JP2002092751A - Monitoring system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a monitoring system capable of taking over to another monitoring camera when an intruding object comes out of a visual field of a monitoring camera under an environment that plural monitoring cameras are located. SOLUTION: An intruding object detecting means 20 detects the intruding object by comparing an image obtained by the monitoring camera 10 and an image having no intruding object which is kept in an image storage means 17. An object position calculating means 21 calculates a position of the detected intruding object in actual environment. An environment map data 18 keeps a map of whole environment wherein the monitoring camera is located and the visual field of the monitoring camera. A visual field determination means 22 determines the monitoring camera capable of supervising the intruding object with the usage of the position of the intruding object in the actual environment and the environment map data 18. A turning device 14 receives a turning command from a turning position determination means 23 so that an image of the intruding object can be produced near the center of an optical axis of the camera. An image of the monitoring camera selected by the visual field determination means 22 is outputted to a monitor 26.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a surveillance system using a surveillance camera that can take over surveillance to another surveillance camera in an environment where a plurality of surveillance cameras are installed, when an intruding object goes out of the field of view of the surveillance camera. About.

[0002]

2. Description of the Related Art In recent years, technical development relating to automation of a surveillance system using a surveillance camera has been advanced. As this type of monitoring apparatus, a technique described in Japanese Patent Application Laid-Open No. Hei 8-205513 is known. In the monitoring device described in this publication, when an abnormality in the screen of the monitoring camera is detected, camera control, turning device control, switcher control, and the like are performed even if there is no operator of the monitoring device, and video recording of an intruding object is performed. It is configured to be.

[0003]

However, in the conventional surveillance device, if the intruding object goes out of the field of view monitored by the surveillance device, in order to continue monitoring the intruding object, the number of surveillance cameras installed is required. The operator, who was familiar with the position and the field of view of the individual surveillance cameras, had to decide which surveillance camera to use next.

An object of the present invention is to provide a surveillance system capable of automatically continuing surveillance by another surveillance camera even when an object to be monitored is out of the field of view of a surveillance camera. And

[0005]

In order to achieve the above object, the present invention provides a video storage device for obtaining video information, storing at least two monitoring cameras in the same area, and storing the video information obtained by the monitoring cameras. Means, an intruding object detecting means for detecting an intruding object based on the video information acquired by the monitoring camera, an object position calculating means for calculating a position of the intruding object detected by the intruding object detecting means, and a monitoring environment. Environmental map data on the field of view of the surveillance camera,
Based on the environment map data and the position of the intruding object, a visual field determining means for calculating which visual field of the surveillance camera enters, and an image of the selected surveillance camera based on an output of the visual field determining means. A video signal selection control means for selecting, and video signal switching means for outputting a video from the selected surveillance camera to a monitor according to a selection control signal from the video signal selection control means.

Further, the present invention provides a single surveillance camera for acquiring video information, video storage means for storing the video information acquired by the surveillance camera, and an intruding object based on the video information acquired by the surveillance camera. Object position detecting means for detecting the position of the intruding object detected by the intruding object detecting means, environmental map data on the monitoring environment and the field of view of the monitoring camera, and the environmental map data. A visual field determining means for calculating whether to enter the visual field of the surveillance camera based on the position of the intruding object and an input / output means for inputting / outputting data of the visual field determining means. At least two sets of monitoring devices are provided so as to be monitored, and the selected monitoring device is determined based on a visual field judgment output from each input / output means in the intruding object monitoring device. A video signal selection control means for selecting a video camera, a video switching means for outputting an image from the selected surveillance camera by selection control means from said video signal selection control means on the monitor, in constructed.

Further, according to the present invention, the video from the surveillance camera, which initially detects that an abnormal object has entered, starts monitoring the information in the object position calculating means and the visual field determining means. Next, when the approaching object is moving to a position outside the surveillance area of the surveillance camera, the visual field determination unit selects a surveillance camera that can newly be monitored. Since the turning position is determined and the data is sent to the camera,
The surveillance camera can be monitored. Surveillance is taken over by the surveillance camera selected in this way.

Further, according to the present invention, since necessary and sufficient devices for monitoring are combined as an approaching object monitoring device, processing operations can be distributed.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The embodiments of the present invention shown in the drawings will be described below.

