicle type, and the color of the vehicle. Each tracking target object information is registered in the database 16 in advance. If there is, the information is accepted, and if not registered, the operator inputs via the I / F unit 20.

  Next, operations of the situation determination unit 17 and the display switching unit 18 will be described in detail.

  Here, the processing of the situation determination unit 17 when four types of display modes are prepared in advance will be described with reference to FIG. However, the display modes do not have to be four types, and display modes for displaying other information may be prepared.

  The four display modes consist of a notification mode (mode 1), a verification mode (mode 2), an incident mode (mode 3), and a lost mode (mode 4). Is a mode that displays the content of the notification when there is a notification, and the mode 2 verification mode is a mode that assists the operator in making an accurate tracking target vehicle determination when a tracking target vehicle is found. The incident mode of mode 3 is a mode for assisting the operator to quickly track and find the tracked vehicle when the incident occurs, and the lost mode of mode 4 is lost to the operator when the tracked vehicle is lost. In this mode, an operation is requested so that the detected object can be detected and tracked.

  Next, an operation flow for monitoring a moving object will be described with reference to FIG.

  First, in S801, when the situation determination unit 17 receives a report that the tracking target vehicle has been found via the network 21, the process proceeds to S802, and when the report is not received, the process proceeds to S804.

  Next, in S802, the display switching unit 18 causes the monitor 19 to display the notification mode.

  Next, in S <b> 803, the notification processing unit 174 adds the notification information 92 to the database 16.

  Next, in S804, the moving object detection tracking unit 14 detects the tracking target vehicle based on the report information 92.

  Next, in S805, the database processing unit 173 collates the detected image 1002 with the database image. If they match, the process proceeds to S806, and if they do not match, the process proceeds to S804.

  Next, in S806, the display switching unit 18 displays the collation mode on the monitor 19.

  Next, in S807, the supervisor confirms the detection image 1002, inputs the confirmation result from the I / F unit 20, and if the detection image 1002 matches the tracking target vehicle, the process proceeds to S808, and the detection If the image 1002 and the tracking target vehicle do not match, the process proceeds to S804.

  Next, in S808, the display switching unit 18 displays the incident mode on the monitor 19.

  Next, in S809, when the situation determination unit 17 has lost the tracking target vehicle, the process proceeds to S810, and during tracking, the process proceeds to S809.

  Next, in S810, the display switching unit 18 displays the lost mode on the monitor 19.

  Next, in S811, the moving object detection tracking unit 14 detects and tracks the tracking target vehicle in consideration of the search route, and the process proceeds to S804.

  Next, when there is no report, if the moving object detection tracking unit 14 detects a tracking target vehicle in S812, the process proceeds to S813. If not detected, the process proceeds to S801.

  Next, in S813, the database processing unit 173 collates the detected image 1002 and the database image. If they match, the process proceeds to S814, and if they do not match, the process proceeds to S801.

  Next, in S814, the display switching unit 18 displays the collation mode on the monitor 19.

  Next, in S815, the monitoring person confirms the detection image 1002, inputs the confirmation result from the I / F unit 20, and if the detection image 1002 matches the tracking target vehicle, the process proceeds to S816 to detect If the image 1002 and the tracking target vehicle do not match, the process proceeds to S801.

  Next, in S816, the display switching unit 18 displays the incident mode on the monitor 19.

  Next, in S817, when the situation determination unit 17 has lost the tracking target vehicle, the process proceeds to S818, and during tracking, the process proceeds to S801.

  Next, in S818, the display switching unit 18 displays the lost mode on the monitor 19.

  Next, in S819, the moving object detection tracking unit 14 detects and tracks the tracking target vehicle in consideration of the search route, and the process proceeds to S801.

  The above is an operation flow for monitoring a moving object.

  Here, FIG. 9 shows a display of the notification mode (mode 1). When the photographed image, the characteristic data of the target person, and the existing position are notified as the contents of the report, the report image 91, the report information 92, and the report position map 93 indicating the present position are displayed, and the monitoring target is displayed to the monitor. Indicates the presence of a vehicle. Here, the position of the vehicle is shown on the notification position map 93 using a star-shaped icon.

  Further, in the collation mode (mode 2), as shown in FIG. 10, the tracking target vehicle image 1001 registered in the database 16 and a plurality of actually detected detection images 1002 are simultaneously displayed.

  Further, when tracking target vehicle information 1003 such as a vehicle number, a vehicle type, a color, an owner, and a user of a tracking target vehicle is registered in the database 16 in advance, the tracking target vehicle information 1003 can be displayed at the same time. It is.

