JP3045406B2 - Surveillance image change detection method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting a change in a monitored image for remotely monitoring an intruder from the outside.

[0002]

2. Description of the Related Art When an intruding object is detected by image processing and an intruding object is determined, an area of a changed portion on a screen is calculated, and if the area is larger than a specified value, it is determined that an intruding object is present. The method of doing is general.

[0003] This method is roughly classified into a method in which the determination is made only by the area of the changed portion on the screen and a method in which the determination is made by obtaining the actual size of the intruding object from the position on the screen. Use under conditions that are limited so that the size of the intruding object on the screen can be predicted to some extent (for example, only the doors for entering and leaving the room, and only the windows are set as the monitoring area). In order to know the actual size of the intruding object, monitor the same area with multiple cameras, calculate the distance from the camera to the object from the difference in the size of the intruding object captured by each camera,
From this distance and the focal length of the camera lens, the actual size of the intruding object is calculated. In order to know the actual size of the intruding object, the coordinates of the monitoring area on the screen and the distance to the camera are input in advance.

[0004]

In the method described above, the size of the intruding object on the screen cannot be used unless the size on the screen is known in advance. In the method (1), a plurality of cameras are required, so that the cost is increased. Since the focal length of the lens and the position of the camera must be determined in advance, the monitoring area cannot be set freely. In this method, the focal length of the lens must be determined in advance. In addition, when the monitoring area is large, it takes a lot of trouble to measure and input the distance from each coordinate on the screen to the camera. An object of the present invention is to solve the above problems.

[0005]

According to the present invention, there is provided a method for detecting a change in a monitoring image, comprising a monitoring camera for capturing an image of a monitoring area, an image processing monitor for displaying a processing result of the captured image, and an image processing device. The image processing monitor screen is divided into areas, a reference object whose length or area is measured in advance is placed in the area, and a reference value for each area is stored, and a change occurs in the monitoring area in the monitoring state. When the image processing device calculates in which area of the divided area of the image processing monitor the change is located, calculates the area of the changed image based on the reference value of the area, An alarm is issued when the area is within a predetermined range.

[0006]

The actual area of the intruding object can be obtained by dividing the area of the monitor screen and placing a reference object whose length and area have been measured in advance in the area, and making a judgment by comparing with the reference value. it can.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be specifically described with reference to embodiments shown in the drawings. FIG. 1 shows an embodiment of the present invention. FIG. 1A is a block diagram of an image processing apparatus.
(B) is an explanatory view showing an example of a monitoring area. FIG. 1 (a)
, An image processing apparatus 3 includes an image processing monitor 1 for displaying a processing result of an image and an operation monitor 2 for instructing an operation procedure.
A monitoring camera 4 for capturing an image of the monitoring area 7 shown in (b), a mouse 5 for inputting a reference value, and a keyboard 8 for inputting an operation command are connected. In this example, the reference object to be imaged in the monitoring area 7 is a reference bar 6 having a length of 1 m.

First, a reference value is input. The operation procedure is displayed on the operation monitor 2. According to the first instruction (see FIG. 6), the image of the image processing monitor 1 is viewed and the monitoring camera 4 is set (FIG. 2A). When the setting of the camera is completed, the number of divisions of the image on the image processing monitor 1 is selected from the display on the operation monitor 2, and is input using the keyboard 8.

When a key input is made, the image from the camera 4 and the selected number of cross hatch images are combined and output to the image processing monitor 1 (FIG. 2B).
Displays the next instruction (FIG. 7). In this example, four divisions are selected. In general, the number of divisions is set to be large when the depth of the monitoring image is deep, and small when the monitoring image is shallow.

[0010] Looking at the synthesized image, press the "G" key if it is good, and press the "ESC" key to change the number of divisions. When the "ESC" key is pressed, the image processing monitor 1
The cross hatch disappears from the image displayed in FIG. 2 (FIG. 2B) (FIG. 2A), the image on the operation monitor 2 also returns to the previous state (FIG. 6), and the number of divisions can be input again. When the "G" key is pressed, a cross cursor is displayed on the image processing monitor 1 (FIG. 2C), the next instruction is displayed on the operation monitor 2 (FIG. 8), and input from the mouse 5 is awaited. . FIG. 13 shows a flowchart for selecting the number of divisions.

A reference bar 6 is set up vertically on an area to be monitored among the areas divided by the image processing monitor 1. When the mouse 5 is operated to position the cross cursor at one end of the reference bar 6 (FIG. 2D) and left-clicked, a point is drawn at the position of the cross cursor (FIG. 2E). Next, reference bar 6
When the cross cursor is positioned at the other end of the image (FIG. 2F) and left-clicked, a straight line is drawn between the previously drawn point (FIG. 2G).

The image processing device 3 calculates the area in which the straight line is drawn based on the coordinates of the lower end of the straight line, and displays the area number and the length of the straight line on the operation monitor 2 (FIG. 1).
2) Remember. The area numbers are assigned in order from the upper left for convenience (FIG. 2H). For example, the calculation in the case of four divisions is such that if the x coordinate of the lower end of the straight line to be examined is larger than の of the number of pixels in the horizontal direction of the screen, it is in the right half of the screen (FIG. 2I), and the y coordinate is vertical. If it is smaller than half the number of pixels in the direction, this point is in the upper half of the screen (FIG. 2J),
It can be seen that the position of this straight line is the area 1 at the upper right of the screen (FIG. 2K).

