JP2008028605A - Image recording device for monitoring - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image recording device for monitoring which can certainly confirm all images photographed by all cameras installed in a room, and can easily and certainly search an image to find a cause of a fault when a certain fault occurs. <P>SOLUTION: Many cameras are installed in the room to be monitored, and imaging directions are fixed for respective cameras so that all the room can be covered. Images photographed by respective cameras are input in an image composition unit 204, a sheet of composite image viewed from the above is created. An image input to the image composing unit 204 and an composite image output from the image composing unit 204 are combined at every time. A database is created, and stored in a recording unit 213. The database stores position information of an image from a camera before composition which is used when the composite image is created, and an image from the camera before composition can be called from the composite image. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a monitoring image recording apparatus that records images taken by a plurality of cameras.

  Conventionally, provided with a multi-display control device that generates a multi-display image from images captured by a plurality of cameras, and a sensor for abnormality detection corresponding to each of the plurality of cameras, and when an abnormality is detected by the sensor, A multi-display monitoring system is devised that inputs a detection signal from the sensor and an image taken by each camera to a multi-display control device, and superimposes an abnormality detection result by the sensor on an output image of the multi-display control device. (For example, refer to Patent Document 1).

  Further, conventionally, it is used in a device that supports driving such as parking by combining images taken by a plurality of photographing means (cameras) mounted on a car and presenting the driver with the situation around the car. An image composition device has been devised (see, for example, Patent Document 2). An image composition device disclosed in Patent Document 2 detects a plurality of photographing units, a photographing state detecting unit that detects photographing states of the plurality of photographing units, and a plurality of images photographed by the plurality of photographing units. A plane projection unit that projects the projected image on the projection unit to reflect the captured shooting situation, and generates a second projection image by projecting the plurality of first projection images onto a three-dimensional projection model. 3D projection model projecting means, virtual projection means for virtually projecting the second projection image, 3D projection model determining means for determining the position and shape of the 3D projection model, and imaging by the virtual projection means Virtual projection means for determining the situation, and projection status determining means.

JP-A-11-184448 JP 2002-83285 A

  However, in the conventional multi-display monitoring system, when a plurality of cameras are installed in a room and the room is monitored, the number of images that can be displayed in a multi display is limited, and images taken by all the cameras can be displayed at a time. It is impossible to display. In other words, the image displayed on the monitor needs to be large enough to enable human identification of the content, and even if the monitor is large, the number of images that can be displayed increases, but the number is limited. In addition, even if multiple images are displayed as many as possible, if there is any abnormality, it is very often used to check the images taken by all cameras and identify the cause of the abnormality. Requires labor. Since more images will be searched, important videos may be missed.

  Also, the images from each camera are not linked to each other. For example, when the operator wants to see the right side of the area captured by a certain camera, select another camera that captures the corresponding area. However, it is necessary to perform an operation for displaying an image from the camera. In addition, in order to perform this work, it is necessary to grasp the area to be photographed by each camera and grasp the relative positional relationship between the cameras. In addition, when any abnormality occurs, it is necessary to repeatedly perform the above-described work in a short time, and for that purpose, advanced technology and experience are required.

  The present invention has been made in view of such circumstances, and can surely check all images taken by all cameras installed indoors, and if any abnormality occurs, the cause of the abnormality It is an object of the present invention to provide a monitoring image recording apparatus capable of easily and surely searching for an image capable of locating an image, and capable of mutually linking images from each camera.

The above object is achieved by the following configuration.
(1) Sorting for aligning a plurality of images photographed by a plurality of cameras at the same time in a monitoring image recording apparatus that captures and records images photographed by a plurality of cameras connected to the network via the network And a plurality of images at the same time subjected to alignment processing by the sorting means, and motion detection is performed, and a seam position at the time of combining the plurality of images as one image is selected and selected. An image synthesizing unit that synthesizes an image while moving the joint position and evaluating a change amount of the synthesized image, and an image that compresses the synthesized image synthesized by the image synthesizing unit and a plurality of images that are the basis of the synthesized image. A compression unit, and a database that generates a database that associates the composite image compressed by the image compression unit with the plurality of images that are the basis of the composite image. The database generation means, the recording means for recording the database generated by the database generation means, the database search means for searching the database recorded in the recording means in accordance with an operator's instruction, and the search result by the database search means Image decompression means for decompressing the compressed image read from the recording means and returning it to the original state before compression.

(2) In the monitoring image recording apparatus according to (1), the image synthesizing unit includes a plurality of motion detecting units that detect a motion from images captured by the plurality of cameras, a current synthesized image, A change detection unit that detects a difference from the composite image generated immediately before, and a difference detected by the change detection unit is compared with a predetermined threshold value. A seam position determining means for determining a motion of the image captured by the camera from the detection result and determining a seam position of the image to be moved when generating the composite image from the determination result; and a seam of the image Seam position holding means for holding the position, and seam position movement for moving the current seam position held by the seam position holding means to the seam position determined by the seam position determining means And the difference between the synthesized image after movement of the joint position and the current synthesized image is the threshold value after the joint position of the current synthesized image is moved to the joint position determined by the joint position determining means. The image synthesis, change detection, joint position determination, and joint position movement processing are repeated until the difference between the current composite image and the composite image after moving the joint position falls within the threshold. The moved seam position is stored in the seam position holding means.

(3) In the monitoring image recording apparatus according to (1) or (2), the plurality of cameras are installed in a room for monitoring so that respective shooting ranges overlap each other, and the image synthesis is performed. The means converts a plurality of images taken at the same time to generate a composite image in which the room is viewed from above.

(4) In the monitoring image recording apparatus according to (1) or (2) above, the plurality of cameras are installed on the entire floor of the building to be monitored so that the respective shooting ranges overlap each other. The image synthesizing unit performs image conversion on images taken by the plurality of cameras for each room or each specific shooting location, and generates a synthesized image viewed from above for each room or shooting location.

(5) The monitoring image recording apparatus according to any one of (1) to (4), further including floor information management means for managing position information of each of the plurality of cameras, wherein the database generation means includes the image The position information of each of the plurality of cameras managed by the floor information management unit is generated when generating a database in which the combined image compressed by the compression unit and the plurality of images that are the basis of the combined image are generated. By utilizing this, a database is generated so that each of the plurality of images before synthesis can be called.

(6) In the monitoring image recording apparatus according to (5), the floor information management unit manages information on a structure of a building floor including a doorway and a window of the room to be monitored, and the database generation unit Is a database so that a composite image can be called over a plurality of monitoring areas by utilizing the position information of each of the plurality of cameras managed by the floor information management means and the information on the structure of the building floor. Is generated.

