JP5055962B2 - Intruder detection device - Google Patents

Description

  The present invention relates to an intruder detection device.

  Conventionally, an intruder is provided with a monitoring camera and an image processing unit in order to detect an intruder in the restricted area by analyzing the image of the surveillance camera installed in the restricted area such as facilities and roads. A technique for configuring a detection device is known. The surveillance camera is installed at a position overlooking the restricted entry area, and images the restricted entry area and its surroundings. The captured image is transmitted to the image processing unit. The image processing unit inputs the image of the surveillance camera, divides the input image into appropriate small areas, the background image acquired in advance for each small area and the corresponding small area image in the input image Correlation calculation is performed to find the temporal change point. Based on the calculation result of the correlation calculation, a small change area and a background candidate small area in the input image are detected, and adjacent change small areas are combined to determine a change detection area. Then, by determining that there is an intruder in this change detection area, the intruder is detected and the position in the input image is specified. In such an intruder detection device, normalized correlation is used for correlation calculation between a background image and an image of a corresponding small area in the input image, but the normalized correlation reacts to a change in the pattern of the image. Since it does not react to changes in overall brightness, it is an effective method in places where the brightness changes frequently due to sunlight, such as outdoors. An example of such an intruder detection device is disclosed in Patent Document 1 below.

JP 2004-282368 A

However, in the conventional intruder detection device as described above, the property of not reacting to the change in the brightness of the entire image is that the background is monochromatic and the detection object in the input image is monochromatic and does not move. Since the change in the image gradation value in the small area is the same as the change in the overall brightness, the change in the image in the small area cannot be detected. This problem becomes more prominent because the smaller the size of the small area, the larger the proportion of pixels in the small area corresponding to the place where the monochrome object is stopped becomes a single color. In this case, the following problem occurs as an intruder detection device.
(1) When a person wearing a single-colored clothing such as a suit stops in a place where there are few patterns such as an indoor wall, the person may not be detected.
(2) When a person wearing a single-colored outfit such as work clothes stops in a place with few patterns such as an outdoor road or lawn, the person may not be detected.
(3) If a single color object such as a cardboard box is left behind in a place with little pattern such as an outdoor road or lawn, the object may not be detected.

  In view of the above, an object of the present invention is to provide an intruder detection device that detects an intruder or the like in an area where entry is prohibited such as an important facility or road using an image of a surveillance camera.

An intruder detection device according to a first invention (corresponding to claim 1) for solving the above-mentioned problem is
In an intruder detection apparatus comprising a camera that captures an image and an image processing unit that processes an image captured by the camera,
Wherein the image processing unit,
Input means for inputting the image to the image processing unit;
A small area dividing means for dividing the image into small areas;
Correlation calculating means for calculating a temporal correlation between the small area and a background image previously captured by the camera corresponding to the small area;
Change detecting means for detecting the small area having a temporal change as a small change area;
A change region combining means that combines the change small regions into a change detection region;
A change area outline extracting means for extracting the outline of the change detection area as a change detection area outline;
A geometric feature amount evaluating means for evaluating whether or not the geometric feature amount of the change detection region outline is within a range of a geometric feature amount corresponding to a preset intruder;
Intruder detection means for determining whether there is an intruder in the change detection area based on the result of the evaluation ,
An intruder detection device further comprising an inner contour line deleting unit that deletes an inner contour line from the contour lines extracted from one of the change detection areas .

An intruder detection device according to a second invention (corresponding to claim 2) for solving the above problem is
In an intruder detection apparatus comprising a camera that captures an image and an image processing unit that processes an image captured by the camera,
Wherein the image processing unit,
Input means for inputting the image to the image processing unit;
A small area dividing means for dividing the image into a small area having a relatively large size around the outline in order to detect the outline of the detection object with a width;
Correlation calculating means for calculating a temporal correlation between the small area and a background image previously captured by the camera corresponding to the small area;
Change detecting means for detecting the small area having a temporal change as a small change area ;
A division changing small area setting means for setting a division changing small regions and further dividing the pre-Symbol changes subregion,
Change area combining means for determining a divided change detection area by combining adjacent divided change small areas;
A geometric feature amount evaluating means for evaluating whether or not the geometric feature amount of the divided change detection region is within a range of a geometric feature amount corresponding to a preset intruder;
Intruder detection means for determining whether there is an intruder in the change detection area based on the result of the evaluation ,
A change detection area filling processing unit is further provided which changes a portion of the background small area inside the division change detection area into a division change small area, thereby making a hole filling division change detection area clogged up to the inside. An intruder detection device characterized by that.