Furthermore, the number of detected images 1002 actually detected can be changed based on the reliability of the result of collation by the database processing unit 173. As the reliability, for example, the similarity between the detected image 1002 and the tracking target vehicle image 1001 registered in the database 16 can be used. As a result, when the reliability is low, more detection images 1002 can be displayed, and the monitoring person determines whether or not the collation result is appropriate.

  Here, for the purpose of preventing misrecognition, the monitor uses a confirmation button 1004 YES / NO to determine whether or not the detected plurality of detected images 1002 match the tracking target vehicle image 1001 in the database 16. When NO of the confirmation button 1004 is selected, a signal for continuing detection and tracking is output from the I / F unit 20 to the moving object detection tracking unit 14, and YES of the confirmation button 1004 is selected again. Alternatively, the detection image 1002 is displayed as long as the tracking target vehicle image 1001 can be tracked.

  If YES on the confirmation button 1004 is selected, it is determined that the vehicle is a tracking target vehicle, and the fact is output from the I / F unit 20 to the attribute processing unit 175 to recognize that the incident has occurred, and the incident mode (mode) Process 3) is performed.

  Note that when the notification processing unit 174 receives a notification of occurrence of an emergency from the network 21, the incident mode of mode 3 can be set.

  The incident mode corresponds to the case where the database processing unit 173 determines that the vehicle is a tracking target vehicle as described above and the case where the notification processing unit 174 in FIG. When an incident occurs, the incident mode is set as mode 3 and the incident mode is displayed by the display switch 18. When an incident occurs, the place where the incident occurred is an important issue. Therefore, in the incident mode, as shown in FIG. 11, a tracking position map 1101 and an address near the place where the incident occurred are displayed, and an icon is displayed on the tracking position map 1101 so that it can be seen that the vehicle is a vehicle. Here, it is a star icon. It is also possible to display a tracking image as a tracking camera image 1103. Further, a representative image 1102 selected from the detected image 1002 can be displayed.

  At the same time, the tracking target vehicle information 1104 such as the number information, the vehicle type and the color of the vehicle body held by the attribute processing unit 175 is also displayed. The tracking target vehicle information 1104 is data registered in advance in the database 16 or data input by the monitor via the I / F unit 20.

  It is also possible to display an icon on the tracking position map 1101 and further display a tracking locus of the tracking target. Here, the movement trajectory is indicated by a dotted line.

  Further, the representative image 1102 of the vehicle extracted by the video processing unit 12 is displayed. As a result, the monitor can easily perform an operation for arranging the guard.

  Next, the display mode (mode 4) at the time of occurrence of lost in FIG. 12 will be described. The occurrence of lost refers to the case where the tracking target vehicle is being detected and tracked, or the tracking target is lost while the tracking target vehicle is being tracked when the incident occurs. The lost determination unit 171 of the situation determination unit 17 determines whether or not the tracking target tracked by the moving object detection tracking unit 14 can be tracked.

Here, in order to allow lost due to temporary shielding or the like, the lost determination is performed from a plurality of images. When it is determined that it has been lost, the lost mode is displayed as mode 4. As shown in FIG. 12, the lost mode is displayed by searching for an escapeable route based on the lost position map 1201 including information on the buildings and houses in the vicinity that have been lost and the elapsed time from the time of the loss. It is displayed on the position map 1201. Here, the escape route is indicated by a dotted arrow.

  At this time, since the scale of the lost position map 1201 to be displayed is determined by the elapsed time and the escape route, the scale of the lost position map 1201 represents a wide area as time passes after the lost. Further, the camera image 1203 near the lost, the tracking target vehicle information 1204 of the lost target, and the representative image 1202 taken at the time of tracking are displayed at the same time, thereby urging the observer to discover the lost target.

  Similarly, when the vehicle being automatically tracked goes out of the monitoring area, the lost determination unit determines that the vehicle has been lost in order to fail the tracking process. At this time, the lost mode is displayed in the same manner. In addition, the nearest monitoring area is searched using the data of the map data section 15 and the monitoring areas for the cameras 10-1 to 10-n are displayed in different colors. . Also, by displaying a monitoring area that can be moved from outside the monitoring area, it assists the operation of the monitoring person such as placing a person outside the monitoring area, and reduces monitoring errors.

  As described above, when the monitoring target is a vehicle, the situation determination unit 17 determines a situation such as an emergency call or a loss of a plurality of data such as a vehicle number, a vehicle color, a vehicle type, and a position where a tracked vehicle is present. By selecting and switching the necessary display according to the determination result, efficient and accurate monitoring of the monitoring staff can be realized, and monitoring errors can be reduced. Also, it is possible to display information that is effective for operations such as the placement of guards.