(7) In the monitoring image recording apparatus according to any one of (2) to (6), when monitoring is performed in a specific room, each of the plurality of motion detection units of the image synthesis unit is When the predicted motion vector of the monitoring target is calculated, the movement destination of the monitoring target is predicted based on the predicted motion vector of the monitoring target calculated by each of the motion detection means, and when the monitoring target is predicted to go out of the room Includes control means for performing control to read out a composite image of the predicted destination area.

(8) In the monitoring image recording apparatus according to any one of (2) to (7) above, head / face detection for detecting a human head and face with respect to images captured by the plurality of cameras. The joint position determining means determines the joint position of the image composition so that the human face is displayed in the composite image when the head / face detection means detects a human face.

(9) In the security system, a plurality of cameras connected directly to a network or via a wireless transceiver, the monitoring image recording apparatus according to any one of (1) to (8), and the monitoring image A monitor for displaying an image recorded on the recording apparatus; and an input unit for giving an operation command to the monitoring image recording apparatus.

  In the monitoring image recording apparatus described in (1) above, motion detection is performed on monitoring images taken by a plurality of cameras for image synthesis, and a seam position for image synthesis is selected from the motion detection result, and the seam is selected. Since the image is synthesized while moving the position and evaluating the amount of change in the synthesized image, the monitoring target moving from the synthesized image is moved from the synthesized image by synthesizing images taken by a plurality of cameras into one image. It is possible to prevent disappearance or unnatural display, and thus, a lot of information obtained from a plurality of cameras can be easily grasped.

  In addition, since information included in images taken by a plurality of cameras can be viewed as a single composite image, changes in the situation can be easily grasped. Furthermore, even if some abnormality occurs, it is possible to search by tracing a plurality of camera images that are the basis of one composite image, so that the labor required for the search can be reduced.

  In the monitoring image recording apparatus described in (2) above, the joint position at the time of image composition is selected, and the joint position is moved to perform image composition while evaluating the amount of change in the composite image. It is possible to prevent the monitoring target from disappearing from the composite image or being displayed unnaturally.

  In the monitoring image recording apparatus described in (3) above, a plurality of cameras are installed in a room to be monitored so that respective shooting ranges overlap with each other, and a plurality of images shot at the same time by these cameras are provided. Since the image is converted to generate a composite image in which the room is viewed from above, it is possible to very easily grasp how the monitoring target moves.

  In the monitoring image recording apparatus described in (4) above, a plurality of cameras are installed on the entire floor of the building to be monitored so that the respective shooting ranges overlap each other, and images taken by these cameras are stored in the room. Image conversion is performed every time or every specific shooting location, and a composite image seen from above is generated for each room or shooting location, so when tracking the movement of the monitoring target across multiple rooms, There is no need to sequentially switch the monitoring video in consideration of the structure, and the monitoring target can be traced very easily.

  In the monitoring image recording apparatus described in (5) above, when generating a database in which a composite image and a plurality of images that are the basis of the composite image are generated, the position information of each of the plurality of cameras is used for the synthesis. Since the database is generated so that each of the plurality of previous images can be called, when any abnormality occurs, an image that can determine the cause of the abnormality can be easily and reliably searched.

  In the monitoring image recording apparatus described in (6) above, it is possible to manage information related to the structure of the building floor, such as the entrance / exit of a room to be monitored and windows, and to call up a composite image over a plurality of monitoring areas. When some abnormality occurs, the cause can be searched over a plurality of monitoring areas.

  In the monitoring image recording apparatus described in (7) above, each of the plurality of motion detecting means of the image synthesizing means calculates the predicted motion vector to be monitored, and predicts the destination of the monitoring target based on this predicted motion vector. When the monitoring target goes out of the room, the composite image of the predicted destination area is read out, so that the monitoring target moving across a plurality of monitoring areas can be traced very easily.

  In the monitoring image recording apparatus described in (8) above, since the image composition is performed so that the face can be recognized by determining the joint position of the image composition so that the human face is displayed in the composite image, the person can be easily Can be recognized.

  In the security system described in (9) above, the number of surveillance cameras that can be handled is practically infinite, and monitoring by the security system in the ubiquitous sensor network can be realized.

  DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments for carrying out the invention will be described in detail with reference to the drawings.

(Embodiment 1)
FIG. 1 is a block diagram showing a schematic configuration of a security system according to Embodiment 1 of the present invention. In the figure, the security system of the present embodiment includes a wired camera 101, a wireless camera 102, a wireless receiver 103, a network 104, a monitoring image recording device 105, a monitor 106, and an input unit 107. It is prepared for. Usually, a plurality of wired cameras 101 and wireless cameras 102 are used. The wired camera 101 converts the video signal into an electrical signal, compresses the image, and sends the compressed image data to the network 104. The wireless camera 102 converts the video signal into an electrical signal, performs image compression, and transmits the compressed image data wirelessly. A wireless signal transmitted from the wireless camera 102 is received by the wireless receiver 103.

  The wireless receiver 103 performs wireless data reception and sends the received data to the network 104. In the present embodiment, the network 104 is a dedicated network for supplying compressed image data transmitted from the wired camera 101, the wireless camera 102, and the wireless receiver 103 to the monitoring image recording apparatus 105, and provides high-speed data communication. Has sufficient network capacity. The monitoring image recording apparatus 105 accumulates compressed image data sent via the network 104. In addition, the stored image can be searched and reproduced. The monitor 106 is composed of an LCD (Liquid Crystal Display) or the like, and displays control information necessary for operation of the monitoring image recording apparatus 105, a playback image, and the like. The input unit 107 includes a keyboard, a mouse, and the like, and inputs control information by an operator operation to the monitoring image recording apparatus 105.

  FIG. 2 is a block diagram illustrating a schematic configuration of the monitoring image recording apparatus 105. In the figure, a monitoring image recording apparatus 105 includes a network I / F (interface) 201, a sorting unit (alignment means) 202, an image expansion unit 203, an image composition unit 204, a floor information management unit 205, A control unit 207, an image compression unit 209, a database generation unit 210, a database search unit 211, a data write control unit 212, a recording unit 213 such as a hard disk, a data read control unit 214, and an image expansion unit 215. And a display control unit 216.

  The network I / F 201 receives compressed image data from a plurality of cameras (wired camera 101 and wireless camera 102) supplied via the network 104 and inputs the compressed image data to the sorting unit 202. The sorting unit 202 arranges compressed image data from a plurality of cameras input from the network I / F 201 for each photographing time and each camera, and inputs the images to the image expansion unit 203 and the database generation unit 210. The image expansion unit 203 expands the compressed image data arranged for each shooting time and camera by the sorting unit 202 and inputs the compressed image data to the image composition unit 204. Based on the information from the floor information management unit 205, the image composition unit 204 generates a composite image such that the expanded image input from the image expansion unit 203 is viewed from the top for each room in which the camera is installed. .

