JP2002140711A - Method for setting size of invasion object detecting device, invasion object detecting method and invasion object detector - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily set the size of an invasion object to be monitored without the need of the basic knowledge of a camera image pickup system even if sizes of monitor view field and an image pickup device, the focal distance and zoom ratio of a using lens and a distance to the object are changed. SOLUTION: An image signal with which a reference object is photographed is inputted. The reference object is detected from the inputted image signal. The size of the detected reference object is measured, and the detection position of the detected object and the measured size are recorded. The size of the object considered to be the invasion object is decided based on the recorded detection position and the size of the recorded reference object.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a monitoring apparatus using an imaging device, and more particularly to a method for automatically measuring conditions for determining whether a detected object is an intruding object to be monitored from a detection result of a reference object. The present invention relates to a method and a device for detecting an intruding object.

[0002]

2. Description of the Related Art In recent years, an intruding object detecting device using an image pickup device such as a camera as an image input means has not been conventionally used for manned monitoring by an observer, but automatically detects an intruding object in a monitoring field of view. In many cases, the type of the object is automatically confirmed, and a predetermined notification or alarm is automatically obtained. In order to realize such a system, first, processing for detecting a predetermined intruding object from an input image obtained by an image input unit such as a camera is required. As a method of realizing such processing, for example, an input image is compared with a reference background image (that is, an image in which an object to be detected is not captured), a difference is obtained for each pixel, and an area having a large difference is determined. There is a method of extracting as an object. This method is called a difference method and has been widely used conventionally.

The processing of the difference method will be described with reference to FIG. FIG. 5 is a diagram for explaining the processing principle of object detection in the difference method. 501 is the input image, 502 is the reference background image, 503
Is a difference image, 504 is a binarized image of the difference image 503, 505 is a subtractor, 506 is a binarizer, 507 is a region of a humanoid object in the input image 501, and 508 is a difference image The resulting difference area 509 is a binarized area 510 extracted in the binarized image 504.
Is a circumscribed rectangle of the binarization area 509.

In FIG. 5, a subtractor 505 calculates a difference in luminance value between an input image 501 and a reference background image 502 for each pixel, and outputs a difference image 503. The binarizer 506 sets the luminance value (difference value) of each pixel of the difference image 503 to a predetermined threshold value Th.
A binary value where the luminance value of less than (difference value <Th) is “0” and the luminance value of a pixel that is equal to or greater than the threshold value Th (difference value ≧ Th) is “255” (the luminance value of one pixel is calculated with 8 bits) A modified image 504 is obtained.
Here, for example, the threshold value Th is set to “20”. Thus, the human-like object 5 in the input image 501
In step 07, the region 508 where the difference has occurred is calculated by the subtractor 505, and the binarizer 506 detects an image 509 of a cluster of the luminance value “255” as an intruding object. As an object detection method applying the difference method, for example,
No. 9429.

Further, in the conventional technique, the size of a detected object is evaluated to determine whether or not the detected object is an intruding object to be monitored.
For the object (image 509) detected by the difference method, a circumscribed rectangle 510 (width w and height h) is measured based on the cluster area. When the width w and the height h are equal to or larger than a predetermined size that is regarded as an intruding object to be monitored, the detected object is determined as an intruding object to be monitored.

Here, the size regarded as an intruding object to be monitored is, for example, a 1/2 type CCD (element size:
Monitoring using a lens of width dx = 6.5mm, height dy = 4.6mm), f = 25mm, distance from camera to monitoring area L = 50m, input image size: width X0 = 320pix, height Y0 = 240pix , The intruder to be monitored should have a width of w0 = 1m and a height of h0 = 1.5m
Then, the following equation (1) is obtained.

(Equation 1) The width w of the circumscribed rectangle 510 is 25 or more and the height h of the circumscribed rectangle 510 is 39
In this case, the object 509 is determined to be an intruding object to be monitored. In this way, it is possible to distinguish between an intruding object to be monitored in the monitoring field of view and an intrusion such as erroneous detection due to shaking trees or invasion of small animals or fallen leaves. Furthermore, by changing the size of an object regarded as an intruding object to be monitored, an object such as a person or a vehicle can be distinguished.

In this method, it is necessary to set in advance the size of an object regarded as an intruding object to be monitored. Therefore, the size of an object considered as an intruding object to be monitored must be determined each time the monitoring field of view, the size of the image sensor, the focal length and zoom ratio of the lens used, and the distance to the target are changed. Was complicated. Further, since the calculation of the size of an object regarded as an intruding object to be monitored requires basic knowledge of a camera imaging system, the setting operation requires skill.

[0008]

The above-mentioned prior art includes the following:
In order to determine whether the detected object is an intruding object to be monitored, it is necessary to set in advance the size of an object to be regarded as an intruding object to be monitored. This has to be performed every time the focal length and zoom ratio of the lens to be used and the distance to the object are changed. Furthermore, there is a drawback that skill is required because the setting operation is complicated. An object of the present invention is to eliminate the above-described drawbacks and automatically calculate a size regarded as an intruding object to be monitored by detecting a reference object captured in an image captured in an imaging field of view. It does not require basic knowledge of camera imaging system, and it is regarded as an intruding object that should be easily monitored even if the monitoring field of view, the size of the image sensor, the focal length and zoom ratio of the lens used, and the distance to the target change. An object of the present invention is to provide a highly reliable intruding object detection method and intrusion object detecting device that can accurately detect an intruding object by enabling the size of the object to be set.

[0009]

In order to achieve the above-mentioned object, a method for setting the size of an object regarded as an intruding object according to the present invention comprises placing an image of a predetermined reference object at a plurality of predetermined positions in a monitoring field of view. The size of an object regarded as an intruding object in the monitoring visual field is calculated based on the calculated detection position and size of the reference object, and calculates the size of the reference object detected at each of a plurality of predetermined positions. To set. That is, the method for setting the size of an object regarded as an intruding object according to the present invention is a method for detecting an intruding object in an image signal sequentially input from an imaging device for imaging the inside of a monitoring visual field. Is imaged at a plurality of predetermined positions in the monitoring field of view by the imaging device, and from the imaged image signal,
The reference object is detected, the size of the reference object detected at each of the plurality of predetermined positions is calculated, and intrusion in the monitoring visual field is performed based on the calculated plurality of detection positions of the reference object and the size corresponding to each. The size of an object to be regarded as an object is set.

In the method of setting the size of an object considered as an intruding object according to the present invention, the size of the object considered as an intruding object is determined by the width, height, and area of a detection area of the circumscribed rectangle of the detected mass of the reference object. At least one of Further, in the method for setting the size of an object considered as an intruding object according to the present invention, the size of the object considered as an intruding object is obtained by an approximate expression using the position of the detected reference object as a variable.

In the method for detecting an intruding object according to the present invention, an image of a predetermined reference object is picked up at a plurality of predetermined positions in a monitoring field of view, and the size of the reference object detected at each of the plurality of predetermined positions is calculated. Then, based on the calculated detection position and size of the reference object, the size of an object regarded as an intruding object in the monitoring visual field is set, and the size is detected from an image signal sequentially input from an imaging device that captures an image of the monitoring visual field. By comparing the size of the detected object with the size of the object regarded as the intruding object, it is determined whether or not the detected object is an intruding object, and the intruding object in the monitoring visual field is detected. The determination as to whether the detected object is an intruding object is performed when the size of the detected object is within the range of the first predetermined ratio and the second predetermined ratio of the size of the object regarded as the intruding object. It depends on whether or not. In addition, a plurality of sizes of an object considered as an intruding object are set, and a different type of intruding object is determined according to each of the plurality of settings. Further, in the method for detecting an intruding object according to the present invention, the size of an object regarded as an intruding object is updated using the detected intruding object as a reference object.