(First Embodiment) FIG. 1 shows a first embodiment of the present invention.
It is a block diagram which shows the structure of the monitoring system in embodiment. In FIG. 1, the surveillance system of the present invention includes surveillance cameras 10 to 13, turning means 14, A / D conversion means 16, image storage means 17, environmental map data 18, intruding object detection means 20,
The apparatus includes an object position calculation unit 21, a visual field determination unit 22, a turning position determination unit 23, a video signal selection control unit 24, a video switching unit 25, and a video monitor 26.

Each of these will be described below. Surveillance camera
10, 11, 12, 13 are for obtaining video information, A / D conversion means
Reference numeral 16 performs A / D conversion of an image obtained by the monitoring camera 10 or the like. The image storage means 17 holds the A / D converted image data. The environmental map data 18 is environmental map data on the surveillance environment and the visual field of the surveillance camera 10 and the like. The intruding object detection means 20 detects an intruding object based on the monitoring environment and image data from the monitoring camera 10 and the like. The object position calculating means 21 calculates the position of the detected intruding object. Field of view judgment means 22
Calculates, from the position of the intruding object, which camera can be monitored, that is, what camera is in view.

The turning position determining means 23 calculates a turning position at which the surveillance camera selected based on the visual field determining means 22 can be monitored. The turning means 14 turns the monitoring camera in the direction calculated by the turning position determining means 23. The video signal selection control means 24 selects the video of the selected monitoring camera based on the visual field determination means 22. The image switching means 25 outputs the selected image to the monitor 26.

Next, the operation of the above embodiment will be described with reference to the top view of FIG. 2, the same components as those in FIG. 1 are denoted by the same reference numerals. Entry object C is a wall
It is assumed that the vehicle has entered through a gap of 19.

First, video information is obtained using the monitoring camera 10. The video information is converted to digital image data via A / D conversion means 16. The image storage means 17 stores in advance image data in a state where the intruding object C is not shown.
The image data not showing the intruding object C is acquired, for example, by the installer confirming that there is no intruding object when setting the camera, and is stored in the image storage means 17.

The intruding object detecting means 20 converts the image data stored in the image storing means 17 in which the intruding object C is not reflected beforehand and the image information obtained by the current surveillance camera 10 into A / D converting means. The difference between the luminance and color data from the current image data converted into digital data by step 16 is calculated, and the presence of an intruding object is determined on condition that the difference between the calculated luminance and color data is equal to or greater than a certain threshold. To detect.

The installation position of the monitoring camera 10 is fixed and known. Although the direction of the camera is determined by the operation of the turning device 14, the operation of the turning device 14 is determined by the turning position determining means 23, so that the operation angle and the like of the turning device 14 are known. When the position and orientation of the monitoring camera 10 are known,
When the camera is considered as a pinhole model, one point on the imaging surface and one point on a plane (for example, floor) in the real environment are 1
There is a one-to-one correspondence. That is, by using the principle of triangulation, the position in the real environment can be calculated using the position on the camera image.

For this reason, the object position calculating means 21 uses the principle of triangulation from the position on the image to calculate the intruding object detecting means 20.
The position of the intruding object detected in the real environment is calculated.
If the lowest point of the intruding object detected at this time is selected as a representative point, the position on the floor in the real environment can be calculated.

For this reason, two points whose distances are known for a certain surveillance camera 10 are determined in advance, and an angle between a straight line connecting the points and the intruding object and a straight line connecting the two points is measured on an image. Then, the operation may be performed.

Next, a turning command is issued from the turning position determining means 23 so that the intruding object can be imaged near the center of the optical axis of the camera. The turning position of the surveillance camera 10 is calculated by performing the inverse operation of the triangulation calculated by the object position calculating means 21 from the position of P1 in the real environment of the intruding object. When the surveillance camera 10 turns to the turning position, the surveillance camera 10 looks directly at the intruding object, and an image is obtained near the center.

When the entering object C moves from the position P1 and reaches P2, the position of P2 is assumed to be a position that can be monitored by the monitoring camera 11. When the monitoring camera 10 detects the position of P2 and associates it with the map data 18, it is determined that the position of P2 can be monitored by the monitoring camera 11. Then, the turning amount of the surveillance camera 11 is calculated by the turning position calculating means 21 and the turning camera 14 turns the monitoring camera 11. That is, at the position of P2, monitoring can be performed from both the monitoring cameras 10 and 11.