  The floor information management unit 205 inputs the installation information of cameras installed in the room to be monitored and the floor structure information of the building being monitored to the image composition unit 204 and the database generation unit 210. In this case, the camera installation information includes camera position information (horizontal position and vertical position) in a room to be monitored, a camera shooting direction, a shooting range, and the like. The floor structure information is the position information of each room, the position information of the entrance / exit of the room (door, etc.), and the like. The control input 206 is a control signal input from the input unit 107 (see FIG. 1). The control unit 207 controls each unit of the apparatus 105. For the control unit 207, a microcomputer is usually used. A display output 208 is a display signal output toward the monitor 106.

  The image compression unit 209 compresses the composite image generated by the image composition unit 204 and also compresses a plurality of images taken at the same time as the base of the composite image. The database generation unit 210 includes the floor structure information from the floor information management unit 205, the compressed image data from the sorting unit 202 (compressed image data arranged for each shooting time and camera), and the compression composition from the image compression unit 209. Create a database based on images. The database search unit 211 searches for a corresponding image using the time information, alarm information, camera number, camera position, and the like as keywords. The data writing control unit 212 controls data writing to the recording unit 213. The recording unit 213 stores various image data based on the database generated by the database generation unit 210. The data read control unit 214 performs data read control on the recording unit 213 based on the search result of the database search unit 211. The image decompression unit 215 decompresses the compressed image data read from the recording unit 213 by the data read control unit 214. The display control unit 216 converts information necessary for controlling the apparatus 105 and the image expanded by the image expansion unit 215 into a format that can be displayed on the monitor 216, and outputs it as a display output 208.

  Next, the image composition unit 204 will be described in detail. FIG. 3 is a block diagram illustrating a schematic configuration of the image composition unit 204. In the figure, an image composition unit 204 can synthesize images from four cameras, and includes four frame memories 302 to 305, four motion detection units 306 to 309, a composition processing unit 310, and change detection. A unit 311, a joint position determining unit 312, a joint position moving unit 313, and a joint position buffer 314. Here, the four cameras are only the wired camera 101, for example. Needless to say, only the wireless camera 102 may be used, or the wired camera 101 and the wireless camera 102 may be mixed. The frame memory (A) 302 stores a plurality of frames of images taken by the camera 101-1. The frame memory (B) 303 stores a plurality of frames of images taken by the camera 101-2. The frame memory (C) 304 stores images taken by the camera 101-3 for a plurality of frames. The frame memory (D) 305 stores a plurality of frames of images taken by the camera 101-4.

  The motion detection unit (A) 306 detects the motion of the image captured by the camera 101-1, by calculating the difference between the images of the plurality of frames stored in the frame memory (A) 302, and synthesizes the result. Input to the processing unit 310. The motion detection unit (B) 307 detects the motion of the image captured by the camera 101-2 by calculating the difference between the images of the plurality of frames stored in the frame memory (B) 303, and synthesizes the result. Input to the processing unit 310. The motion detection unit (C) 308 detects the motion of the image captured by the camera 101-3 by calculating the difference between the images of the plurality of frames stored in the frame memory (C) 304, and synthesizes the result. Input to the processing unit 310. The motion detection unit (D) 309 detects the motion of the image captured by the camera 101-4 by calculating the difference between the images of a plurality of frames stored in the frame memory (D) 305, and synthesizes the result. Input to the processing unit 310.

  The compositing processing unit 310 is stored in each of the frame memory (A) 302, the frame memory (B) 303, the frame memory (C) 304, and the frame memory (D) 305 according to the joint position stored in the joint position buffer 314. The image synthesis processing is performed on the obtained images, and a single synthesized image as if the room was viewed from above is generated. The change detection unit 311 calculates the difference between the current composite image and the composite image of the previous frame for each region, and determines which region of the composite image has changed when a preset threshold value is exceeded. The joint position determination unit 312 determines which joint position to move when generating a composite image from the output of the change detection unit 311. The joint position moving unit 313 moves the joint position determined by the joint position determining unit 312 and inputs the movement result to the joint position buffer 314. The joint position buffer 314 holds the joint position input from the joint position moving unit 313. An image input 301 is image data input from the image expansion unit 203. The composite image output 315 is an output of the composition processing unit 310 and is input to the image compression unit 209.

  Next, an installation example of the cameras 101-1 to 101-4 will be described with reference to FIG. In the figure, a monitoring room 400 is a room to be monitored. The camera 101-1 captures the area (A) 405 and the area (B) 406, and the area on the left side of the area (C) 407 and the area (D) 408 (toward the drawing) is also an imaging range. The camera 101-2 captures the area (B) 406 and the area (A) 405, and the area on the left side (toward the drawing) of the area (D) 408 and the area (C) 407 is also an imaging range. The camera 101-3 captures the area (C) 407 and the area (D) 408, and the area on the right side (toward the drawing) of the area (A) 405 and the area (B) 406 is also an imaging range. The camera 101-4 captures the area (D) 408 and the area (C) 407, and the area on the right side (toward the drawing) of the area (B) 406 and the area (A) 405 is also an imaging range. In this way, the shooting ranges of the cameras 101-1 to 101-4 overlap each other, and shooting is performed by at least two cameras in the room to be monitored.

  The area (A) 405 is photographed mainly by the camera 101-1 and the camera 101-2. The area (B) 406 is mainly captured by the camera 101-2 and the camera 101-1. The area (C) 407 is photographed mainly by the camera 101-3 and the camera 101-4. The area (D) 408 is mainly captured by the camera 101-4 and the camera 101-3. The monitoring target 409 is a moving object such as an intruder into the room, and is located near the boundary between the area (A) 405 and the area (B) 406 in FIG.

  Next, a composite image generated by the image composition unit 204 will be described with reference to FIG. For the sake of simplicity, description will be given at the place where four images are combined. In the figure, a composite image 500 is generated by the image composition unit 204. The composite image 500 is generated based on the section (A) 501, the section (B) 502, the section (C) 503, and the section (D) 504. It becomes an image. An image area (A) 501 is an image obtained by cutting an image captured by the camera 101-2 with the joint (AB) 509 and the joint (AC) 510 as a boundary. At this time, coordinate transformation is also performed so as to be rectangular according to the shape of the room. A section (B) 502 is an image obtained by cutting an image captured by the camera 101-1 with the joint (AB) 509 and the joint (BD) 512 as a boundary. At this time, coordinate transformation is also performed so as to be rectangular according to the shape of the room. A section (C) 503 is an image obtained by cutting an image captured by the camera 101-4 with the joint (AC) 510 and the joint (CD) 511 as a boundary. At this time, coordinate transformation is also performed so as to be rectangular according to the shape of the room. A section (D) 504 is an image obtained by cutting an image captured by the camera 101-3 with the joint (CD) 511 and the joint (BD) 512 as a boundary. At this time, coordinate transformation is also performed so as to be rectangular according to the shape of the room. The monitoring target 505 is a moving object such as an intruder, and is located near the joint (AB) 509.