  As described above, according to the wide area monitoring apparatus of the present invention, in order to collate information stored in the database with information obtained from the camera, information is selectively displayed so as to reduce the burden on the observer. In the event of an emergency, the occurrence location and the attributes of the tracking target are displayed, and tracking the target reduces the burden by facilitating the operation of the observer, or when the device loses the tracking target It is possible to prompt the supervisor to operate by notifying the supervisor of the occurrence. Thus, by judging the situation and displaying according to the situation, it is possible to improve the efficiency of monitoring and reduce monitoring errors.

Schematic configuration diagram of a wide area monitoring apparatus according to Embodiment 1 of the present invention Schematic configuration diagram of a situation determination unit used in the wide area monitoring apparatus implementation according to Embodiment 1 of the present invention The flowchart which shows operation | movement of the wide area monitoring apparatus in Embodiment 1 of this invention. The figure which shows the display of the report mode by the wide area monitoring apparatus in Embodiment 1 of this invention The figure which shows the display of the collation mode by the wide area monitoring apparatus in Embodiment 1 of this invention The figure which shows the display of the incident mode by the wide area monitoring apparatus in Embodiment 1 of this invention The figure which shows the display of the lost mode by the wide area monitoring apparatus in Embodiment 1 of this invention. The flowchart which shows operation | movement of the wide area monitoring apparatus in Embodiment 2 of this invention. The figure which shows the display of the report mode by the wide area monitoring apparatus in Embodiment 2 of this invention The figure which shows the display of the collation mode by the wide area monitoring apparatus in Embodiment 2 of this invention The figure which shows the display of the incident mode by the wide area monitoring apparatus in Embodiment 2 of this invention The figure which shows the display of the lost mode by the wide area monitoring apparatus in Embodiment 2 of this invention.

Explanation of symbols

10-1 to n Camera 11, 21 Network 12 Video processing unit 13 Position detection unit 14 Moving object detection tracking unit 15 Map data unit 16 Database 17 Situation determination unit 18 Display switching unit 19 Monitor 20 I / F unit 171 Lost determination unit 172 Route search unit 173 Database processing unit 174 Report processing unit 175 Attribute processing unit 41, 91 Report image 42, 92 Report information 43, 93 Report location map 51 Tracking target person image 52, 1002 Detection image 53, 64, 74 Tracking target person information 54, 1004 Confirmation button 61, 1101 Tracking position map 62, 72, 1102, 1202 Representative image 63, 1103 Tracking camera image 71, 1201 Lost position map 73, 1203 Camera image near lost 1001 Tracking target vehicle image 1003, 1104, 120 4 Tracking target vehicle information

Claims (9)

With respect to the tracking target object included in the image, using the obtained information, a situation determination means for judging the situation, a display switching means for switching necessary information based on the judgment of the situation judgment means, and the necessary information And a wide area monitoring device. The wide-area monitoring apparatus according to claim 1, wherein the obtained information is information included in an external report or reliability information obtained by comparing the tracking target object with a previously stored image. 3. The wide area according to claim 2, wherein the display switching unit is configured to switch to information obtained by combining information included in the external notification and information stored in advance when the external notification is input. Monitoring device. The wide-area monitoring device according to claim 2 or 3, wherein the display switching means is configured to switch to information for requesting confirmation from a user when the reliability information is equal to or greater than a predetermined value. The wide-area monitoring device according to any one of claims 2 to 4, wherein the display switching means is configured to switch necessary information when the reliability information is a predetermined value or more and the user approves. . 3. The display switching unit is configured to switch necessary information when a reliability obtained by comparing a tracking target object approved by a user with an image stored in advance is less than a predetermined value. Item 6. The wide area monitoring device according to any one of Items 5 to 6. The necessary information includes map information and image information, and the map information is a position where reliability information obtained by comparing a tracking target object approved by the user with a previously accumulated image is less than a predetermined value. The wide area monitoring apparatus according to claim 6, wherein an escape prediction route in consideration of a nearby traffic situation and means of transportation is displayed. The map information is displayed by changing the scale and escape prediction route according to the elapsed time when the reliability information obtained by comparing the tracking target object approved by the user with the previously accumulated image is less than a predetermined value. The wide-area monitoring apparatus according to claim 7. The wide-area monitoring device according to any one of claims 4 to 6, wherein the display switching unit switches to information for displaying a plurality of images of the tracking target object according to the reliability information.

Images