Next, when the approaching object C moves to the position P3, monitoring by the monitoring camera 10 becomes impossible, and monitoring by the monitoring camera 11 is continued. That is, camera 10
Monitoring is taken over from to 11.

By performing such an operation, in the environment where a plurality of surveillance cameras 10, 11, 12, and 13 are installed, when an intruding object goes out of the field of view of one surveillance camera, the surveillance is monitored by another surveillance camera. You can take over.

(Second Embodiment) FIG. 3 shows a second embodiment of the present invention.
It is a block diagram which shows the structure of the monitoring system in embodiment. In FIG. 3, the monitoring camera 10, the A / D conversion means 16, the image storage means 17, the environment map data 18, the intruding object detection means 20, the object position calculation means 21, the visual field judgment means 22, the turning position determination means 23, the video signal The selection control means 24, the video signal switching means 25, and the monitor 26 correspond to the same elements as those shown in FIG.

In FIG. 3, an input / output means 27 inputs and outputs turning position information, visual field determination information, and video signals. As the intruding object monitoring device 90, the components from the monitoring camera 10 to the visual field determining means 22 are collectively shown. Therefore, intruding object monitoring devices having the same configuration as the intruding object monitoring device 90 are denoted by 91 to 93.

The operation of each of the intruding object monitoring devices 90 to 93 is the same as the operation of elements in the same range as shown in FIG. Then, input / output of the turning position information, the visual field determination information, and the video signal is performed via the input / output device 27. Further, the video of the surveillance camera selected based on the visual field determination information output from each input / output device 27 of each of the intruding object monitoring devices 90 to 93 is displayed by video signal selection control means.
Perform at 24.

In the configuration shown in FIG. 3, each intruding object monitoring device performs processing from acquisition of an image to determination of a turning position in a separate device, so that when a failure occurs, it is possible to easily deal with it.

Next, a specific configuration of the object position calculating means and the visual field judging means in each of the above devices will be described as an embodiment. FIG. 4 shows an example of the object position calculation means 21. The image storage means 30 stores an image. The position calculating means 31 calculates the position of the intruding object. The moving direction calculating means 32 calculates the moving direction of the intruding object. The image storage means 30 and the position calculation means 31 store at least two pieces of image data with a time difference, and use them to calculate the current position of the intruding object using, for example, the concept of triangulation. .

Next, the moving direction of the object is calculated by looking at the position of the intruding object calculated by the moving direction calculating means 32 in chronological order. According to this configuration example, since the moving direction of the intruding object is known, it is possible to manage that the intruding object enters and exits from the visual field.

FIG. 5 is a block diagram showing another object position calculating means. 5, an image storage unit 30 and a position calculation unit 31 are the same as those in FIG. The speed vector calculation means 33 calculates a speed vector of the intruding object. The time difference measuring means 34 measures the time between the images of the time-series images stored in the image storage means 30.

The image storing means 30 and the position calculating means 31 store the image data at least two times with a time difference as described above.
The current position of the intruding object is calculated by using the stored information, for example, using the concept of triangulation.

The speed vector calculating means 34 looks at the position of the intruding object calculated by the position calculating means 33 in chronological order and uses the moving direction and the moving distance, and the time difference measuring means 34 uses the time between the time-series images. , Calculate the velocity vector of the intruding object. With such a configuration, when an intruding object comes out of the field of view, it is possible to predict the time required to enter the next field of view.

In another embodiment shown in FIG. 6, the object position calculating means 35 calculates where the position on the image is located in the real environment when the intruding object is in the visual field. The look-up table 36 stores in advance where the position on the image is located in the real environment.

The position on the image of the intruder detected by the intruder detecting unit 21 is input to the object position calculator 35. Lookup table
36 stores the correspondence between the position on the image and the position on the monitoring environment.

For example, if the position on the monitoring environment is (X1, Y1) when the position on the image is (x1, y1), the look-up table 36 stores (X1, Y1) at the address (x1, y1). Is held. The object position calculating means 35 uses (x1, y1) as an argument and (X1, Y1).
Return 1). The look-up table 36 is created when the surveillance camera is installed.