  The joint (AB1) 506 is a starting point for moving the joint when the joint (AB) 509 is obtained. The joint (AB2) 507 is an end point for moving the joint when the joint (AB) 509 is obtained. The joint movement direction 508 indicates the direction in which the joint is moved from the joint (AB1) 506 to the joint (AB2) 507 when the joint (AB) 509 is obtained. A joint (AB) 509 means a boundary when the image segment (A) 501 and the image segment (B) 502 are connected. The joint (AC) 510 means a boundary when the image segment (A) 501 and the image segment (C) 503 are connected. The joint (CD) 511 means a boundary when connecting the section (C) 503 and the section (D) 504. The joint (BD) 512 means a boundary when connecting the section (B) 502 and the section (D) 504.

  Next, the operation of the first embodiment will be described. FIG. 6 is a flowchart for explaining the operation of the first embodiment. Images captured by the cameras 101-1 to 101-4 are transmitted as compressed image data, and are sorted by the sorting unit 202 for each shooting time and for each camera. Next, the image is expanded by the image expansion unit 203 and then input to the image composition unit 204. And with respect to the image image | photographed with each camera 101-1-101-4 installed in the monitoring room 400, each motion detection part (Motion detection part (A) 306, Motion detection part (B) 307, Motion detection) (C) 308 and motion detection unit (D) 309) perform motion detection processing (step S601). The composition processing unit 310 performs image composition processing on the images captured by the cameras 101-1 to 101-4 installed in the monitoring room 400, and a single composite image as seen from above the monitoring room 400. Is generated (step S602).

  Next, the change detection unit 311 compares the current composite image generated in step S602 with the composite image generated in the previous frame, and calculates the difference (change) (step S603). Then, the calculated difference is compared with a preset threshold TH, and if it exceeds the threshold TH, the process proceeds to step S605, and if not, the process proceeds to step S609 (step S604). Here, when a large change that exceeds the threshold value TH is detected in the composite image, it is necessary to move the joint of the composite image to an appropriate position. The joint position determination unit 312 determines which camera has detected the motion from the motion detection result for the images captured by the cameras 101-1 to 101-4 (step S605). The joint position of the image to be moved when generating the composite image is determined from the processing result of step S605 (step S606).

  Here, an image to be combined with the joint position of the combined image will be described with reference to FIG. When two images are combined and a monitoring target exists in the vicinity of the intermediate line, there are three possible relationships between the monitoring target and the joint. The three relationships are shown in a composite image (A) 700, a composite image (B) 706, and a composite image (C) 713. Here, the intermediate line is a line equidistant from the two cameras that take an image.

  The composite image (A) 700 is a case where the monitoring target (A) 703 is displayed in the section (AA) 701, and the joint (ABA) 704 at this time is below the intermediate line (A) 705. To position. In the composite image (B) 706, the joint (ABB) 710 is positioned in the vicinity of the intermediate line (B) 712, and the monitoring target (B) 709 disappears from the composite image and is not displayed in the composite image. In this case, when the joint (ABB) 710 is moved in the moving direction (B) 711, the monitoring target (B) 709 is displayed on the section (AB) 707 like the composite image (A) 700. When the joint (ABB) 710 is moved in the direction opposite to the moving direction (B) 711, the monitoring target (C) 716 has the joint (ABC) 717 from the intermediate line (Cwww) 719 like the composite image (C) 713. The monitoring target (C) 716 is displayed on the section (BC) 715 like the composite image (C) 713.

  Returning to the flowchart of FIG. 6, the value of the current seam position is read from the seam position buffer 314 (step S607). In the case of the composite image 700, the joint (ABA) 704 exists below the intermediate line (A) 705, and position information thereof is stored. The joint position moving unit 313 moves the joint position read from the joint position buffer 314 from the current position in the vertical direction according to the processing result of Step S606 (Step S608). After moving the joint, the process proceeds to step S602, and the processing from step S602 to step S608 is repeated until the difference between the current composite image and the composite image after moving the seam falls within the threshold. When the difference between the composite image and the composite image after moving the joint falls within the threshold TH, the joint position moved at that time is stored in the joint position buffer 314 (step S609).

  As described above, according to the present embodiment, motion detection is performed on the monitoring video captured by the plurality of cameras 101 and 102 that perform image synthesis, and a joint position at the time of image synthesis is selected from the motion detection result, The image to be monitored is moved at the time of image synthesis by synthesizing images taken by a plurality of cameras 101 and 102 by synthesizing images while moving the joints and evaluating the amount of change in the synthesized image. It is possible to prevent the composite image from disappearing or being unnaturally displayed, whereby the operator can easily grasp a lot of information obtained from a plurality of cameras.

  In addition, since information included in images taken by the plurality of cameras 101 and 102 can be viewed as a single composite image, changes in the situation can be easily grasped. Furthermore, even if some kind of abnormality occurs, it is possible to search by tracing the images of the plurality of cameras 101 and 102 that are the basis of one composite image, so that the labor of the operator in the search can be reduced.

  In addition, since a single composite image as if the room to be monitored was viewed from above is generated, the movement of the monitoring object can be grasped extremely easily.

(Embodiment 2)
Next, a second embodiment of the present invention will be described. FIG. 8 is a diagram showing an installation example of the camera of the security system according to Embodiment 2 of the present invention. In the figure, a monitoring floor 801 indicates a floor with a building to be monitored using the monitoring image recording apparatus 105. The monitoring room (A) 802 is a room existing on the monitoring floor 801, and four cameras 101-1 to 101-4 are installed in the same manner as in FIG. 4, and these cameras 101-1 to 101- are installed. From the image captured in step 4, the image composition unit 204 generates a single composite image as if the room was viewed from above. The same applies to the monitoring room (B) 803, the monitoring room (C) 804, the monitoring room (D) 805, and the monitoring room (E) 806, and a composite image is generated one by one by the image composition unit 204. The shared space 807 is a space such as a corridor existing on the monitoring floor 801.

  From the shared space 807 to the monitoring room (A) 802, the monitoring room (B) 803, the monitoring room (C) 804, the monitoring room (D) 805, and the monitoring room (E) 806, the door (A) 813, the door (B ) 814, door (C) 815, door (D) 816, and door (E) 817. The camera 101-5 is installed in the shared space 807 and captures a monitoring video in the shared space 807. The same applies to the camera 101-6, the camera 101-7, the camera 101-8, the camera 101-9, and the camera 101-10. One composite image is generated by the image composition unit 204 from the images taken by the cameras 101-5 to 101-10 installed in the shared space 807. The door (A) 814 is an entrance to a shared space 807 provided in the monitoring room (A) 802. The door (B) 815 is an entrance to a shared space 807 provided in the monitoring room (B) 803. The door (C) 816 is an entrance to a shared space 807 provided in the monitoring room (C) 804. The door (D) 817 is an entrance to a shared space 807 provided in the monitoring room (D) 805. The door (E) 818 is an entrance to a shared space 807 provided in the monitoring room (E) 806.