Further, the intruding object detection device of the present invention is a method for detecting an intruding object by capturing an image in a monitoring visual field and analyzing the captured image to detect an intruding object. An image input interface for sequentially converting a video signal obtained by the imaging device into an input image; an image storage unit for storing the input image; and an image processing unit for processing the input image, wherein the image processing unit is configured to convert the input image into a reference image. Detects the object, records the detected position and size of the detected reference object, determines the size of the object to be regarded as an intruding object from the recorded detected position and size of the detected object, and sequentially inputs the image signal By comparing the size of the object detected from the object and the size of the object regarded as the intruding object, it is determined whether or not the detected object is the intruding object. Further, an intruding object detection device of the present invention is an intruding object detection method for detecting an intruding object by capturing an image in a monitoring visual field and analyzing the captured image, wherein an imaging device for capturing an image in the monitoring visual field and an imaging device An image input interface for sequentially converting the acquired video signal into an input image, an image storage means for storing the input image, and an image processing means for processing the input image, wherein the image processing means converts a reference object from the input image. Detects and records the detected position and size of the detected reference object, determines the size of the object regarded as an intruding object from the recorded detected position and size of the detected object, and detects from the sequentially input image signal The size of the detected object is compared with the size of the object considered as the intruding object to determine whether the detected object is the intruding object.
Furthermore, the intruding object detection device of the present invention has input means for setting whether or not to update the size of an object considered as an intruding object, and updates the size of the object considered as an intruding object by the input means. Is set, the detection position and the size of the detected intruding object are calculated, and the size of the object regarded as the intruding object is updated.

[0013]

FIG. 4 shows an embodiment of the intruding object monitoring apparatus according to the present invention. FIG. 4 is a block diagram illustrating a hardware configuration of the intruding object monitoring device. In FIG. 4, reference numeral 401 denotes a television camera (hereinafter, referred to as a TV camera); 402, an input device; 403, an image input interface (hereinafter, the interface is referred to as an I / F);
I / F, 405 is image memory, 406 is work memory, 407 is CPU
(Central Processing Unit), 408 is a program memory, 409 is an output I / F, 410 is an image output I / F, 411 is a warning light, 4
12 is a monitoring monitor, and 413 is a data bus.

In FIG. 4, a TV camera 401 has an image input I / F
403, the input device 402 is coupled to the input I / F 404,
The warning light 411 is coupled to the output I / F 409, and the monitor 412 is coupled to the image output I / F 410. In addition, image input I / F40
3, input I / F404, image memory 405, work memory 406, CPU
407, program memory 408, output I / F 409 and image output I / F
F410 is coupled to data bus 413.

Referring to FIG. 4, a TV camera 401 captures an image of the field of view including the monitoring target area. TV camera 401
The captured video is converted into an electrical video signal (for example, an NTSC video signal), and the converted video signal is provided to the image input I / F 403. The image input I / F 403 is a format (for example, width 320 pix, height
240pix, 8bit / pix) image data and data bus
This is given to the image memory 405 via 413. Image memory 405
Accumulates the transmitted image data. Input device 402
Detects at least one input operation of a specific operation such as a voice or gesture of a human or an auxiliary animal, an input by an external input device such as a button, a keyboard, a mouse, and the like, and detects an electrical ON or OFF binary signal (contact point). Signal), and give the converted binary signal to the input I / F 404.

The input I / F 404 converts the input binary signal into binary data in a format (for example, “255” when turned on and “0” when turned off) to be handled by the intruding object monitoring apparatus,
The data is supplied to the work memory 406 via the data bus 413. The work memory 406 stores the sent binary data. C
The PU 407 analyzes the image stored in the image memory 405 in the work memory 406 according to the program stored in the program memory 408. As a result of the above analysis, information such as whether or not an intruding object has entered the imaging field of view of the TV camera 401 is obtained.

The CPU 407 supplies a video signal to the monitor 412 via the image output I / F 410, and displays, for example, a processing result image. At the same time, the CPU 407 gives a control signal for turning on the warning light 411 to the warning light 411 via the output I / F 409. The warning lamp 411 is turned on by the input control signal. Image output I
The / F 410 converts the video signal from the CPU 407 into a format (for example, an NTSC video signal) that can be used by the monitor 412 and sends the format to the monitor 412. The monitoring monitor 412 is, for example,
An intruding object detection result image is displayed.

FIG. 2 is an example of a flowchart showing the processing operation of the first embodiment of the present invention. This flowchart is executed using FIG. 4 which is an example of a hardware configuration of the intruding object monitoring device. In FIG. 4, a predetermined reference object is arranged in the monitoring visual field, and the TV camera 401
And input to the intruding object monitoring device via the image input I / F 403. The given image signal detects the reference object by, for example, the difference method described with reference to FIG. 5, calculates the detection position of the detected reference object and the size of the circumscribed rectangle of the detection area, and records the size in the work memory 406. From the recorded detection position of the reference object and the size of the circumscribed rectangle of the detection area, the size of an object regarded as an intruding object to be monitored is determined. Then, it is determined whether or not the reference object detected in the field of view of the TV camera 401 is an intruding object to be monitored, thereby monitoring the intruding object.

Before starting the process of FIG. 2, a reference object (for example, a human, a doll, etc.) of the same size as the intruding object to be monitored
Are prepared in advance and arranged at a predetermined position in the field of view of the TV camera 401. In FIG. 2, in a video selection step 101, a video in which a reference object prepared in advance is within the field of view is captured by the TV camera 401. Next, in the image input step 102, the input video signal captured by the TV camera 401 is
For example, it is acquired as an input image 501 of 320 × 240 pixels.
Next, in the reference object detection step 103, a reference object is detected by, for example, the difference method described with reference to FIG. In the size measurement step 104, a cluster of pixels having a value of “255” in the binarized image 504 obtained in the reference object detection step 103 is detected using, for example, a method such as labeling,
The detection position (the displacement y between the circumscribed rectangle 510 of the detected object and the lower part of the image in FIG. 5) and the size (the width w, the height h, or the detected area (number of pixels) of the circumscribed rectangle 510 of the detected object in FIG. 5) (At least one). In the size recording step 105, the detection position and size of the reference object obtained in the size measurement step 104 are recorded in the work memory 406. Next, in a predetermined frame end determination step 106, a predetermined number of frames N (when the size of an object regarded as an intruding object is expressed by a quadratic expression using a detection position as a variable, N is preferably 3 or more for different y. When the processing described in detail later is completed, the process branches to the size determination step 201, and when the number of frames does not reach the predetermined number N, the process branches to the image input step.

In the size determining step 201, the size of an object regarded as an intruding object to be monitored is determined from the detected position and size of the reference object recorded in the work memory. This processing will be described with reference to FIGS. FIG. 6 is a diagram illustrating a state of a binarized image when the reference object moves to various positions. The reference object is observed so large that it is detected on the near side of the binarized image. In contrast, the smaller the image is detected on the far side of the binarized image 504, the smaller the observation is. FIG.
, The detected position of the reference object measured in each binarized image is yi, the width of the circumscribed rectangle is wi, and the height of the circumscribed rectangle is h
i, and the detection area is si. Here, the detection position yi of the reference object is a distance from the lower side of the binarized image to the lowermost side of the detected object. The size of the reference object indicates at least one or more of the width wi of the circumscribed rectangle, the height hi of the circumscribed rectangle, and the area si. For example, the detection position y ₁ , height h ₁ , and width w of the circumscribed rectangle 601a of the reference object detected in the binary image 601 where the distance between the TV camera 401 and the reference object is small
₁ , a binary image with a large distance between the TV camera 401 and the reference object
Detection position of the circumscribed rectangle 603a of the reference object detected by 603
y ₃ , height h ₃ , width w _3, and the detection position of the circumscribed rectangle 602a of the reference object where the distance between the TV camera 401 and the reference object is detected in the binary image 602 between them, y ₂ , height h ₂ and the width w ₂ , that is, when y ₁ <y ₂ <y ₃ , h ₁ > h ₂ >
h ₃ , w ₁ > w ₂ > w ₃ , s ₁ > s ₂ > s ₃ (s ₁ = h ₁ × w ₁ , s ₂ = h ₂ × w ₂ ,
s ₃ = h ₃ × w ₃ ).

FIG. 7 is a graph showing the detection position yi of the reference object and the width wi of the circumscribed rectangle with the horizontal axis representing the detection position and the vertical axis representing the width of the circumscribed rectangle. And the width of the circumscribed rectangle (yi, wi (i = 1, ‥‥‥, N), respectively) measured in an image of one frame (an image of the reference object is taken at an appropriate position). FIG.
As described above, there is a relationship between the detection position and the width of the circumscribed rectangle such that the width w of the circumscribed rectangle decreases as the detection position y increases. In FIG. 7, a curve shown by a broken line represents a curve that approximates the detection position detected in each frame and the width of the circumscribed rectangle.

This approximation curve is called a regression curve, and can be calculated using a method called a least squares method.
For example, when calculating a regression curve by a quadratic equation,
In the case of the approximation expressed by (2), the approximation error E
The cumulative value of the square of the value of W (y) and the measured value wi) can be obtained by Expression (3).

(Equation 2)

(Equation 3)

Aw, bw, cw that minimize this approximation error E
Can be obtained by the following equation (4).

(Equation 4)

Here, the matrix [A] ^-1 represents the inverse matrix of the matrix [A]. When calculating a regression curve by a quadratic formula, if there are three or more yi with different values (regression curve order + 1 or more),
There is an inverse matrix [A] ^{-1 of the} matrix [A]. That is, by using W (y), any y not measured using the reference object
It is possible to calculate the size of an object regarded as an intruding object to be monitored and to use it as a condition for determining whether or not the object is an intruding object to be monitored.

An approximation formula for calculating the size of an object considered as an intruding object is only one position with respect to the detected position of the detected object.
Although only one calculation is performed, there is a case where it is necessary to provide a certain range for the calculated values because the size of the intruding object may vary. Therefore, in the above embodiment, the size calculated for the detection position is:
It is multiplied by a coefficient for providing predetermined upper and lower ranges (for example, the size of an object regarded as an intruding object × 0.8 and the size of an object regarded as an intruding object × 1.3).

Similarly, FIG. 8 shows the detected position yi of the reference object.
FIG. 9 is a graph in which the horizontal axis represents the detection position and the vertical axis represents the height of the circumscribed rectangle with respect to the height hi of the circumscribed rectangle.
Is a graph showing the detection position of the reference object yi and the detection area si of the reference object, with the horizontal axis representing the detection position and the vertical axis representing the detection area. The approximate curve 801 represents the relationship between the detected position of the reference object and the height of the circumscribed rectangle, and the approximate curve 901 represents the relationship between the detected position of the reference object and the detected area. Approximate curves 801 and 901
Can be calculated as the following Expressions (5) and (6), similarly to the approximation curve 701.

(Equation 5)

(Equation 6)

Subsequently, in the image input step 109 of FIG.
As in the image input step 102, an input video signal captured by the TV camera 401 is obtained as an input image 501 of, for example, 302 × 240 pixels (pix). Next, in an object detection step 110, an object is detected from the input image 501 that is sequentially input.
In the size measuring step 111, a cluster of pixels having a value of “255” is detected in the binarized image 504 obtained in the object detecting step 110 using a method such as labeling, and the detected object is detected. Position y and object size (Fig. 5
At least one of the width w, the height h, or the detection area (the number of pixels) of the circumscribed rectangle 510 of the detection object in step (a) is measured.

In the intruding object determination step 202, a regression curve is calculated by, for example, equation (2) using the detection position y obtained in the size measurement step 111, and the detected object obtained in the size measurement step 111 is calculated. If the size w of w is w ≧ W (y), the detected object is regarded as an intruding object to be monitored. This process is performed for all detected objects present in the binarized image 504, and if there is an intruding object to be monitored, the process branches to the alarm / monitor display step 203.If there is no intruding object to be monitored, The flow branches to an image input step 109.

In this description, an example will be described in which the width of the circumscribed rectangle of the detected object is used to determine whether or not the object is an intruding object to be monitored. Alternatively, the determination can be made using the area of the circumscribed rectangle of the detected object using Expression (6).

Also, the determination can be made by combining two or more of the width, height, area, peripheral length, diagonal length (straight line connecting opposite vertices) of the circumscribed rectangle of the detected object, and the like. When the width and height of the circumscribed rectangle of the object are used as indicators for determination, when w ≧ W (y) and h ≧ H (y),
The detected object can be set as an intruder to be monitored, or when w ≧ W (y) or h ≧ H (y), the detected object can be set as an intruder to be monitored.

In the alarm / monitor display step 203, for example, the processing result is displayed on the monitor 412 via the image output I / F 410, and the warning lamp 411 is turned on via the output I / F 409. Then, the process returns to the image input step 109 to continue monitoring. Therefore, according to the present embodiment, it is possible to determine the size of an object regarded as an intruding object to be monitored from the detection position and the size measured using the reference object, and to set the size of the intruding object to be monitored. And it is possible to detect an intruding object in the field of view of the imaging device.