In FIG. 7, the pattern storage means 37 stores a human model pattern. Pattern comparison means
38 performs matching between the model pattern and the image data. The human judging means 39 detects whether or not the intruding object is a human from the comparison result of the comparing means. Image storage means 40
Holds image data. The position calculating means 41 calculates the position of the detected intruding object.

The image storage means 40 receives a difference image for detecting an intruding object as an image obtained by the intruding object detecting means 20 shown in FIG. This difference image is (image not showing the intruding object)-(image showing the current intruding object), and this difference image is input to the image storage means 40.

In the difference image, there is almost no difference in portions other than the intruding object, and only the intruding object portion has a value other than "0".
At this time, a portion having a value other than “0” in the difference image indicates the outer shape of the intruding object. The pattern storage means 37 stores the external shape of the surveillance camera 10 when a person is reflected on the surveillance camera 10. In general, the external shape of a human has a vertically long shape having a certain aspect ratio when viewed from the horizontal direction.
The upper part, which hits the head, is thinner than the others, and the lower part, which hits the feet, can split in two. Looking diagonally from above,
The aspect ratio of the outer shape changes. The pattern storage means 37 holds a human outline conforming to the above possibilities.

Next, the pattern comparing means 38 compares the portion having the value of the difference image with the external shape held in the pattern storing means 37. For example, the value close to “0” in the difference image is ×, the value of the other (invading object) is ○, and the human outline pattern is also ○. Calculate the number of failed matches at the location.

If the number of matching failures is equal to or less than a certain threshold value, the intruding object is determined to be a human by the human determining means 39.
The position calculating means 41 calculates the position of the intruding object as described above. The human determination means 39 determines whether the intruding object is a human from the comparison result of the pattern comparison means 38,
The judgment result is added to the position information. The human determination means 39 outputs information on the position of the intruding object and whether the intruding object is a human.

Next, FIG. 8 shows an example of a specific configuration for determining the visual field. In FIG. 8, when the intruding object goes out of the field of view, the judging means 42 judges from the position out of the field of view that there is a possibility that the next supervised camera may enter the field of view.
The look-up table 43 is a table in which, when the intruding object goes out of the field of view, the surveillance camera which is likely to enter the field of view next from the position out of the field of view is stored in advance.

The position of the intruding object calculated by the object information detecting means 21 is input to the judging means 42. The lookup table 43 stores the correspondence between the position of the intruding object and the camera that can be monitored. When the position of the intruding object is x, if the surveillance camera A can monitor by pointing in the direction (a1, a2), the look-up table 43 holds data (A, a1, a2) at the address x. I just want to. The judging means 42 returns (A, a1, a2) using x as an argument. The lookup table 43 may be created when the surveillance camera is installed.

The image storage means 17 shown in FIG. 1 stores image data in which an intruding object is not reflected in advance. At this time, as a method for acquiring image data in which an intruding object is not reflected, an operator needs In addition to the method of confirming that an intruding object is not reflected and holding the image in the image storage means as described above, the following method is also available.

That is, a plurality of image data is acquired at intervals of time, the difference between the luminance and the color data is calculated between the plurality of image data, and there is no or almost no difference between the calculated luminance and the color data. To make image data without an intruding object. In this case, it is possible to cope with aging of the monitoring environment and environmental changes such as time and weather.

In the above embodiment, the surveillance camera with the turning device has been described, but the same applies to a fixed camera.

[0045]

As described above, according to the present invention, in an environment where a plurality of surveillance cameras are installed, when an intruding object goes out of the field of view of the surveillance cameras, surveillance can be taken over by another surveillance camera, and It is possible to provide a monitoring system capable of capturing an object more accurately.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a monitoring system according to a first embodiment of the present invention;

FIG. 2 is a diagram for explaining the operation of the monitoring system shown in FIG. 1;

FIG. 3 is a block diagram showing a configuration of a monitoring system according to a second embodiment of the present invention;

FIG. 4 is a block diagram showing a configuration of an object position calculating unit in FIG. 1;

FIG. 5 is a block diagram illustrating a configuration of an object position calculating unit in FIG. 1;