  Next, an example of image call using a motion vector will be described with reference to FIG. In the figure, a monitoring object 902 is a moving object such as an intruder into the room. In this example, the object to be monitored 902 moves with a motion vector (MVA) 904 one frame before, and currently has a motion vector (MVB) 905. Has moved. The door (C) 816 is an entrance that connects the monitoring room (C) 804 and the shared space 807. The predicted motion vector (MV) 912 is an expected motion vector indicating the position to which the monitoring target 902 moves in the next frame. The cutout area 913 is an area obtained by cutting out a place where the end point of the predicted motion vector (MVP) 912 exists from the monitoring floor 801.

  Next, the operation of the present embodiment will be described. FIG. 10 is a flowchart for explaining the operation of the present embodiment. In the figure, with respect to the monitoring video output from each of the cameras 101-1 to 101-4 installed in the monitoring room (C) 804, each motion detector (motion detector (A) 306, motion detector ( B) 307, motion detection unit (C) 308, motion detection unit (D) 309) performs motion detection processing (step S1001). If a motion is detected in the monitoring room (C) 804 in the motion detection process in step S1001, the process proceeds to step S1003, and if not detected, the process proceeds to S1001 (step S1002). Each of the motion detection units 306 to 309 calculates an expected motion vector of the monitoring target 902 existing in the monitoring room (C) 804 (step S1003). Each of the motion detection units 306 to 309 holds a motion vector (MVA) 904 detected in the previous frame, and based on a difference from the motion vector (MVB) 905 detected in the current frame, the motion vector (MVP) ) 912 is calculated.

  Next, from the information of the predicted motion vector (MVP) 912 calculated in step S1003, the control unit 207 determines whether the movement destination in the next frame of the monitoring target 902 is in or outside the current monitoring area. To do. If it is within the monitoring area, the process proceeds to step S1001, and if it is outside the area, the process proceeds to step S1005 (step S1004). The predicted motion vector (MVP) 912 has information on the position of the starting point of the vector and information on the distance and direction to move to the next frame. It can be determined whether or not to go out of the area (monitoring room (C) 804).

  Next, the entrance / exit information of the monitoring room (C) 804 is read (step S1005). The entrance / exit information is the location information of the entrance / exit provided in the monitoring room (C) 804, and means the location information of the door (C) 816 in this embodiment.

  Next, it is determined whether the monitoring object 902 has exited the monitoring room (C) 804 through the door (C) 816. If the user leaves the monitoring room (C) 804, the process proceeds to step S1007, and if not, the process proceeds to step S1001 (step S1006). When the movement destination predicted by the next frame of the predicted motion vector (MVP) 912 of the monitoring target 902 calculated in step S1003 passes through the door (C) 816 provided in the monitoring room (C) 804 In addition, it can be determined that the monitoring target 902 has left the monitoring room (C) 804. Further, when the movement destination predicted by the predicted motion vector (MVP) 912 of the monitoring target 902 does not pass through the door (C) 816, it can be determined that the monitoring target 902 does not leave the monitoring room (C) 804. . When the monitoring object 902 leaves the monitoring room (C) 804, an alarm is displayed on the monitor 106 and the operator is notified of this (step S1007).

  Next, the monitoring area is changed to the movement destination of the monitoring target 902 (step S1008). The door (C) 816 is connected to the shared space 807, and it can be seen that the monitoring object 902 has moved to the shared space 807 via the door (C) 816. In addition, the next monitoring area from the position of the door (C) 816 in the shared space 807 is captured by the camera 101-8, the camera 101-9, and the camera 101-10 that photograph the periphery of the door (C) 816. It can be determined that the cutout area 913 is desirable.

  The compressed composite image of the next monitoring area is read from the recording unit 213 (step S1009). That is, the compressed composite image of the cutout area 913 that becomes the next monitoring area is read from the recording unit 213. The compressed composite image of the next monitoring area read in step S1009 is expanded and displayed on the monitor 106 (step S1010).

  As described above, according to the present embodiment, a plurality of cameras 101 are installed on the entire floor of a building to be monitored so that respective shooting ranges overlap each other, and images taken by these cameras 101 are displayed in a room. Every time or every specific shooting location, generate a composite image seen from above for each room or shooting location, calculate the predicted motion vector to predict the destination to be monitored, Information and the installation information of the camera 101 are used to predict and display a composite image of the movement destination of the monitoring target. Therefore, when tracking the movement of the monitoring target across a plurality of rooms, be aware of the structure of the room. Therefore, it is not necessary to switch the monitoring video sequentially, and the monitoring target can be traced very easily.

(Embodiment 3)
Next, a third embodiment of the present invention will be described. FIG. 11 is a diagram for explaining the calling process of images taken by each camera from the composite image in the security system according to Embodiment 3 of the present invention. The composite image 1101 is a piece of surveillance video output from the camera 101-1, the camera 101-2, the camera 101-3, and the camera 101-4 as seen from above the room generated by the image composition unit 204. It is a composite image. The section (A) 1102 is a part in which the monitoring image captured by the camera 101-2 is reflected, and the section (B) 1103 is a part in which the monitoring image captured by the camera 101-1 is reflected. A section (C) 1104 is a part in which a monitoring image taken by the camera 101-4 is reflected, and a section (D) 1105 is a part in which a monitoring image taken by the camera 101-3 is reflected. is there.

  The monitoring video 1106 captured by the camera 101-2 is an image corresponding to the segment (A) 1102 of the composite image 1101, and includes more detailed information than the segment (A) 1102. The monitoring video 1107 captured by the camera 101-1 is an image corresponding to the segment (B) 1103 of the composite image 1101, and includes more detailed information than the segment (B) 1103. The monitoring video 1108 photographed by the camera 101-4 is an image corresponding to the section (C) 1104 of the composite image 1101, and includes more detailed information than the section (C) 1104. A monitoring video 1109 photographed by the camera 101-3 is an image corresponding to the section (D) 1105 of the composite image 1101, and includes more detailed information than the section (D) 1105.

  The position information 1110 of the monitoring video 1106 photographed by the camera 101-2 has information regarding the photographing range of the camera 101-1. The position information 1111 of the monitoring video 1107 captured by the camera 101-1 has information regarding the imaging range of the camera 101-2. The position information 1112 of the monitoring video 1108 photographed by the camera 101-4 has information related to the photographing range of the camera 101-3. The position information 1113 of the monitoring video 1109 photographed by the camera 101-3 has information related to the photographing range of the camera 101-4. The position information 1110, 1111, 1112, 1113 is stored in the floor information management unit 205.