An intruder detection device according to a third invention (corresponding to claim 3) for solving the above-described problem is
In an intruder detection apparatus comprising a camera that captures an image and an image processing unit that processes an image captured by the camera,
Wherein the image processing unit,
Input means for inputting the image to the image processing unit;
A small area dividing means for dividing the image into a small area having a relatively large size around the outline in order to detect the outline of the detection object with a width;
Correlation calculating means for calculating a temporal correlation between the small area and a background image previously captured by the camera corresponding to the small area;
Change detecting means for detecting the small area having a temporal change as a small change area ;
A division changing small area setting means for setting a division changing small regions and further dividing the pre-Symbol changes subregion,
Change area combining means for determining a divided change detection area by combining adjacent divided change small areas;
A change area outline extracting means for extracting a division change detection area outline which is an outline of the division change detection area;
A geometric feature amount evaluating means for evaluating whether or not a geometric feature amount of the division change detection region contour line is within a range of a geometric feature amount corresponding to a preset intruder;
Intruder detection means for determining whether there is an intruder in the change detection area based on the result of the evaluation ,
An intruder detection apparatus further comprising an inner contour line deleting unit that deletes an inner contour line from contour lines extracted from one of the divided change detection areas .

In the fourth invention intruder detector according to (corresponding to claim 4), intruder detection apparatus according to the first invention or the third invention for solving the above problems, the change detection region contour or the divided change detection region intruder detecting apparatus further comprising a change subregion reconfigure Teite stage to set the reset change small region of the small region within the range surrounded by the contour line.

The present invention has the following effects.
(1) Even when a person dressed in a single color such as a suit stops in a place where there are few patterns such as an indoor wall, the person can be detected.
(2) Even when a person wearing a single-colored clothing such as work clothes stops on an outdoor road or lawn where there is little pattern, the person can be detected.
(3) Even when an object of a single color such as a cardboard box is left behind in a place with few patterns such as an outdoor road or lawn, the object can be detected.

  Embodiments of an intruder detection device according to the present invention will be described with reference to FIGS. FIG. 1 is a configuration diagram of an intruder detection device according to the first embodiment, FIG. 2 is a diagram showing a small region and an outline according to the present invention, and FIG. 3 is a diagram showing a small change region according to the first embodiment. FIG. 4 is an apparatus configuration diagram of the intruder detection apparatus according to the second embodiment, FIG. 5 is an apparatus configuration diagram of the intruder detection apparatus according to the third embodiment, and FIG. FIG. 7 is a diagram illustrating a small change region according to the embodiment, FIG. 7 is a diagram illustrating a small change region having two contour lines according to the present invention, and FIG. 8 is a diagram illustrating a small change region according to the fourth embodiment. 9 is a diagram showing a small change region according to the fifth embodiment, FIG. 10 is a diagram showing a small change region according to the sixth embodiment, and FIG. 11 is a change detection region according to the sixth embodiment. FIG. 12 is a diagram showing a minimum rectangle for a contour line, and FIG. 12 is a diagram showing a minimum rectangle for a change detection region contour line according to the seventh embodiment.

[First Embodiment]
In the intruder detection apparatus 1 according to the present embodiment, in order to detect an intruder into the restricted entry area by analyzing the video of the surveillance camera 2 installed in the restricted entry area such as a facility or a road, FIG. As shown in FIG. 3, the camera includes a monitoring camera 2 that captures an image and an image processing unit 3 that performs image processing on the image captured by the monitoring camera 2.

  The surveillance camera 2 is installed at a position overlooking the restricted entry area, and images the restricted entry area and its surroundings. An image captured by the monitoring camera 2 is transmitted to the image processing unit 3. The image processing unit 3 inputs the image transmitted from the monitoring camera 2 to the image processing unit 3 by the image input unit 10. The input image input by the image input means 10 is divided into images of appropriate small areas d by the small area dividing means 11 as shown in FIG.

  The small areas divided by the small area dividing means 11 are the background small area of the background image acquired by the correlation calculating means 12 in advance by the monitoring camera 2 and the image of the small area corresponding to this background image. The correlation calculation is performed to calculate the correlation with respect to the time change. Based on the calculation result calculated by the correlation calculation means 12, the change detection means 13 detects a change small area which is a small area having a temporal change and a background small area other than the change small area. The change small areas detected by the change detection means 13 are combined with adjacent change small areas by the change area combining means 14 to determine a change detection area.