FIG. 6 is a block diagram showing a configuration of an object position calculating unit in FIG. 1;

FIG. 7 is a block diagram showing a configuration of an object position calculating unit in FIG. 1;

FIG. 8 is a block diagram illustrating a configuration of a visual field determination unit in FIG. 1;

[Explanation of symbols]

 10 to 13 Surveillance camera 14 Turning means 16 A / D conversion means 17 Image storage means 18 Environmental map data 20 Intruding object detection means 21 Object position calculation means 22 Field of view judgment means 23 Turning position determination means 24 Video signal selection control means 25 Video signal Switching means 26 Monitor 27 Input / output means 30 Image storage means 31 Object position calculation means 32 Moving direction calculation means 33 Speed vector calculation means 34 Time difference measurement means 35 Object position calculation means 36, 43 Look-up table 37 Pattern storage means 38 Pattern comparison means 39 Human judgment means 40 Image storage means 41 Position calculation means 42 Field of view judgment means 90-93 Intruding object monitoring device

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5C022 AA01 AB62 AB65 AC27 5C054 AA02 CF05 CG06 EB05 FC01 FC12 FC13 FC15 FE22 FF01 GB01 HA19 5C084 AA01 AA06 BB01 BB31 CC16 DD11 GG18 GG42 GG51 GG54 A04A DD03 DD13 EE05 EE06 EE08 EE20 GG03 GG19 GG52 GG55 GG65

Claims (7)

[Claims] 1. A video information acquisition device, at least two monitoring cameras in the same area, video storage means for storing video information acquired by the monitoring camera, and intrusion based on the video information acquired by the monitoring camera. An intruding object detecting means for detecting an object; an object position calculating means for calculating a position of the intruding object detected by the intruding object detecting means; environmental map data on a monitoring environment and a field of view of the monitoring camera; Field-of-view judging means for calculating which surveillance camera enters the field of view based on data and the position of the intruding object, and an image for selecting an image of the selected surveillance camera based on an output of the field-of-view judging means Selection control means, and video signal switching means for outputting video from the selected surveillance camera to a monitor according to a selection control signal from the video selection control means Monitoring system characterized by comprising a. 2. A single surveillance camera for acquiring video information, a video storage unit for storing video information acquired by the surveillance camera, and an intrusion detecting an intruding object based on the video information acquired by the surveillance camera. Object detecting means; object position calculating means for calculating the position of the intruding object detected by the intruding object detecting means; environmental map data on the monitoring environment and the field of view of the monitoring camera; the environmental map data and the intruding object And an input / output means for inputting / outputting data of the visual field determining means, based on the position of the monitoring camera. A set of monitoring devices are provided so as to be capable of monitoring, and the monitoring camera selected based on the visual field judgment output from each input / output means in the intruding object monitoring device. A surveillance system comprising: video selection control means for selecting a video signal; and video switching means for outputting video from the surveillance camera selected by the selection control means from the video selection control means to a monitor. . 3. An object position calculating means, comprising: storage means for holding a plurality of image data having a time difference; position calculating means for calculating position information of an intruding object from the plurality of image data; The surveillance system according to claim 1, further comprising a moving direction calculating unit configured to calculate a moving direction of the intruding object from the position information having the following. 4. An object position calculating means, comprising: storage means for storing a plurality of image data having a time difference; position calculating means for calculating position information of an intruding object from the plurality of image data; The monitoring system according to claim 1, further comprising: a speed vector calculating unit configured to calculate a speed vector of the intruding object from the positional information having the time information and the time difference. 5. The apparatus according to claim 1, wherein said object position calculating means comprises a look-up table which stores in advance where in the real environment the position on the image is located. Monitoring system. 6. An object position calculating means, comprising: a position calculating means for calculating a position of a detected intruding object; an image storing means for storing image data; a pattern storing means for storing a human model pattern; A pattern comparing unit that matches the image data with the image data; and a human determining unit that detects whether the intruding object is a human based on an output of the pattern comparing unit. The monitoring system according to claim 1 or claim 2. 7. The visual field judging means comprises a look-up table in which, when an intruding object goes out of the field of view, the surveillance camera is likely to enter the next field of view from the position outside the field of view. 3. The monitoring system according to claim 1, wherein