  Next, the operation of the present embodiment will be described. FIG. 12 is a flowchart for explaining the operation of the present embodiment. In the figure, the control unit 207 reads the compressed composite image stored in the recording unit 213, decompresses it with the image decompression unit 215, and the display control unit 216 updates the composite image displayed on the monitor 106 (step S1201). The composite image is updated at regular time intervals. It is determined whether or not the operator operates the input unit 107 and inputs an instruction for displaying in detail the specific area displayed on the monitor 106 (step S1202). If there is an instruction to display details, the process proceeds to step S1203. If there is no instruction, the process proceeds to step S1201.

  The control unit 207 recognizes which section of the composite image 1101 displayed on the monitor 106 corresponds to the area where the operator has instructed the detailed display (step S1203). The camera 101 that captures the image area is identified from the image areas obtained from the processing result of step S1203 (step S1204). In the case of the composite image 1101, the segment (A) 1102 is synthesized based on the monitoring image 1106 taken by the camera 101-2, and the segment (B) 1103 is a surveillance image taken by the camera 101-1. The image is synthesized based on the image 1107, the image segment (C) 1104 is composed based on the monitoring image 1108 captured by the camera 101-4, and the image segment (D) 1105 is captured by the camera 101-3. The images are synthesized based on the monitored image 1109. These pieces of information are stored in the floor information management unit 205. The monitoring image 1107 photographed by the camera 101-1 is captured by the position information 1111 of the camera 101-1, the monitoring image 1106 photographed by the camera 101-2 is photographed by the position information 1110 of the camera 101-2, and the camera 101-3. The position information 1113 of the camera 101-3 corresponds to the monitored image 1109, and the position information 1112 of the camera 101-4 corresponds to the monitoring image 1108 captured by the camera 101-4.

  The control unit 207 determines the shooting time of the image of the camera 101 determined in step S1204 to be displayed next on the monitor 106 based on the shooting time of the composite image currently displayed on the monitor 106 (step S1205). Normally, the monitoring video imaged by the corresponding camera 101 should be displayed on the monitor 106 one frame after the imaging time of the composite image currently displayed on the monitor 106. The control unit 207 inputs a control signal to the database search unit 211 based on the camera 101 determined in step S1204 and the shooting time determined in step S1205. The database search unit 211 searches for and reads out the corresponding compressed image data from the recording unit 213 (step S1206). The compressed image read by the processing in step S1206 is expanded by the image expansion unit 215 (step S1207). The image expanded in step S1207 is input to the display control unit 216. The display control unit 216 displays the monitoring video captured by the camera corresponding to the monitor 106 (step S1208).

  As described above, according to the present embodiment, monitoring information from a plurality of cameras 101 can be easily grasped by viewing a single composite image, and more detailed information can be viewed. The corresponding monitoring video can be easily viewed, and the labor for monitoring can be greatly reduced.

(Embodiment 4)
Next, a fourth embodiment of the present invention will be described. FIG. 13 is a block diagram showing a schematic configuration of the image composition unit of the security system according to Embodiment 4 of the present invention. The image composition unit 204A of the security system according to the present embodiment has a function of detecting the head / face. An image input 1301 is an output of the image expansion unit 203. A frame memory (A) 1302 stores images from the camera 101-1 for a plurality of frames. A frame memory (B) 1303 stores images from the camera 101-2 for a plurality of frames. A frame memory (C) 1304 stores an image from the camera 101-3 for a plurality of frames. A frame memory (D) 1305 stores an image from the camera 101-4 for a plurality of frames.

  The motion detection unit (A) 1306 calculates the difference between the images of a plurality of frames stored in the frame memory (A) 1302 and inputs the motion information of the image captured by the camera 101-1 to the synthesis processing unit 1314. The head / face detection unit (A) 1307 detects the human head from the image stored in the frame memory (A) 1302 and inputs the result to the joint position determination unit 1316. Further, face detection is performed on the area determined as the head, and the result is input to the joint position determination unit 1316. The motion detection unit (B) 1308 calculates the difference between the images of a plurality of frames stored in the frame memory (B) 1303 and inputs the motion information of the image captured by the camera 101-2 to the synthesis processing unit 1314. The head / face detection unit (B) 1309 detects the human head from the image stored in the frame memory (B) 1303 and inputs the result to the joint position determination unit 1316. Further, face detection is performed on the area determined as the head, and the result is input to the joint position determination unit 1316.

  The motion detection unit (C) 1310 calculates the difference between the images of a plurality of frames stored in the frame memory (C) 1304, and inputs the motion information of the image captured by the camera 101-3 to the synthesis processing unit 1314. The head / face detection unit (C) 1311 detects the human head from the image stored in the frame memory (C) 1304 and inputs the result to the joint position determination unit 1316. Further, face detection is performed on the area determined as the head, and the result is input to the joint position determination unit 1316. The motion detection unit (D) 1312 calculates the difference between the images of a plurality of frames stored in the frame memory (D) 1305, and inputs the motion information of the image captured by the camera 101-4 to the synthesis processing unit 1314. The head / face detection unit (D) 1313 detects the human head from the image stored in the frame memory (D) 1305 and inputs the result to the joint position determination unit 1316. Further, face detection is performed on the area determined as the head, and the result is input to the joint position determination unit 1316.

  The composition processing unit 1314 is stored in each of the frame memory (A) 1302, the frame memory (B) 1303, the frame memory (C) 1304, and the frame memory (D) 1305 in accordance with the joint position stored in the joint position buffer 1318. Then, image synthesis processing is performed on the obtained image and output as a synthesized image output 1319. The change detection unit 1315 calculates the difference between the current composite image and the previous frame for each region, and determines which region of the composite image has changed when a preset threshold value is exceeded.

  The joint position determination unit 1316 includes an output result of the change detection unit 1315, a head / face detection unit (A) 1307, a head / face detection unit (B) 1309, a head / face detection unit (C) 1311, a head A joint for image synthesis is determined from the outputs of the face detection unit (D) 1313. The joint position moving unit 1317 moves the joint position of the composite image in accordance with the output of the joint position determining unit 1316 and outputs the movement result to the joint position buffer 1318. The joint position buffer 1318 holds the joint position output from the joint position moving unit 1317. A composite image output 1319 is an output of the composition processing unit 1314.