FIG. 3 is an example of a flowchart showing the processing operation of the second embodiment of the present invention. In the second embodiment, a plurality of types of intruding objects to be monitored can be determined. FIG. 3 uses a second size determination step 107 instead of the size determination step 201 in the flowchart shown in FIG.
And a second intruding object determination step 112 is used in place of the intruding object determination step 202, and a second alarm / monitor display step 113 is used in place of the alarm / monitor display step 203. Other steps are the same as in the first embodiment described with reference to FIG.

In FIG. 3, the description starts from the point where the processing operation has proceeded to the predetermined frame end determination step 106. If the processing in the predetermined frame end determination step 106 has been completed, the flow branches to a second size determination step 107, and if the predetermined number of frames N is not reached, the image input step 102
Branch to In the second size determination step 107, the approximate curves Wj (y) and Hj are set for each type of the reference object captured in the input image captured in the monitoring visual field in the video selection step 101.
(y) and Sj (y) are obtained. Here, the index j represents the number of the type of the reference object. For example, when two kinds of intruding objects to be monitored are determined (M = 2), j = 1 for a human, and j = 1 for a vehicle. As j = 2, the video selection step 101 is numbered in advance according to the reference object to be captured in the input image. This numbering may be arbitrarily classified into adults and children in the case of humans and large and small vehicles in the case of vehicles. Next, in the predetermined type end determination step 108, when the reference object number becomes equal to the predetermined number M of types of intruding objects to be monitored, the image input step
The process branches to 109, and if the number of types is less than the number M, the process branches to the image selection step 101 so that another type of reference object of the intruding object to be monitored is imaged in the imaging field of view.

The process proceeds from step 109 to step 111. In the second intruding object determination step 112, Wj (y) (j
= 1, ‥‥‥, M) (Here, an example will be described in which the width of the circumscribed rectangle of the detected object is used to determine whether or not the object is an intruding object to be monitored). When the detected size w of the detected object is w ≧ Wj (y) (j = 1, ‥‥‥, M), the detected object is determined as a j-th kind of intruding object to be monitored. Here, an object detected using the minimum j that satisfies w ≧ Wj (y) may be determined as a jth type of intruding object to be monitored, or w ≧ Wj (y).
If there is a plurality of js, the type of the detected object may be determined as having a plurality of possibilities. That is, for example, when j satisfying w ≧ Wj (y) is j = 1,3,
It is assumed that the detected object may be a first type of intruding object to be monitored or a third type of intruding object to be monitored.

Next, a second alarm / monitor display step 11
In 3, the image output I / O is performed in accordance with the type of the intruding object to be monitored obtained in the second intruding object determining step 112.
The processing result and the type of the intruding object are displayed on the monitoring monitor 412 via the F410, and the warning light 411 is turned on via the output I / F 409 according to the type of the intruding object (for example, while changing the blinking speed and the color). Or let it. Therefore, according to the present embodiment, the size of an object considered as an intruding object to be monitored can be determined for each type from the detection position and the size measured using the reference object, and the size of the intruding object to be monitored can be determined. Setting can be made easier, and an intruding object in the field of view of the imaging device can be detected.

FIG. 1 is an example of a flowchart showing the processing operation of the third embodiment of the present invention. In the third embodiment, an approximate curve representing the size of an object regarded as an intruding object is updated at an arbitrary timing such as an input by a supervisor or at a certain time interval. FIG. 1 is obtained by adding a size condition update determination step 114, a second size recording step 115, and a third size determination step 116 to the flowchart shown in FIG. The other steps are the same as those described with reference to FIG. 2 or FIG. In FIG. 1, a predetermined frame end determination step 106
The description will be started from the point where the processing operation has been performed up to this point. In the alarm / monitor display step 203, for example, an image output I / F
The processing result is displayed on the monitoring monitor 412 via the 410, and the warning lamp 411 is turned on via the output I / F 409. Thereafter, the process proceeds to step 114.

In the size condition update determination step 114, the signal of the input device 402 obtained via the input I / F 404 is turned on / off.
Branches off. That is, when the switch is ON, the size condition is updated.
The process branches to 115, and if OFF, the size condition is not updated, so the process returns to the image input step 109 and monitoring is continued under the same condition. If the input device 402 can be turned on by an arbitrary input of a monitor, for example,
An arbitrary device such as a button, a keyboard, and an input device such as a mouse can be selected.

In the second size recording step 115, the detection position and the size obtained in the size measurement step 111 are recorded in the work memory 406. At this time, the number N of detected position and size data recorded so far is increased by 1, or the oldest detected position and size data recorded so far is deleted (in this case, N does not change). Subsequently, a third size determination step 115
Then, similar to the second size determination step 107, an approximate curve Wj representing the size of an object regarded as an intruding object to be monitored
(y), Hj (y), and Sj (y) (j = 1, ‥‥‥, M) are updated. Therefore, it is possible to update the approximate curve representing the size of an object considered as an intruding object at an arbitrary timing such as an input of a monitor or at a certain time interval, and to easily set the size of the intruding object to be monitored, It is possible to detect an intruding object in the field of view of the imaging device. Third size determination step 115
Is completed, the process returns to the image input step 109, and monitoring is further continued.

In the above-described embodiment, the approximate expression for calculating the size of an object regarded as an intruding object is a quadratic equation. However, depending on the range to be monitored, the object to be monitored, and the monitoring conditions, other than the quadratic equation. It is self-evident that it may apply. In addition, when there are a plurality of types of objects to be determined as intruding objects, and when the size of the object regarded as each intruding object is updated by the detected intruding object, the type of the detected object is It is determined whether it is present. Therefore, the size of an object regarded as another type of intruding object is not updated.

[0040]

As described above, according to the present invention, the size for judging whether an object detected by the difference method or the like is an intruding object to be monitored is determined by a reference object or an arbitrary By determining from the detection position and size of the detection object obtained at the timing of, it is easy even if the monitoring field of view, the size of the image sensor, the focal length and zoom ratio of the lens used, and the distance to the target change. And the skill required for the setting operation can be reduced, and the applicable range of the intruding object detection device can be greatly expanded.

[Brief description of the drawings]

FIG. 1 is a flowchart for explaining the operation of one embodiment of the present invention.

FIG. 2 is a flowchart for explaining the operation of one embodiment of the present invention.

FIG. 3 is a flowchart for explaining the operation of one embodiment of the present invention.

FIG. 4 is a block diagram showing a configuration of one embodiment of the present invention.

FIG. 5 is a diagram for explaining a conventional background image difference method.

FIG. 6 is a diagram for explaining a change in size of an intruding object in an input image with respect to a detection position.

FIG. 7 is a graph showing a distribution of a width of a circumscribed rectangle with respect to a detection position of an intruding object and an approximate curve thereof.

FIG. 8 is a graph showing a distribution of heights of a circumscribed rectangle with respect to a detection position of an intruding object and an approximate curve thereof.

FIG. 9 is a graph showing a distribution of a detection area with respect to a detection position of an intruding object and an approximate curve thereof.

[Explanation of symbols]

101: video selection step, 102: image input step,
103: reference object detection step, 104: size measurement step, 105: size recording step, 106: predetermined frame end determination step, 107: size determination step, 10
8: predetermined type end determination step, 109: image input step, 110: object detection step, 111: size measurement step, 112: intruding object determination step, 113: alarm / monitor display step, 114: size condition update determination step , 115: size recording step, 116: size determination step, 201: size determination step, 202: intruding object determination step, 203: alarm / monitor display step, 4
01: TV camera, 402: Input device, 403: Image input I /
F, 404: Input I / F, 405: Image memory, 406: Work memory, 407: CPU, 408: Program memory,
409: Output I / F, 410: Image output I / F, 411: Warning light, 412: Monitoring monitor, 413: Data bus, 501: Input image, 502: Reference background moving image, 503: Difference image,
504: binarized image, 505: differencer, 506: binarizer.

Continued on the front page F term (reference) 5B057 AA19 BA02 CA12 CA16 DA07 DB02 DC04 DC32 5C054 AA04 CA04 CC02 EA01 FA09 FC15 FE28 GB12 GB14 GB15 GD00 HA18 5L096 BA02 FA02 FA18 FA59 FA69 GA34 HA03

Claims (13)

[Claims] 1. An image of a predetermined reference object is captured at a plurality of predetermined positions in a monitoring visual field, and the size of the reference object detected at each of the plurality of predetermined positions is calculated. A size setting method for an intruding object detection device, comprising: setting a size of an object regarded as an intruding object in the monitoring field of view based on a detection position and a size of a reference object. 2. An intruding object detecting apparatus for detecting an intruding object in an image signal sequentially input from an image pickup device for picking up an image in a monitoring field of view, comprising: Detecting the reference object from the captured image signal, calculating the size of the reference object detected at each of the plurality of predetermined positions, and calculating the calculated reference object A size of an object regarded as an intruding object in the monitoring field of view based on the plurality of detection positions and the size corresponding to each of the detection positions. 3. The size setting method for an intruding object detecting device according to claim 1, wherein the size of the object regarded as the intruding object is a circumscribed rectangle of the detected lump of the reference object. A size of the intruding object detection device, which is at least one of a width, a height, and an area of the detection region. 4. The method according to claim 1, 2 or 3.
The method for setting the size of an intruding object detection device according to any one of claims 1 to 3, wherein the size of the object regarded as the intruding object is obtained by an approximate expression using a position of the detected reference object as a variable. How to set the size of the detector. 5. An image of a predetermined reference object is picked up at a plurality of predetermined positions in a monitoring visual field, imaged, and the size of the reference object detected at each of the plurality of predetermined positions is calculated. Based on the detection position and the size of the reference object, set the size of the object to be regarded as an intruding object in the monitoring field of view, and detect an object detected from an image signal sequentially input from an imaging device that captures an image of the monitoring field of view. And comparing the size of the object considered as the intruding object to determine whether the detected object is an intruding object, and detecting the intruding object in the monitoring visual field. . 6. An intruding object detection device for detecting an intruding object in an image signal sequentially input from an imaging device for capturing an image in a monitoring visual field, wherein an image of a predetermined reference object is detected by the imaging device in a plurality of positions in the monitoring visual field. Detecting the reference object from the captured image signal, calculating the size of the reference object detected at each of the plurality of predetermined positions, and calculating the calculated reference object Setting the size of an object to be regarded as an intruding object in the monitoring field of view based on the plurality of detection positions and the size corresponding to each of the detection positions, and detecting the object detected from the image signal sequentially input from the imaging device. And comparing the size of the object considered as the intruding object to determine whether the detected object is an intruding object, and to detect the intruding object in the monitoring visual field. Intruding object detecting method. 7. The intruding object detection method according to claim 5, wherein the size of the object regarded as the intruding object is a width and a height of a circumscribed rectangle of the detected mass of the reference object. And at least one area of the detection area. 8. The method of claim 5, 6 or 7
The intruding object detection method according to any one of claims 1 to 3, wherein the size of the object regarded as the intruding object is obtained by an approximate expression using the position of the detected reference object as a variable. 9. The method for detecting an intruding object according to claim 5, wherein the determination as to whether the detected object is an intruding object includes determining the size of the detected object as the intruding object. A method for detecting an intruding object, which is performed based on whether or not the size is within a range between a first predetermined ratio and a second predetermined ratio. 10. The object detection method according to claim 5, wherein a plurality of sizes of the object regarded as the intruding object are set, and a different type of intruding object is determined according to each of the plurality of settings. A method for detecting an intruding object. 11. The object detecting method according to claim 5, wherein a size of an object regarded as the intruding object is updated using a detected intruding object as a reference object. Object detection method. 12. An intruding object detection method for detecting an intruding object by capturing an image in a monitoring visual field and analyzing the captured image, comprising: an imaging device for capturing an image in the monitoring visual field; and a video acquired by the imaging device. An image input interface for sequentially converting signals into an input image; an image storage unit for storing the input image; and an image processing unit for processing the input image, wherein the image processing unit converts a reference object from the input image. Detect, record the detected position and size of the detected reference object, determine the size of the object to be regarded as an intruding object from the recorded detected position and size of the detected object, from the sequentially input image signal An intruding object detection device, which determines whether the detected object is an intruding object by comparing the size of a detected object with an object regarded as the intruding object. 13. The intruding object detection device according to claim 12, further comprising: input means for setting whether or not to update the size of the object regarded as the intruding object, and If the setting is made to update the size of the object to be considered, the detection position and the size of the detected intruding object are calculated, and the size of the object to be considered as the intruding object is updated. Intrusion object detection device.