After drawing a straight line in one area, the reference bar 6 is moved to the next area, and the input is performed in the same manner (FIG. 3).
A). Correction after drawing points and straight lines is performed as follows. After the point is drawn, if the position of the point is shifted (FIG. 3B), the mouse 5 is right-clicked to delete the point (FIG. 3C). After drawing the straight line, if the position of the straight line is shifted from the reference bar 6 (FIG. 3D), right-click is performed to delete the line (FIG. 3E). At this time, the point at one end of the straight line remains as it is, indicating that the length of the straight line stored in the operation monitor 2 has been deleted (FIG. 9). Further, when the user tries to draw a straight line again in the area where the line has been drawn, the movement of the cross cursor stops (FIG. 3F), and a message asking whether to overwrite is displayed on the operation monitor (FIG. 10).
Pressing the "Y" key deletes the previous line, leaving only the new line (FIG. 3G), and stores the length of the new line as the reference value for this area. When "N" is pressed, the straight line to be newly drawn is deleted (FIG. 3H).

Enter a reference value in all areas, or
When the "ESC" key is pressed, the reference value input processing is terminated, and a confirmation message as to whether to create a data file is output on the operation monitor (FIG. 11). When the "Y" key is pressed, the length of the straight line is saved in the area where the straight line is drawn, and "0" is saved in the data file together with the selected number of divisions in the area where the straight line is not drawn. When the "N" key is pressed, the reference value is again input (FIG. 8), and the straight line can be drawn again. FIG. 14 shows a flowchart of the reference value input.

The data file is read into the memory of the image processing device 3 when the intrusion monitoring state is set. When a change occurs in the monitoring area (FIG. 4A), the image processing apparatus 3 extracts the changed area (FIG. 4B), and determines where the coordinates of a specific point of the area are located in the divided area. Calculate (FIG. 4C). Further, the reference value of the area number is searched, and the actual size is calculated from the measured value of the size of the extracted area on the screen and the reference value. For example, in the case of FIG. 4A,
The actual size is calculated as follows.

FIG. 4C shows this changed area as area 1
Is determined to belong to The reference value for area 1 is 50,
Assuming that the height of the extracted region on the screen is 80 and the area is 1500, the reference bar 100 cm is 50 cm above the screen in this area.
Then, take the ratio, and 100 cm: 50 = xcm: 80 ∴x = 160 cm Therefore, the actual height of the extracted region is 160 cm. The area is the square of the ratio: 100 ² cm ² : 50 ² = x: 1500 @ x = 6000 cm ² Therefore, the actual area of the extracted region is 6000 cm ²
It is.

In this way, the intruding object is determined based on the actual size of the changed area. When there is a change outside the monitoring area (FIG. 5A), the actual size is as follows. First, the changed area is extracted (FIG. 5).
B), calculate the area of the changed area (FIG. 5C). At this time, since the reference value of this area 0 is 0, the above calculation is not performed, and the actual height and area are set to 0.

Therefore, even if there is a change on the screen outside the monitoring area, the change is canceled in the process of processing by the image processing device 3, and no erroneous determination is made. FIG. 15 shows a flowchart for obtaining the actual size.

[0019]

As described above, the present invention has the following effects. There are no special installation conditions for performing image processing,
The monitoring area can be set freely. Therefore, even when image processing is added to the conventional image monitoring system, there is no need to change the camera settings. After the setting of the camera is completed, the actual size of the intruding object can be known only by inputting the reference value according to the instruction on the monitor screen, so that the monitoring area can be easily changed. By inputting the reference value, it is possible to simultaneously mask an area (eg, a tree or a cloud) of a factor that causes a change other than the intruding object.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration example of an apparatus for implementing the present invention.

FIG. 2 is an image processing monitor screen of a monitoring area showing a processing procedure according to the present invention.

FIG. 3 is an image processing monitor screen showing a processing procedure in another area.

FIG. 4 is an image processing monitor screen showing a processing procedure when a change occurs in a monitoring area.

FIG. 5 is an image processing monitor screen showing a processing procedure when a change occurs outside the monitoring area.

FIG. 6 is an operation monitor screen showing contents for instructing an operation.

FIG. 7 is an operation monitor screen showing contents for instructing an operation.

FIG. 8 is an operation monitor screen showing contents for instructing an operation.

FIG. 9 is an operation monitor screen showing contents for instructing an operation.

FIG. 10 is an operation monitor screen showing contents for instructing an operation.

FIG. 11 is an operation monitor screen showing contents for instructing an operation.

FIG. 12 is an operation monitor screen showing contents for instructing an operation.

FIG. 13 is a flowchart for selecting the number of divisions.

FIG. 14 is a flowchart of a reference value input.

FIG. 15 is a flowchart for obtaining an actual size.

[Explanation of symbols]

 Reference Signs List 1 image processing monitor 2 operation monitor 3 image processing device 4 monitoring camera 5 mouse 6 reference rod 7 monitoring area 8 keyboard

──────────────────────────────────────────────────続 き Continuing on the front page (72) Ryoichi Matsunaga, Kyushu Electric Power Co., Inc. 2-82, Watanabe-dori, Chuo-ku, Fukuoka City, Fukuoka Prefecture (56) References JP-A-2-71397 (JP, A) JP-A-1-153903 (JP, A) JP-A-62-222108 (JP, A) (58) Survey Field (Int.Cl. ⁷ , DB name) G06T 7/60 G08B 13/196 H04N 7/18

Claims (1)

(57) [Claims] 1. An image processing apparatus comprising: a monitoring camera that captures an image of a monitoring area; an image processing monitor that displays a processing result of the captured image; and an image processing device. A reference object whose length or area is measured in advance is stored, and a reference value for each area is stored. When a change occurs in the monitoring area in the monitoring state, the change in any area of the divided area of the image processing monitor. Is calculated by the image processing apparatus, the area of the changed image is calculated based on the reference value of the area, and an alarm is issued when the area is within a predetermined range. A method for detecting a change in a monitoring image, the method comprising: