JP4722537B2 - Monitoring device - Google Patents

Description

  The present invention relates to a monitoring apparatus that performs monitoring using a plurality of cameras.

  In recent years, a monitoring system has been put into practical use in which cameras are installed at various locations in a facility such as a building and the video captured by these cameras is centrally monitored at one location. However, when a plurality of captured images are displayed on a plurality of monitors and monitored simultaneously, the monitor often overlooks an abnormality such as an intruder intrusion. Even when an intruder is found on the screen, it is difficult to intuitively understand where the images of multiple monitors are reflected in the building. is there. This tendency increases in proportion to the size of the building and the number of cameras. In response to this situation, measures are being taken to demand a burden on people, such as increasing the number of observers and instructing observers on the structure of the building.

  As a means for reducing such a human burden, there is a monitoring system disclosed in Patent Document 1. When this surveillance system captures a surveillance site with a plurality of cameras and detects an intruder from the video, the video output of the camera that has detected the intruder is displayed on a single monitor, and the intruder faces from the image. A direction is predicted, another camera installed within a certain range centered on the direction is selected, and the video is displayed on another monitor. Therefore, since the video of the camera that detected the intruder and the video of the camera that is predicted to show the intruder are selected and presented to the monitor, the monitor can easily grasp the intruder. it can.

  Further, Patent Document 2 discloses a monitoring system that can quickly reproduce only a portion where a suspicious person is shown from an enormous amount of recorded video of a monitoring camera. In this monitoring system, when a suspicious person is detected by an entrance camera, an abstract image that abstracts the characteristics of the person is generated and stored together with a person ID that identifies the suspicious person. When a suspicious person with the characteristic is detected from the image of the camera in the facility, the suspicious person appears in the field of view of the camera together with the person ID in the monitor log database and the suspicious person disappears from the field of view. The “end time” is recorded for each camera. At the time of reproduction, the retrieval system extracts and reproduces the captured images of each camera within the range specified by the “start time” and “end time”. Therefore, in this monitoring system, since the video showing the suspicious person is extracted from the recorded image and presented, the burden on the monitoring person is reduced.

Japanese Patent No. 9-016863 JP 2004-187115 A

  However, even in the monitoring system of Patent Document 1, since the captured images from the respective cameras are displayed on separate monitors, it is difficult for the supervisor to intuitively recognize which part of the facility the monitor video is captured. is there. In addition, since the movements of intruders that can be displayed by individual surveillance cameras are limited, there is a possibility that errors or oversights may occur when predicting the movements of intruders based on the images of individual surveillance cameras. .

  When an intruder is specified from an image of an individual surveillance camera as in the surveillance system of Patent Document 2, the intruder's position depends on the shooting position and angle of the camera or the orientation of the face when passing through the camera. Features may not be captured sufficiently, which may lead to cases where an intruder is overlooked or a person who is not an intruder is detected, and many false images may be presented to the observer. As a result, the burden on the observer is not reduced.

  The present invention has been made in view of the above circumstances, and provides an image that allows the monitor to intuitively know the position of the monitoring target and to easily grasp the abnormal state such as the movement and characteristics of the suspicious person. The object is to provide a monitoring device that can be presented.

In the present invention, a monitoring apparatus that performs monitoring using a plurality of cameras, the composite image generating means for generating a composite image using a camera image captured by the camera, and the camera image and the composite image are held. When a recording unit, a display unit for displaying an image, and an arbitrary region of the composite image displayed on the display unit are selected, the camera image corresponding to the selected region or the region of the camera image is searched. Tag information in the search means displayed on the display means, motion detection means for detecting motion from the composite image generated using a plurality of camera images, and tag information in the composite image detected by the motion detection means. Tag information adding means for adding, and searching for the camera image used in the combined image to which the tag information is added, and tag information transmission for adding the tag information to the camera image. Those comprising a means. According to the present invention, the supervisor can intuitively know the position of the monitoring target such as various facilities from the composite image, and selects a desired region from the composite image to view a detailed camera image. Can do. Further, by detecting a motion from a composite image including a plurality of camera images, a motion that is not clear in the analysis of a single camera image can be detected accurately and stably.

  The monitoring apparatus of the present invention further includes an erasing unit that determines the priority of the composite image and the camera image to which the tag information is added based on the tag information, and determines an image to be deleted based on the priority. Including things. According to the present invention, it is possible to effectively utilize the storage capacity of the apparatus by automatically erasing an image having a low priority.

  The monitoring apparatus according to the present invention can present a monitor with a composite image that allows the user to know at a glance where the facility is being monitored. By searching, a detailed camera image can be presented. Therefore, the supervisor can monitor the inside of the facility in detail and efficiently.

  In addition, the monitoring device of the present invention can stably and accurately detect an unclear movement on an image simply by analyzing individual camera images, and tag information assigned to an image based on the detection result. The playback image can be accurately narrowed down using.

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

(First embodiment)
FIG. 1 shows a schematic configuration of a monitoring apparatus according to the first embodiment of the present invention. The monitoring device of FIG. 1 monitors facilities such as buildings, and includes cameras 11 to 13, a monitoring image control unit 1, a data holding unit 2, an image composition unit 3, a display unit 4, and a search / area designation unit 5. , A time code generation unit 6, a motion detection unit 7, and a recording unit 9.

  The cameras 11 to 13 are installed in the facility and shoot the facility. FIG. 2 shows an example of the arrangement of the cameras 11 to 13. In FIG. 2, the cameras 11 to 13 are installed in almost the same position in the facility with different monitoring directions in order to monitor the passage 20. At this time, as shown in FIG. 3, an image 31 on the right side of the passage is output from the camera 11, an image 32 on the front side is output from the camera 12, and an image 33 on the left side of the passage is output from the camera 13. The Here, it is assumed that the output images of the cameras 11 to 13 have a resolution of horizontal 720 × vertical 480 pixels.

  The monitoring image control unit 1 controls display and recording / reproduction of the monitoring image. The monitoring image control unit 1 is designated by the recording unit 1a that controls the recording of the camera image or the composite image in the recording unit 9 and the search / area designation unit 5. A search unit 1b for searching for the monitoring video from the recording unit 9, a playback unit 1c for controlling the playback of the monitoring video searched by the search unit 1b, and a tag information attachment unit for attaching tag information to the image detected by the motion detection unit 7 1d and a tag information propagation unit 1e for attaching the tag information attached to the composite image 155 to the corresponding camera image 154. Each unit of the monitoring image control unit 1 is realized by a processor executing an operation defined by a program.

  The data holding unit 2 holds camera parameter data such as the angle of view and direction of each of the cameras 11 to 13 and composite map data generated based on the camera parameter data.

  The image composition unit 3 generates a composite image from the output images of the cameras 11 to 13. As shown in FIG. 2, when the cameras 11 to 13 are installed at different positions in the monitoring direction, the images 31 to 33 of the cameras 11 to 13 are set as camera parameters 2 such as shooting directions and angles of view of the cameras 11 to 13. Based on this, the image is transformed into images 41, 42, and 43 as shown in FIG. 4 by projection conversion or the like, and the transformed images are synthesized. The composite image has a resolution of horizontal 720 × vertical 200 pixels. By this synthesis, as shown in FIG. 5, a horizontally long composite image 50 overlooking the passage 20 by 180 degrees, which is taken using a fisheye camera, is generated.

  This composition can also be realized using composition map data generated based on camera parameter data. The composite map data defines the output data of each pixel on the composite image by the output data of the pixel of each camera image. When the cameras 11 to 13 are arranged as shown in FIG. 2, a composite map 60 as shown in FIG. 6 is stored. In FIG. 6, the composite map 60 is such that the output data of the pixel 61 on the composite image is composed of (x1, y1) pixel data of the camera 11, and the output data of the pixel 62 is (x2, y2) of the camera 12. ) And (x3, y3) pixel data of the camera 13 are calculated with weights (1-α) and (α), respectively. Similarly, all the pixels and the cameras 11 to 11 are calculated. The relationship with 13 pixels is stored in the data holding unit 2 as composite map data. The synthesized image synthesized based on the synthesized map 60 also has a resolution of horizontal 720 × vertical 200 pixels.

  The display unit 4 displays the image output from the monitoring image control unit 1, and the monitoring image control unit 1 outputs this composite image to the display unit 4. Therefore, the supervisor 8 who has seen the composite image can intuitively grasp at a glance where the surveillance image shows in the facility.

  The time code generation unit 6 generates a time code to be added to the recorded image. The recording unit 1a of the monitoring image control unit 1 generates the time code generation unit 6 in the output images of the cameras 11 to 13. Is recorded in the recording unit 9 and the synthesized image synthesized from the image is recorded in the recording unit 9 with the same time code. The recording unit 9 records a camera image 154 taken by each of the cameras 11 to 13 and a synthesized image 155 synthesized by the image synthesizing unit 3.

  The motion detection unit 7 detects a predetermined specific motion from the composite image 155, and the tag information attachment unit 1 d of the monitoring image control unit 1 creates a composite image indicating the specific motion detected by the motion detection unit 7. On the other hand, motion tag information is given.

  For example, consider a case where a person passes through a passage as shown in FIG. Assume that the composite image is obtained in the order of FIG. 7A, FIG. 7B, and FIG. 7C. At this time, if the motion detection unit 7 is set in advance to detect a moving object moving from the right end to the left end on the composite image, an object moving from the right end to the left end on the composite image is detected as shown in FIG. Only when it is done, tag information is added to the composite image. Tag information is not attached to movements of a person turning back in the middle of a passage, disappearing to a middle door, or appearing from a middle door.

  It is also possible to attach different tag information for different movements of the moving object by changing the setting of the motion detection unit 7. For example, the motion detection unit 7 detects the movement of the moving object that is turned back in the middle of the passage, the movement of the moving object that disappears to the middle door, and the movement of the moving object that appears from the middle door. If it sets so that the detection signal according to a motion may be output to the monitoring image control part 1, the tag information attachment part 1d of the monitoring image control part 1 will attach separate motion tag information to a synthesized image according to a detection signal.

  It should be noted here that when such movement is detected from the images of individual cameras, a complicated analysis in which each image is linked is necessary. However, when motion detection is performed on a composite image, The point is that the movement of the moving object can be detected stably.

  The tag information propagation unit 1e of the monitoring image control unit 1 applies the same tag information attached to the composite image to the camera image in which the moving object is reflected among the camera images used to create the composite image. To do. The operation at this time will be described later.

  The search / area designating unit 5 serves as an interface when the supervisor 8 designates image display. The supervisor 8 searches the recorded video of the suspicious person who has passed through the passage. When designated by the area designation unit 5, the search unit 1b of the monitoring image control unit 1 extracts only the synthesized image to which the motion tag information indicating passage of passage is added from the synthesized image 155 of the recording unit 9, and reproduces it. The image is displayed on the display unit 4 through the unit 1c. When the monitor 8 who has seen this composite image obtains the display of the high-resolution camera image from the search / area specifying unit 5, the search unit 1 b of the monitor image control unit 1 includes the camera image 154 in the recording unit 9. Then, a camera image to which the same tag information is added is extracted, and this image is displayed on the display unit 4.

  In this way, the suspicious person's movement can be accurately grasped from the composite image, and tag information can be added to the composite image and the camera image based on it. Since the video to be displayed can be narrowed down using this tag information, the monitor can efficiently monitor.

  Further, when the supervisor 8 wants to confirm the face of the person shown in the composite image, as shown in FIG. 9A, the monitor 8 selects a part from which the person's face can be recognized from the composite image, and through the search / area designating unit 5 A search / area designation request is issued to the monitoring image control unit 1. Receiving this, the search unit 1b of the monitoring image control unit 1 uses the composite map data 2 to search and extract the data of the camera image corresponding to the designated area on the composite image, and the face image of FIG. Present as shown in b).

  At this time, the synthesized image is 720 × 200 pixels as a whole and has a low resolution, but each camera image is 720 × 480 pixels and is clear. However, a composite image that can track a person for a long time is more suitable for selecting a place where the person's face can be best seen. Accordingly, by selecting a person's face area from the composite image and extracting the camera image of the area, the supervisor can see a clear face image of the person and can more surely confirm the passing person. .

  Processing for extracting a face image will be described with reference to FIGS. FIG. 10 shows an example of a monitoring image series based on a composite image and area designation by the supervisor. The supervisor 8 looks at the image series (moving image) as shown in FIG. An image (a) where the face can be seen well is selected, a rectangular area where the face can be seen in the composite image is designated, and an enlargement display request is issued from the search / area designation unit 5 to the monitoring image control unit 1. At this time, the search / area designating unit 5 receives command data indicating an enlarged display request, time code data indicating an image (a), data indicating a rectangular area (the upper left end point 101 (X4, Y4) and the lower right end point). 102 (X5, Y5 coordinate data) is sent to the monitoring image control unit 1.

  When the search unit 1b of the monitoring image control unit 1 receives command data indicating an enlarged display request from the search / area designating unit 5, the search unit 1b accesses the data holding unit 2 and acquires the composite map 60. Then, as shown in FIG. 11, the data indicating the input camera corresponding to the upper left corner point 101 (X4, Y4) and the lower right corner point 102 (X5, Y5) on the composite image and the coordinate data on the input image are displayed. obtain. In the example of FIG. 11, the pixel data of the upper left end point 101 is (x6, y6) pixel data of the camera 13, and the pixel data of the lower right end point 102 is (x7, y7) pixel data of the camera 13. It is shown that.

  Thus, when the input cameras corresponding to the two points are the same, the camera image to which the same time code data as the time code data indicating the image (a) is attached from the camera images 154 of the input camera 13 is attached. And a rectangular area image indicated by the coordinate data ((x6, y6), (x7, y7)) is extracted from the retrieved camera image and output to the display device 4. By such processing, a clear face image of the input camera shown in FIG. 9B can be presented to the supervisor.

  If the rectangular area designated by the supervisor 8 is an area based on a plurality of camera images as shown in FIG. 12, the camera image is extracted by another method. In the example of FIG. 12, the pixel data of the upper left end point 121 on the composite image is (x8, y8) pixel data of the camera 12, and the pixel data of the lower right end point 122 on the composite image is (x9 of the camera 13). , Y9) pixel data). As described above, when the input cameras corresponding to the two points are different, the points corresponding to (the minimum coordinate value of X and the maximum coordinate value of Y) of the composite map data corresponding to the camera 13 on the outer periphery of the rectangle. Calculated as the upper left end point of the camera 13 (camera 13: x10, y10)), the point corresponding to (the maximum coordinate value of X, the minimum coordinate value of Y) of the composite map data corresponding to the camera 12 is the lower right end point of the camera 12. A point (camera 12: (x11, y11)) is calculated, and a camera image to which the same time code data as the time code data of the image (a) is attached is searched from the recorded images of the input cameras 12 and 13, respectively. Further, coordinate data (camera 12: ((x8, y8), (x11, y11))) and (camera 13: ((x9, y9), (x10, y10))) are indicated from the retrieved camera image. Extracts an image of the rectangular region, and outputs to the display unit 4. By such processing, the face image shown in FIG. 13B can be presented to the supervisor.

  In this way, even when the part where the input camera is switched is selected on the composite image, the face image of a clear person of both input cameras is searched, extracted, and presented at the same time. The face image of can be confirmed reliably.

  The method for obtaining a camera image corresponding to a composite image using the composite map data is when the tag information propagation unit 1e of the monitoring image control unit 1 adds the tag information attached to the composite image to the camera image. Also used. At this time, as in the case of FIG. 11 and FIG. 12, the tag information propagation unit 1e indicates that the motion area on the composite image detected by the motion detection unit 7 is the coordinate data of one input camera on the composite map data. Whether it is related to the coordinate data of a plurality of input cameras. Then, a camera image with the same time code as the synthesized image is selected from the camera images of the related cameras, and the same motion tag information as that of the synthesized image is attached.

  In this way, the supervisor who monitors the video of this monitoring device can grasp at first glance where the facility is being monitored by looking at the composite image, and can also view the composite image to be reproduced. However, a clear image of the corresponding input camera can be reproduced and confirmed for the portion of the composite image that is desired to be noticed. Also, in this monitoring device, since motion detection is performed using a composite image and motion tag information is attached to the composite image and the camera image based on the detection, whether or not it is an intruder can be determined only by analyzing individual camera images. Even if it is uncertain, an intruder can be detected stably, and monitoring efficiency in subsequent search and reproduction is improved. Therefore, intruders in the facility can be monitored in detail and efficiently.