  As shown in FIG. 2, the change detection area combined by the change area combination means 14 is extracted by the change area outline extraction means 15 as shown in FIG. The change detection area outline extracted by the change area outline extraction means 15 is applied to the intruder whose geometric feature quantity of the change detection area outline is preset by the geometric feature quantity evaluation means 16. It is evaluated whether or not it is within the range of the corresponding geometric feature amount (for example, area, circumference, minimum rectangle).

  The evaluation result evaluated by the geometric feature quantity evaluation means 16 is detected by the intruder detection means 17 by determining whether there is an intruder in the change detection area based on the evaluation result. The position in the input image input from the surveillance camera 2 is specified.

  Thus, according to the present embodiment, whether the background or the detection object is a single color or not, as shown in FIG. The contour E of the detection target object is extracted and the intruder is determined based on the geometric feature amount of the contour E (change detection region contour in this case). Even if a person wearing single color clothes stops or a single color object is left behind, the person or object can be detected.

[Second Embodiment]
In the intruder detection device 1 according to the present embodiment, in addition to the functions of the intruder detection device 1 according to the first embodiment, as shown in FIG. after extraction, the change subregion reconfigure Teite stage 18, it sets a small region within the range surrounded by the change detection region contour to reconstruct changes subregion.

  The reconstruction change small area set by the change small area reconstruction detection means 18 constitutes a reconstruction change detection area by the change detection area reconstruction means 19. The reconstruction change detection area constituted by the change detection area reconstruction means 19 is within the range of the geometric feature quantity corresponding to the intruder set in advance by the geometric feature quantity evaluation means 16 (for example, (Area, circumference, minimum rectangle) is evaluated whether there is a geometric feature amount of the change detection region outline.

  The evaluation result evaluated by the geometric feature quantity evaluation means 16 is detected by the intruder detection means 17 by determining whether there is an intruder in the change detection area based on the evaluation result. The position in the input image input from the surveillance camera is specified.

  As described above, according to the present embodiment, in addition to the effects of the intruder detection device 1 according to the first embodiment, the geometric feature amount for determining the intruder is changed in the same manner as the conventional intruder detection device. Since the geometric feature of the detection area is used, the same geometric feature amount range as that of the conventional intruder detection device can be set as it is.

[Third Embodiment]
In the intruder detection device 1 according to the present embodiment, in order to detect an intruder into the restricted entry area by analyzing the image of the surveillance camera 2 installed in the restricted entry area such as a facility or a road, FIG. As shown in FIG. 3, the camera includes a monitoring camera 2 that captures an image and an image processing unit 3 that performs image processing on the image captured by the monitoring camera 2.

  The surveillance camera 2 is installed at a position overlooking the restricted entry area, and images the restricted entry area and its surroundings. An image captured by the monitoring camera 2 is transmitted to the image processing unit 3. The image processing unit 3 inputs the image transmitted from the monitoring camera 2 to the image processing unit 3 by the image input unit 10. The input image input by the image input means 10 is divided by the small area dividing means 20 into images of appropriate small areas d as shown in FIG. At this time, as a small region for detecting a change in the input image, the contour of the detection target is detected to have a width, so that the contour has a width of, for example, 20 [pix] or more around the contour. Use small areas of large size.

  The small areas divided by the small area dividing means 20 are the background small area of the background image that has been acquired by the correlation calculating means 12 and captured in advance by the monitoring camera 2, and the image of the small area corresponding to this background image. The correlation calculation is performed to calculate the correlation with respect to the time change. Based on the calculation result calculated by the correlation calculation means 12, the change detection means 13 detects a change small area which is a small area having a temporal change and a background small area other than the change small area.

  The change detection area detected by the change detecting means 13 is further divided by the divided change small area setting means 21 to set a divided change small area. The divided change small areas set by the divided change small area setting unit 21 are combined with adjacent divided change small areas by the change area combining unit 14 to determine a divided change detection area.

  The divided change detection area determined by the change area combining unit 14 is detected by the geometric feature amount evaluating unit 16 in which the geometric feature amount of the divided change detection area corresponds to the geometrical shape corresponding to the intruder set in advance. It is evaluated whether it is within the range of the target feature amount (for example, area, circumference, minimum rectangle). The evaluation result evaluated by the geometric feature amount evaluation means 16 is detected by the intruder detection means 17 by determining whether or not an intruder exists in the divided change detection area based on the evaluation result. Thus, the position in the input image input from the monitoring camera 2 is specified.

  Thus, according to the present embodiment, as shown in FIG. 6, by using the large small area d (see FIG. 2) for change detection and further combining the small change areas to obtain the change detection area D, Although it is impossible to detect a change in the central portion where the entire small area becomes a single color with respect to a single color object, the change in the vicinity of the outline of the detection target m is detected in a donut shape with a certain width. be able to. For this reason, those persons and objects can be stably detected even in places where there are few patterns indoors and outdoors.

  Further, in the first embodiment, the contour line E (see FIG. 2) is extracted. However, in the small region d (see FIG. 2) divided into small sizes as in the first embodiment, A in FIG. There is a possibility that a portion such as the portion shown in FIG. In this embodiment, since the size of the small region d is increased, the extraction omission of the contour line E can be reduced.

[Fourth Embodiment]
In the intruder detection device 1 according to the present embodiment, in addition to the function of the intruder detection device 1 according to the third embodiment, as shown in FIG. After the division change detection area is determined, when the division change detection area has a donut shape (see FIG. 6), the change detection area filling processing unit 22 uses the change detection area filling processing unit 22 to subtract the background small area portion inside the division change detection area. By changing to the division change small region, it is possible to obtain a hole filling division change detection region that is packed to the inside.

  The burial division change detection area filled by the change detection area burial processing means 22 is within the range of the geometric feature quantity corresponding to the intruder set in advance by the geometric feature quantity evaluation means 16. It is evaluated whether (for example, area, circumference, minimum rectangle) has a geometric feature amount of the hole filling division change detection region. The evaluation result evaluated by the geometric feature amount evaluation means 16 is detected by the intruder detection means 17 by determining whether or not an intruder exists in the divided change detection area based on the evaluation result. The position in the input image input from the surveillance camera is specified.

  Thus, according to the present embodiment, in addition to the effects of the intruder detection device 1 according to the third embodiment, even when the background is a single color and the detection target is also a single color, the detection target is not affected. Since it is possible to obtain a change detection area that is clogged up to the inside, it is possible to obtain the geometric feature amount of the change detection area that is the same as when the background and the detection target are not monochromatic, so the feature amount range that is an intruder Can be easily set.

[Fifth Embodiment]
In the intruder detection apparatus 1 according to the present embodiment, in order to detect an intruder into the restricted entry area by analyzing the image of the surveillance camera installed in the restricted entry area such as a facility or a road, FIG. As shown, a monitoring camera 2 that captures an image and an image processing unit 3 that performs image processing on the image captured by the monitoring camera 2 are provided.

  The surveillance camera 2 is installed at a position overlooking the restricted entry area, and images the restricted entry area and its surroundings. An image captured by the monitoring camera 2 is transmitted to the image processing unit 3. The image processing unit 3 inputs the image transmitted from the monitoring camera 2 to the image processing unit 3 by the image input unit 10. The input image input by the image input means 10 is divided by the small area dividing means 20 into images of appropriate small areas d as shown in FIG. At this time, as a small region for detecting a change in the input image, the contour of the detection target is detected to have a width, so that the contour has a width of, for example, 20 [pix] or more around the contour. Use small areas of large size.

  The small areas divided by the small area dividing means 20 are the background small area of the background image that has been acquired by the correlation calculating means 12 and captured in advance by the monitoring camera 2, and the image of the small area corresponding to this background image. The correlation calculation is performed to calculate the correlation with respect to the time change. Based on the calculation result calculated by the correlation calculation means 12, the change detection means 13 detects a change small area which is a small area having a temporal change and a background small area other than the change small area.

  The change small area detected by the change detecting means 13 is further divided by the divided change small area setting means 21 to set a divided change small area. The divided change small areas set by the divided change small area setting means 21 are combined with adjacent change small areas by the change area combining means 14 to determine a divided change detection area.

  As shown in FIG. 2, the divided change detection region combined by the change region combining unit 14 is extracted by the change region outline extracting unit 15, as shown in FIG. 2. The division change detection region outline extracted by the change region outline extraction unit 15 is input to the geometric feature amount evaluation unit 16 in which the geometric feature amount of the division change detection region outline is set in advance. It is evaluated whether it is within the range of the geometric feature amount corresponding to the person (for example, area, circumference, minimum rectangle).

  The evaluation result evaluated by the geometric feature quantity evaluation means 16 is detected by the intruder detection means 17 by determining whether there is an intruder in the change detection area based on the evaluation result. The position in the input image input from the surveillance camera 2 is specified.

  In this way, according to the present embodiment, as shown in FIG. 6, the change detection region D (here, using a large small region d (see FIG. 2) for change detection and further combining the divided change small regions) By determining the division change detection area), it is not possible to detect a change in the central portion where the entire small area becomes a single color for a single color object, but with a certain width in the vicinity of the outline of the detection target m The change can be detected in a donut shape. Further, by extracting an outline E (see FIG. 2) from the divided change detection area D having a width and detecting an intruder based on the outline E, a single-color clothing can be used in a place where there are few patterns indoors and outdoors. Even when a person who has stopped or a single color object is left behind, the person or object can be detected stably.

[Sixth Embodiment]
In the intruder detection device 1 according to the present embodiment, in addition to the functions of the intruder detection device 1 according to the fifth embodiment, as shown in FIG. When the divided change detection region contour line becomes a donut shape (see FIG. 6), the inner contour line deletion means 23 performs double extraction from one division change detection region as shown in FIG. The inner contour line Ei is deleted from the contour lines to be processed. As a method for determining whether or not the inner contour line is Ei, a minimum rectangle parallel to the surrounding axis is obtained for each contour line, and the minimum rectangle corresponds to another contour line. Judgment is made based on whether or not it is included in As shown in FIG. 11, for example, the outline of the change detection area Ds included inside the minimum rectangle r is the change detection area outline Es.

  The outer division change detection area from which the inner contour line Ei has been deleted by the inner contour line deleting unit 23 is the range of the geometric feature amount corresponding to the intruder set in advance by the geometric feature amount evaluating unit 16. It is evaluated whether there is a geometric feature amount of the outer division change detection region outline in the inside (for example, area, circumference, minimum rectangle). The intruder detection means 17 detects the intruder by judging whether or not the intruder exists in the change detection area based on the evaluation result by the intruder detection means 17. The position in the input image input from the monitoring camera 2 is specified.

  Thus, according to the present embodiment, in addition to the effects of the intruder detection device according to the fifth embodiment, even if the background is a single color and the detection target is also a single color, a single detection target is doubled. By deleting the inner contour line from the contour lines extracted and obtaining only the outer contour line, the amount of calculation and the memory usage can be reduced.

[Seventh Embodiment]
In the intruder detection device 1 according to the present embodiment, in addition to the functions of the intruder detection device 1 according to the sixth embodiment, as shown in FIG. After deleting the inner outline from the double extracted outlines, the change small area resetting means 18 changes the small area within the range surrounded by the outer change detection area outline to the reset change small area. Set. For the reset change small area set by the change small area resetting means 18, the change detection area reconstruction means 19 sets the small area within the range surrounded by the outer change detection area outline as the reconstructed outer change small area. To do.

  The reconstructed outside change detection area constituted by the change detection area reconstruction means is within the range of the geometric feature quantity corresponding to the intruder set in advance by the geometric feature quantity evaluation means 16 (for example, (Area, circumference, minimum rectangle) is evaluated whether there is a geometric feature amount of the change detection region outline. The intruder detection means 17 detects the intruder by judging whether or not the intruder exists in the change detection area based on the evaluation result by the intruder detection means 17. The position in the input image input from the surveillance camera is specified.

  As described above, according to the present embodiment, in addition to the effects of the intruder detection device 1 according to the sixth embodiment, the geometric feature amount for determining the intruder is the same change as that of the conventional intruder detection device. Since the geometric feature of the detection area is used, the same geometric feature amount range as that of the conventional intruder detection device can be set as it is.

It is an apparatus block diagram of the intruder detection apparatus which concerns on 1st Embodiment. It is the figure which showed the small area | region and outline which concern on this invention. It is the figure which showed the change small area | region which concerns on 1st Embodiment. It is an apparatus block diagram of the intruder detection apparatus which concerns on 2nd Embodiment. It is an apparatus block diagram of the intruder detection apparatus which concerns on 3rd Embodiment. It is the figure which showed the change small area | region which concerns on 3rd, 5th embodiment. It is the figure which showed the change small area | region which has two outlines based on this invention. It is the figure which showed the change small area | region which concerns on 4th Embodiment. It is the figure which showed the change small area | region which concerns on 5th Embodiment. It is the figure which showed the change small area | region which concerns on 6th Embodiment. It is the figure which showed the minimum rectangle with respect to the change detection area outline which concerns on 6th Embodiment. It is the figure which showed the minimum rectangle with respect to the change detection area outline which concerns on 7th Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Intruder detection apparatus 2 Surveillance camera 3 Image processing part 10 Image input means 11 Small area division means 12 Correlation calculation means 13 Change detection means 14 Change area combination means 15 Change area outline extraction means 16 Geometric feature amount evaluation means 17 Intruder detection means 18 Change small area resetting detection means 19 Change detection area reconstruction means 20 Small area dividing means (large size)
21 division change small area setting means 22 change detection area filling process means 23 inner contour line deletion means

Claims (4)

In an intruder detection apparatus comprising a camera that captures an image and an image processing unit that processes an image captured by the camera,
Wherein the image processing unit,
Input means for inputting the image to the image processing unit;
A small area dividing means for dividing the image into small areas;
Correlation calculating means for calculating a temporal correlation between the small area and a background image previously captured by the camera corresponding to the small area;
Change detecting means for detecting the small area having a temporal change as a small change area;
A change region combining means that combines the change small regions into a change detection region;
A change area outline extracting means for extracting the outline of the change detection area as a change detection area outline;
A geometric feature amount evaluating means for evaluating whether or not the geometric feature amount of the change detection region outline is within a range of a geometric feature amount corresponding to a preset intruder;
Intruder detection means for determining whether there is an intruder in the change detection area based on the result of the evaluation ,
An intruder detection device further comprising an inner contour line deleting unit that deletes an inner contour line from the contour lines extracted from one of the change detection areas . In an intruder detection apparatus comprising a camera that captures an image and an image processing unit that processes an image captured by the camera,
Wherein the image processing unit,
Input means for inputting the image to the image processing unit;
A small area dividing means for dividing the image into a small area having a relatively large size around the outline in order to detect the outline of the detection object with a width;
Correlation calculating means for calculating a temporal correlation between the small area and a background image previously captured by the camera corresponding to the small area;
Change detecting means for detecting the small area having a temporal change as a small change area ;
A division changing small area setting means for setting a division changing small regions and further dividing the pre-Symbol changes subregion,
Change area combining means for determining a divided change detection area by combining adjacent divided change small areas;
A geometric feature amount evaluating means for evaluating whether or not the geometric feature amount of the divided change detection region is within a range of a geometric feature amount corresponding to a preset intruder;
Intruder detection means for determining whether there is an intruder in the change detection area based on the result of the evaluation ,
A change detection area filling processing unit is further provided which changes a portion of the background small area inside the division change detection area into a division change small area, thereby making a hole filling division change detection area clogged up to the inside. An intruder detection device characterized by that. In an intruder detection apparatus comprising a camera that captures an image and an image processing unit that processes an image captured by the camera,
Wherein the image processing unit,
Input means for inputting the image to the image processing unit;
A small area dividing means for dividing the image into a small area having a relatively large size around the outline in order to detect the outline of the detection object with a width;
Correlation calculating means for calculating a temporal correlation between the small area and a background image previously captured by the camera corresponding to the small area;
Change detecting means for detecting the small area having a temporal change as a small change area ;
A division changing small area setting means for setting a division changing small regions and further dividing the pre-Symbol changes subregion,
Change area combining means for determining a divided change detection area by combining adjacent divided change small areas;
A change area outline extracting means for extracting a division change detection area outline which is an outline of the division change detection area;
A geometric feature amount evaluating means for evaluating whether or not a geometric feature amount of the division change detection region contour line is within a range of a geometric feature amount corresponding to a preset intruder;
Intruder detection means for determining whether there is an intruder in the change detection area based on the result of the evaluation ,
An intruder detection apparatus further comprising an inner contour line deleting unit that deletes an inner contour line from contour lines extracted from one of the divided change detection areas . In the intruder detection device according to claim 1 or claim 3,
Intruder, further comprising a change subregion reconfigure Teite stage for setting a small region within the range surrounded by said change detection area contour line or the divided change detection region contour to reconfigure change subregion Detection device.

Images