JPS62147891A - Area-dividing method for monitoring picture - Google Patents

Abstract

PURPOSE:To simplify the constitution, speed up the action, and decrease the cost of the titled method by dividing the area of a monitoring picture by means of a combination of the boundary coordinates of valid projection area on axis X and axis Y, deciding a valid or an invalid picture area by checking the presence or absence of the valid data of respective division areas. CONSTITUTION:The digital picture data of a monitoring picture A taken up by an image pickup means is added up in each column and line. Projection data Px and Py on the axes X and Y are obtained from the resulting addition- data, and parts of the data that exceeds thresholds tx and ty are made valid projection areas on the axes X and Y which are Px1, Px2...Pxi, and Py1, Py2...Pyi.... By means of the combination of the boundary coordinates x11, x12, x21, x22...xi1, xi2...y11, y12, Y21, y22...yi1, yi2... between the said valid projection areas, the monitoring picture A is divided into areas. Whether or not a valid data exists in the division areas B1-B4 whose upper left coordinate is (xi1, yj1) and lower right is (xi2, yj2), is checked. Such ones of the division areas B1-B3 that include the valid data are decided to be the valid picture areas, but one B4 not including the valid data is decided to be the invalid picture area and abandoned.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field 1] The present invention relates to a method for dividing a surveillance image into regions of a surveillance system.

[Background Art] Generally, in a TV monitoring system that monitors the situation based on monitoring images taken with a TV camera,
Changes are detected by taking the difference between the background monitoring image captured by a TV camera and the current monitoring image that is input sequentially, and using projection data obtained by projecting the image data in the difference image in the X-axis and Y-axis directions. In such cases, a threshold level is set to determine whether there is a change in each projection, and the width of the difference between the maximum and minimum coordinates exceeding this threshold level is defined as the area where the change has occurred. It was like that. However, in such a conventional example, even though a change occurs in a plurality of regions, it is regarded as a single change, so there is a problem that accurate change information cannot be obtained. Therefore, as shown in the fJS2 diagram, first, the monitoring image is scanned line by line using a rask operation, and the changed areas detected in each line are sequentially labeled starting from 1. At the same time, each detected area is labeled in one line 611. There is a method called the segment method, which uses boundary tracking to determine whether a label overlaps with a labeled region, and then combines each labeled region based on the determination result to create a region segmentation method. For example, region 3 is divided into a set of segments labeled 3 and 4. However, in such a conventional example, there is a problem in that the algorithm for dividing is complicated and the processing time is long. Similarly, there is a contour tracking method, but the algorithm for segmentation is complex and takes a long time to process.If speeding up is required, dedicated high-speed processing hardware is required, resulting in high costs. There was a problem.

[Object of the Invention] The present invention has been made in view of the above points, and its purpose is to be able to easily and quickly perform area division processing, and to realize it using inexpensive hardware. ,
Furthermore, it is an object of the present invention to provide a method for dividing a monitoring image into regions that can reduce the storage capacity of a storage means for storing divided regions.

Disclosure of the Invention 1 (Example) Fig. 1 shows the method of the present invention, in which digital image data of a surveillance image A taken by an imaging means such as a television camera is cumulatively added for each column and each row. Then, projection data P
y is calculated and this projection data Px.

The part where py exceeds the threshold Lx and ty are the effective projection area P x1'*P X2 on the X and Y axes, respectively.
1...tPxi,..., P F+ * P V2t
...tPV'+*...binding, both effective projection areas PK,
, P x2...Pxi..., Py,, Py2...
...Pyi... boundary coordinates Xz+X+2yXz+*X
2z*・・'*XL*'XIzt”・,y1+1y12
13'2++Fz2t...*'/1IyViz*...
The surveillance image A is divided into regions by the combination of ・, and each divided region P x, tP
P X11””, Pyl, Py2. ..., Pyi,・
It is reasonable to check the presence or absence of valid data within ... to determine the valid image area and invalid image area,
In the illustrated example, these boundary coordinates are zwX+2+Xzth×2
2. By all the combinations of y-th V+z+Fz+*F2z, the upper left on the monitoring image A is (xi,, yL)
, rectangular divided area B whose lower right corner is (X12.Fj2)
,~B.

is determined, and it is checked whether there is valid data in the divided areas 81 to B, and if there is valid data, the SW area B,
~B is determined to be a valid image area, and 'tk 1411 area B, where there is no valid data, is determined to be an invalid image area and discarded. Note that the thresholds tx and ty are set to levels that allow noise in the monitoring image A to be reliably removed.

Now, in the method of the present invention, depending on whether the projection data exceeds the thresholds tx, ty, the effective projection area P x, , P xj+・P xi−1PF+v
Divide P y2"'Pyi..., and divide the effective projection area PX+*Px2, the boundary 7J of py eye PF2! To the target 9 +2! to the 2nd eye 22 turtle y mouth gy+2*S'z+*F
The divided areas 81 to B4 of the monitoring image A are set by the combination of 2□, and the amount of data handled in the area dividing process can be reduced, so the area dividing process can be performed easily and quickly, and is also inexpensive. It is now possible to implement it in hardware.

Furthermore, since each of the divided areas 81 to B4 is rectangular, when storing divided areas B and -B, it is only necessary to memorize the coordinates of the upper left and lower right of the divided areas B, -B, and so on.
, ~B, memory of the storage means that stores it! This allows the amount of ff to be reduced.

Furthermore, in the next-stage image change detection means, parameters such as the center of gravity and area of the image data of only the necessary divided areas 81 to B can be determined, and change detection can be performed more reliably.

[Effect of the Invention J As described above, the present invention obtains projection data on the X-axis and Y-axis of digital image data of a surveillance image taken by an imaging means such as a television camera, and calculates the projection data on the X-axis and the Y-axis. The area beyond the predetermined threshold is defined as the effective projection area on the X and Y axes, respectively, and the monitoring image is divided into areas by combining the boundary coordinates of both effective projection areas, and the effective projection area within each divided area is This method determines valid image areas and invalid image areas by checking the presence or absence of data, and since the amount of data handled during area segmentation processing can be reduced, area segmentation processing can be performed easily and at high speed. It has the advantage that it can be realized using inexpensive hardware, and since each divided area is a rectangle, it is only necessary to memorize the coordinates of the upper left and lower right of the divided area, and the memory that stores the divided area This has the effect of reducing the storage capacity of the means.

[Brief explanation of drawings]

FIG. 1 is an explanatory diagram of the operation of the method of the present invention, and FIG. 2 is an explanatory diagram of the operation of the conventional example. A is a monitoring image, and 81 to B4 are divided areas. Agent Patent Attorney 1) Appearance 7 Figure 1 Figure 2 Procedures Supplementary JE (Spontaneous) February 15, 1986

Claims (1)

[Claims] (1) Obtain the projection data on the X-axis and Y-axis of the digital image data of the surveillance image taken by an imaging means such as a television camera, and The effective projection area is defined as the effective projection area on the axis and the Y axis, and the monitoring image is divided into areas based on the combination of the boundary coordinates of both effective projection areas, and the presence or absence of valid data in each divided area is checked to determine the effective image area and A method for dividing a surveillance image into regions, characterized in that invalid image regions are determined.