  Next, a modified example of this monitoring apparatus will be described. FIG. 14 shows another example of the camera arrangement. The installation positions of the cameras 11 to 13 in FIG. 14 are the same as the installation positions in FIG. 2, but the angles of view of the cameras 11, 12, and 13 are different. As shown in FIG. 14, among the three cameras 11, 12, and 13 that monitor the passage, a telephoto camera is used for the cameras 11 and 13, and a wide-angle camera is used for the camera 12. The angles 141 and 143 are set to 45 degrees or less, and the angle of view 142 of the camera 12 is set to 90 degrees or more, which is twice or more of the angles of view 141 and 143.

  Images taken by the cameras 11, 12, and 13 shown in FIG. 14 are shown in FIG. In the images 31 and 33 obtained by photographing the left and right ends of the passage, the persons 151 and 153 are at a relatively long distance, and in the image 32 obtained by photographing the center, the person 152 is at a relatively short distance. Therefore, when shooting at the same angle of view, the image of the person 152 photographed at a short distance is extremely large and the images of the persons 151 and 153 photographed at a long distance are extremely small. The image of the person being photographed cannot be extremely large or extremely small.

  As in FIG. 4, FIG. 16 synthesizes the captured images 31 to 33, so that the captured images 31 to 33 are converted into images 161 to 163 by projection conversion based on parameters such as the shooting direction and the angle of view of each camera. A deformed state is shown. By synthesizing this deformed image, a horizontally long synthesized image overlooking the passage 180 degrees can be obtained as shown in FIG. Here, the difference from FIG. 4 is that the ranges 161 and 163 on the composite image of the images taken at the left and right ends of the passage are small, and the range 162 on the composite image of the image taken at the center of the passage is large. It is.

  When such a composite image is displayed on the display unit 4 and wants to confirm the face of a person appearing in the composite image, the supervisor 8 first looks at the image composition shown in FIG. 171, 172, and 173 are respectively designated. At this time, the face image of the person is retrieved from the corresponding input camera image (FIG. 15) and presented. As shown in FIG. 15, the image of the person to be photographed is extremely large on each image. Because it is not extremely small, the face image of a person at a long distance can be seen clearly.

  Thus, when monitoring the left and right with a telephoto camera and monitoring the front with a wide-angle camera whose angle of view differs from this telephoto camera by more than twice, an intruder is detected from the composite image of the three camera images, and the person's When the face image is viewed in detail with the camera image, even the face images of persons far away from the left and right can be viewed in more detail.

(Second Embodiment)
FIG. 18 shows a schematic configuration of a monitoring apparatus according to the second embodiment of the present invention. The monitoring device in FIG. 14 monitors facilities such as buildings, and presents a composite image obtained by combining a camera image and a facility structure.

  The monitoring apparatus includes cameras 11 to 16, a monitoring image control unit 181, a data holding unit 2, an image synthesis unit 3, a display unit 4, a search / area designation unit 5, a time code generation unit 6, a motion detection unit 7, and a recording unit. 9 and a facility structure data holding unit 182. As is apparent from the figure, the monitoring apparatus has substantially the same configuration as that of the monitoring apparatus of FIG. 1, but differs in that it includes a facility structure data holding unit 182 and the monitoring image control unit 181 controls display, recording / reproduction of monitoring images. . In addition, the image synthesizing unit 3 includes six cameras, and the image synthesizing unit 3 generates a composite image obtained by back projecting the output images of the cameras 11 to 16 on the facility structure and a floor image obtained by projecting back to the floor. Each floor composite image is recorded together with the camera image 154 and the composite image 155.

  The monitoring image control unit 181 includes an erasing unit 1f that deletes an image having a low priority from the recording unit 9 in addition to the configuration of the monitoring image control unit 1 in the monitoring apparatus of FIG.

  The facility structure data 182 is data that three-dimensionally represents the positions of the plurality of cameras 11 to 16 in the facility, the wall surface of the facility, the passage surface of the floor, and the like, as schematically shown in FIG. . The plurality of cameras 11 to 16 are distributed in the facility, and a plurality of cameras are installed on each floor.

  As shown in FIG. 20, the image composition unit 3 creates an image obtained by back-projecting (202) the captured image of the input camera 201 onto the passage surface 204 and the wall surface 203 in the facility as a composite image. The term “back projection” as used herein refers to an operation of projecting a photographed image on three-dimensional data such as a facility wall or a passage surface in a direction opposite to the actual photographing. The composite map data stored in the data holding unit 2 is obtained by defining the output data of each pixel on the composite image with the output data of the pixel of the camera image, and using the camera parameters and the facility structure data 182, Pixel data can be created by back projecting to the facility structure data 182. The image composition unit 3 uses the composite map data 2 to generate a composite image obtained by back projecting the output images of the cameras 11 to 16 to the facility structure, and reversely outputs the output images of the cameras 11 to 16 to the floor. A projected floor composite image is generated.

  The recording unit 1a of the monitoring image control unit 181 adds the time code generated by the time code generation unit 6 to the camera image of each camera 11 to 13 and records it in the recording unit 9, and the image composition unit 3 also records the camera image. Is recorded on the recording unit 9 with the same time code as that of the camera image. Based on the request from the search / area designating unit 5, the playback unit 1c of the monitoring image control unit 181 plays back a composite image 155 as shown in FIG. Present. At that time, as shown in FIG. 21 (a), the facility is rotated by a predetermined angle in the direction 211 or 212 so that the facility can be seen from various directions, as shown in FIG. ), Only the requested floor 213 is displayed. Therefore, the supervisor can perform monitoring while looking down on the inside of the facility.

  Further, when the monitor designates a camera at a position of interest from the composite image or designates a face region on the composite image through the search / region specification unit 5, the search unit 1 b of the monitor image control unit 181. As in the first embodiment, the corresponding camera image and the face area of the camera image are searched using the composite map data 2 and the time code and displayed on the display unit 4. The motion detection unit 7 detects a predetermined specific motion from each floor composite image 183 generated by the image composition unit 3 and including a plurality of camera images.

  FIG. 22 schematically shows such a state. The motion detection unit 7 uses only the floor composite images 221, 222, and 223 to move only the motion 227 in which the person 224 moves on the floor from the back projection position 225 of the first camera to the back projection position 226 of the second camera. Is detected. Such movement may not be sufficiently grasped even by analyzing the images of individual cameras, but can be accurately detected by using a floor composite image obtained by combining a plurality of camera images projected on the floor.

  The tag information attachment unit 1d of the monitoring image control unit 181 adds motion tag information to the floor composite image 183 that shows the specific motion detected by the motion detection unit 7. The tag information propagation unit 1e is the same as the tag information attached to the floor composite image, the composite image 155 with the same time code as the floor composite image, or the camera image used to create the floor composite image. And attach it to the camera image with the same time code as the floor composite image. The tag information attached to the composite image 155, each floor composite image 183, and the camera image is used to narrow down the images to be reproduced, as in the first embodiment.

  Further, the erasing unit 1f of the monitoring image control unit 181 determines the priority of the monitoring image stored in the recording unit 9 based on the tag information attached to these images, and an image in which the time set according to the priority has elapsed. Is automatically deleted. In this way, the storage capacity of the device can be used efficiently.

  In this way, the monitoring device can present a synthesized image whose position can be intuitively understood to the supervisor, and the supervisor selects individual camera images and camera image areas from the synthesized image. And can be monitored in detail. Therefore, detailed and efficient monitoring is possible even in a complex and large facility.

  INDUSTRIAL APPLICABILITY The monitoring device of the present invention is useful as a monitoring device capable of intuitively knowing the position of a monitoring target for a monitor and presenting an image in which an abnormal state such as a suspicious person's movement or characteristic can be easily grasped. It is. Further, it has an advantage that the video requested by the supervisor can be narrowed down and presented from the accumulated video, and can be widely used in a centralized monitoring system of a large-scale facility.