  The image composition processing in the image composition unit 204A will be described with reference to FIG. In the present embodiment, for the sake of simplicity, a description will be given at a place where two images are combined. A composite image (A) 1400 is a composite image obtained by combining an image captured by the camera 101-1 and an image captured by the camera (B) 101-2 in the monitoring room 400 with a joint (ABA) 1404 as a boundary. An image area (AA) 1401 is an area of an image captured by the camera 101-2. An image area (BA) 1402 is an area of an image captured by the camera 101-1. The monitoring target (A) 1403 is present near the intermediate line (A) 1405, and an image with the face facing the front is displayed. A seam (ABA) 1404 is a seam when the images photographed by the camera 101-1 and the camera 101-2 are combined into one image. The intermediate line is a line that is equidistant from two cameras that capture an image.

  The composite image (B) 1406 is a composite image obtained by combining the image captured by the camera 101-1 and the image captured by the camera 101-2 in the monitoring room 400 with the joint (ABB) 1410 as a boundary. An image area (AB) 1407 is an area of an image captured by the camera 101-1. An image area (BB) 1408 is an area of an image captured by the camera 101-1. The monitoring target (B) 1409 exists near the intermediate line (B) 1411 and an image facing the back is displayed. A seam (ABB) 1410 is a seam when the images photographed by the camera 101-1 and the camera 101-2 are combined into one image.

  Next, the operation of the present embodiment will be described. FIG. 15 is a flowchart for explaining the operation of the present embodiment. In the figure, the monitoring image recording apparatus 105 includes four cameras (camera 101-1, camera 101-2, camera 101-3, and camera 101-4) installed in a monitoring room 400. Connected through. Images captured by the cameras 101-1 to 101-4 are transmitted as compressed image data, and are sorted by the sorting unit 202 for each time and each camera. Next, the image is expanded by the image expansion unit 203 and input to the image composition unit 204.

  For the monitoring video output from each camera (camera 101-1, camera 101-2, camera 101-3, camera 101-4) installed in the monitoring room 400, each motion detector (motion detector (A ) 1306, the motion detection unit (B) 1308, the motion detection unit (C) 1310, and the motion detection unit (D) 1312) perform motion detection processing (step S1501). Each head / face detection unit (head) is output to the monitoring video output from each camera (camera 101-1, camera 101-2, camera 101-3, camera 101-4) installed in the monitoring room 400. The face detection unit (A) 1307, the head / face detection unit (B) 1309, the head / face detection unit (C) 1311, and the head / face detection unit (D) 1312) serve as the head / face detection unit. This is performed (step S1502). Here, head detection is a type of image recognition processing that detects whether there is an object having the shape of a human head with respect to a monitoring video. The face detection process is a kind of image recognition process for detecting whether a human face is present in the area detected by the head detection process.

  Next, image composition processing is performed on the monitoring video input from each camera (camera 101-1, camera 101-2, camera 101-3, camera 101-4) installed in the monitoring room 400, and the monitoring room One composite image as seen from above 400 is generated (step S1503). Next, the current composite image generated in step S1503 is compared with the composite image generated in the previous frame, and the difference is calculated (step S1504). The result of the composite image change detection process calculated in step S1504 is compared with a preset threshold value TH. If the threshold value TH is exceeded, the process proceeds to step S1506, and if not, the process proceeds to step S1511. The seam position is stored (step S1505). When a large change that exceeds the threshold TH occurs in the composite image, it is necessary to move the joint of the composite image to an appropriate position.

  A motion is detected in the monitoring video from which camera from the motion detection result for the monitoring video shot by each camera (camera 101-1, camera 101-2, camera 101-3, camera 101-4) detected in step S1501. It is determined whether it has been done (step S1506). Monitoring from which camera based on the results of head detection and face detection for the monitoring video imaged by each camera (camera 101-1, camera 101-2, camera 101-3, camera 101-4) detected in step S1502. It is determined whether a head and a face are detected in the video (step S1507).

  The joint position of the image to be moved when generating the composite image is determined from the processing result in step S1506 and the processing result in step S1507 (step S1508). The case of the composite image (A) 1400 will be described. When the area where the motion is detected is in the area (AA) 1401, the joint to be moved is the joint (ABA) 1404. When the area where the head and face are detected is in the area (AA) 1401, the front of the monitoring target (A) 1403 is displayed, and the current seam position (ABA) 1404 is in the correct position. It can be judged that there is.

  Next, the case of the composite image (B) 1406 will be described. When the area where the motion is detected is in the block (BB) 1408, the seam to be moved is the seam (ABB) 1410. Here, when the area where the head is detected is in the block (BB) 1408 and no face is detected, it can be seen that the back of the monitoring target (B) 1409 is displayed, and the current seam position (ABB) It can be determined that 1410 needs to be moved. At this time, the direction in which the current seam (ABB) 1410 is moved is the moving direction 1412. By moving the seam (ABB) 1410 in the seam movement direction 1412, the seam (ABB) 1410 is moved so that the front of the monitoring target (B) 1409 is synthesized.

  Next, the current joint position value is read from the joint position buffer 1318 (step S1509). In the case of the composite image (B) 1406, position information of the joint (ABB) 1410 between the image segment (AB) 1407 and the image segment (BB) 1408 is stored. The joint position moving unit 1317 moves the joint position read from the joint position buffer 1318 from the current position in the direction perpendicular to the joint according to the processing result of step S1508 (step S1510). After moving the joint, the process proceeds to step S1503, and the processing from step S1503 to step S1510 is repeated until the difference between the current composite image and the composite image after moving the joint falls within the threshold value TH. When the difference between the composite image and the composite image after moving the joint falls within the threshold TH, the moved joint position is stored in the joint position buffer 1318 (step S1511).

  As described above, according to the present embodiment, the face to be monitored is always included in the synthesized image by controlling the moving direction of the seam during the image synthesizing process using the result of the face / head detection process. It is possible to display the information on the screen, and it is possible to greatly reduce the labor for specifying the monitoring target when an abnormality occurs.

  The present invention can surely check all images taken by all cameras installed indoors, and can easily and reliably find an image that can determine the cause of an abnormality when any abnormality occurs. And can be applied to a security system and the like.