The figure which shows schematic structure of the monitoring apparatus of the 1st Embodiment of this invention. The figure which shows an example of the camera arrangement | positioning in the monitoring apparatus of the 1st embodiment of this invention The figure which shows an example of the image which the some camera in the monitoring apparatus of the 1st embodiment of this invention image | photographed. The figure explaining an example of the image composition method in the monitoring apparatus of the 1st embodiment of this invention The figure which shows an example of the synthesized image in the monitoring apparatus of the 1st embodiment of this invention The figure which shows an example of the composite map for image composition in the monitoring apparatus of the 1st Embodiment of this invention The figure which shows an example of the monitoring image series by the synthesized image in the monitoring apparatus of the 1st embodiment of this invention The figure which shows an example of the motion detection setting on the synthesized image in the monitoring apparatus of the 1st embodiment of this invention. The figure which shows the face image extracted from the synthesized image and the camera image in the monitoring apparatus of the 1st Embodiment of this invention FIG. 4 is a schematic diagram illustrating an example of a monitoring image sequence based on a composite image in the monitoring apparatus according to the first embodiment of the present invention, and region designation by a supervisor. The figure which shows an example of the extraction method of the input camera image using the synthetic | combination map in the monitoring apparatus of the 1st embodiment of this invention. The figure which shows the other example of the extraction method of the input camera image using the synthetic | combination map in the monitoring apparatus of the 1st Embodiment of this invention. The figure which shows the face image extracted from the synthesized image and two camera images in the monitoring apparatus of the 1st embodiment of this invention The figure which shows the other example of the camera arrangement | positioning in the monitoring apparatus of the 1st embodiment of this invention. The figure which shows the other example of the image which the some camera in the monitoring apparatus of the 1st embodiment of this invention image | photographed. The figure explaining the other example of the image composition method in the monitoring apparatus of the 1st embodiment of this invention The figure which shows the other example of the synthesized image in the monitoring apparatus of the 1st embodiment of this invention The figure which shows schematic structure of the monitoring apparatus of the 2nd embodiment of this invention. The figure explaining an example of the facility structure data in the monitoring apparatus of the 2nd embodiment of this invention The figure explaining an example of the synthesized image generation method in the monitoring apparatus of the 2nd embodiment of this invention The figure which shows the synthesized image (a) and the synthesized image (b) for one floor in the monitoring apparatus of the 2nd embodiment of this invention. The figure which shows the motion detection in the floor composite image in the monitoring apparatus of the 2nd embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Monitoring image control part 1a Recording part 1b Retrieval part 1c Reproduction part 1d Tag information attachment part 1e Tag information propagation part 1f Erasing part 2 Data holding part 3 Image composition part 4 Display part 5 Search and area designation part 6 Time code generation part 7 Motion detection unit 8 Monitor 9 Recording unit 11 Camera 12 Camera 13 Camera 14 Camera 15 Camera 16 Camera 20 Passage 31 Camera image 32 Camera image 33 Camera image 41 Deformed image 42 Deformed image 43 Deformed image 50 Synthetic image 60 Synthetic map 61 Pixel 62 Pixel 101 Upper left end point 102 Right lower end point 121 Left upper end point 122 Right lower end point 141 Angle of view 142 Angle of view 143 Angle of view 151 Person 152 Person 153 Person 154 Camera image 155 Composite image 161 Deformed image 162 Deformed image 163 Deformed image 171 Face area 172 Face Area 173 Face area 181 Monitoring image control unit 182 Structure data holding unit 183 Each floor composite image 201 Camera 202 Back projection 203 Wall surface 204 Path surface 211 Angle change direction 212 Angle change direction 213 Composite image 221 for one floor Floor composite image 222 Floor composite image 223 Floor composite image 224 Intruder Image 225 Position 226 Position 227 Movement

Claims (2)

A monitoring device that performs monitoring using a plurality of cameras,
A composite image generating means for generating a composite image using a camera image photographed by the camera;
Recording means for holding the camera image and the composite image;
Display means for displaying an image;
When an arbitrary area of the composite image displayed on the display means is selected, search means for searching the camera image corresponding to the selected area or the area of the camera image and displaying the area on the display means ;
Motion detection means for detecting motion from the composite image generated using a plurality of camera images;
Tag information adding means for adding tag information to the composite image detected by the motion detecting means;
Tag information propagation means for searching the camera image used in the composite image with the tag information added and adding the tag information to the camera image;
Monitoring device. The monitoring device according to claim 1 ,
A monitoring apparatus comprising: an erasing unit that determines a priority of a composite image and a camera image to which the tag information is added based on the tag information, and determines an image to be deleted based on the priority.

Images