The block diagram which shows schematic structure of the security system which concerns on Embodiment 1 of this invention. 1 is a block diagram showing a schematic configuration of the monitoring image recording apparatus of FIG. The block diagram which shows schematic structure of the image synthetic | combination part of FIG. The figure which shows the example of installation in the room | chamber interior of the camera in the security system which concerns on Embodiment 1 of this invention. The figure which shows an example of the synthesized image produced | generated by the monitoring image recording device of the security system which concerns on Embodiment 1 of this invention. Flowchart for explaining composite image processing in the monitoring image recording apparatus of the security system according to the first embodiment of the present invention. The figure which shows the example of a movement of the joint of the synthesized image in the monitoring image recording device of the security system which concerns on Embodiment 1 of this invention The figure which shows the example of installation of the camera of the security system which concerns on Embodiment 2 of this invention. The figure for demonstrating the image call process using the motion vector in the image recording apparatus for monitoring of the security system which concerns on Embodiment 2 of this invention. Flowchart for explaining an image calling process using a motion vector in the monitoring image recording apparatus of the security system according to the second embodiment of the present invention. The figure for demonstrating the call processing of the image which each camera image | photographed from the synthesized image in the security system which concerns on Embodiment 2 of this invention. Flowchart for explaining the operation of the monitoring image recording apparatus of the security system according to the third embodiment of the present invention. The block diagram which shows schematic structure of the image composition part of the security system which concerns on Embodiment 4 of this invention. The figure which shows the example of a movement of the joint of the synthesized image in the monitoring image recording device of the security system concerning Embodiment 4 of this invention Flowchart for explaining the operation of the monitoring image recording apparatus of the security system according to the fourth embodiment of the present invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 Wired camera 102 Wireless camera 103 Wireless receiver 104 Network 105 Image recording apparatus for monitoring 106 Monitor 107 Input unit 201 Network I / F
202 Sorting unit 203 Image decompression unit 204 Image composition unit 205 Floor information management unit 207 Control unit 209 Image compression unit 210 Database generation unit 211 Database search unit 212 Data write control unit 213 Recording unit 214 Data read control unit 215 Image decompression unit 216 Display Control unit 302 Frame memory (A)
303 Frame memory (B)
304 Frame memory (C)
305 Frame memory (D)
306 Motion detection unit (A)
307 Motion detection unit (B)
308 Motion detection unit (C)
309 Motion detection unit (D)
310 Combining Processing Unit 311 Change Detection Unit 312 Seam Position Determination Unit 313 Seam Position Moving Unit 314 Seam Position Buffer 400 Monitoring Room 801 Monitoring Floor 802 Monitoring Room (A)
803 Monitoring room (B)
804 Monitoring room (C)
805 Monitoring room (D)
806 Monitoring room (E)
807 Shared space 814 Door (A)
815 Door (B)
816 Door (C)
817 Door (D)
818 Door (E)
1302 Frame memory (A)
1303 Frame memory (B)
1304 Frame memory (C)
1305 Frame memory (D)
1306 Motion Detection Unit (A)
1307 Head / Face Detection Unit (A)
1308 Motion Detection Unit (B)
1309 Head / Face Detection Unit (B)
1310 Motion Detection Unit (C)
1311 Head / Face Detection Unit (C)
1312 Motion Detection Unit (D)
1313 Head / Face Detection Unit (D)
1314 Combining processing unit 1315 Change detecting unit 1316 Seam position determining unit 1317 Seam position moving unit 1318 Seam position buffer

Claims (9)

In the monitoring image recording apparatus for capturing and recording images taken by a plurality of cameras connected to a network via the network,
Sorting means for aligning a plurality of images taken by a plurality of the cameras at the same time;
Motion detection is performed on each of a plurality of images at the same time that have been aligned by the sorting means, and a seam position for combining the plurality of images as one image is selected from the result, and the selected seam position is selected. Image synthesizing means for synthesizing images while evaluating the amount of change in the synthesized image by moving
Image compression means for compressing the composite image synthesized by the image synthesis means and a plurality of images that are the basis of the composite image;
Database generating means for generating a database associating the composite image compressed by the image compressing means with a plurality of the images that are the basis of the composite image;
Recording means for recording the database generated by the database generating means;
Database search means for searching a database recorded in the recording means according to an operator's instruction;
Image decompression means for decompressing the compressed image read from the recording means according to the search result by the database retrieval means and returning it to the original state before compression;
A monitoring image recording apparatus comprising: The image composition means includes
A plurality of motion detection means for detecting motion from images captured by each of the plurality of cameras;
Change detection means for detecting a difference between the current composite image and the composite image generated immediately before;
The difference detected by the change detection means is compared with a predetermined threshold value, and when the threshold value is exceeded, it is determined from which motion detection result of each of the plurality of motion detection means which image is captured by which camera. A seam position determining means for determining a seam position of an image to be moved when generating a composite image from the determination result;
Seam position holding means for holding the seam position of the image;
Seam position moving means for moving the current seam position held by the seam position holding means to the seam position determined by the seam position determining means,
After the joint position of the current composite image is moved to the joint position determined by the joint position determining means, image synthesis is performed until the difference between the composite image after the joint position movement and the current composite image falls within the threshold value. When the difference between the current composite image and the composite image after moving the joint position falls within the threshold value, each process of change detection, joint position determination, and joint position movement is repeatedly performed. The monitoring image recording apparatus according to claim 1, wherein the position is stored in the joint position holding unit. The plurality of cameras are installed in a room to be monitored so that respective shooting ranges overlap each other,
The monitoring image recording apparatus according to claim 1, wherein the image synthesizing unit converts a plurality of images taken at the same time to generate a synthesized image when the room is viewed from above. The plurality of cameras are installed on the entire floor of the building to be monitored so that the respective shooting ranges overlap each other,
The image synthesizing unit performs image conversion on images captured by the plurality of cameras for each room or each specific shooting location, and generates a synthesized image viewed from above for each room or shooting location. Item 3. The monitoring image recording apparatus according to Item 2. A floor information management means for managing position information of each of the plurality of cameras;
The database generation unit generates a plurality of databases managed by the floor information management unit when generating a database in which the composite image compressed by the image compression unit and a plurality of images that are the basis of the composite image are generated. The monitoring image recording apparatus according to any one of claims 1 to 4, wherein a database is generated so that each of a plurality of images before composition can be called by utilizing position information of each of the cameras. The floor information management means manages information related to the structure of the building floor including the doorway and windows of the room to be monitored,
The database generation unit can call a composite image over a plurality of monitoring areas by utilizing position information of each of the plurality of cameras managed by the floor information management unit and information on the structure of the building floor. The monitoring image recording apparatus according to claim 5, wherein the database is generated so as to be able to.   When monitoring is performed in a specific room, each of the plurality of motion detection units of the image composition unit calculates an expected motion vector of the monitoring target, and the predicted motion of the monitoring target calculated by each of the motion detection units A control unit that performs control for predicting a movement destination of the monitoring target based on a vector and reading a composite image of the predicted movement destination area when the monitoring target is predicted to go out of the room. Item 7. The monitoring image recording apparatus according to any one of Items 6 to 7. A head / face detecting means for detecting a human head and face with respect to images taken by the plurality of cameras;
The seam position determining means determines the joint position of image synthesis so that a human face is displayed in a composite image when the head / face detection means detects a human face. The monitoring image recording apparatus according to any one of claims 7 to 9. A plurality of cameras connected to the network directly or via a wireless transceiver;
A monitoring image recording apparatus according to any one of claims 1 to 8,
A monitor for displaying an image recorded in the monitoring image recording device;
An input unit for giving an operation command to the monitoring image recording apparatus;
Security system with

Images