Black edge detection method and device for video image

Abstract

The invention provides a black edge detection method and a black edge detection device for a video image. The method comprises the following steps of: judging whether a black index value of each edge of the image belongs to a preset black edge range value or a preset non-black edge range value or not sequentially from each edge of the image to the center; when the current detected edge of the image does not belong to either the black edge range value or the non-black edge range value, determining that the current detected edge is an undetermined edge; continuing judging whether black index values of a preset number of edges adjacent to the undetermined edge in a direction close to the center of the image belong to the black edge range value or not; when a black index value of at least one of the preset number of edges belongs to the black edge range value, determining that the undetermined edge is a black edge, and continuing executing the sequential judgment step; and when all the black index values of the preset number of edges do not belong to the black edge range value, determining that the undetermined edge is a non-black edge, and stopping executing the sequential judgment step. By the method and the device, detection robustness is enhanced, and black edge detection accuracy is improved.

Description

technical field [0001] The invention relates to the field of data services, in particular to a method and device for detecting black borders of video images. Background technique [0002] In the process of video processing, especially in video surveillance image processing, black borders often appear in the image. For different situations, the black borders of video images may exist in different positions. For example, for video images caused by image scaling processing, the black borders may only exist on the left and right or up and down of the image; for images of video surveillance scenes, the black borders may It exists on the top, bottom, left and right of the image. At present, there are several black edge detection methods for video images, which are briefly described as follows. [0003] Method 1: Threshold detection method, which compares the pixel value of each row or column along the edge of the video image. When the pixel value exceeds a certain threshold, it ...

AI Technical Summary

Problems solved by technology

The difference comparison method expands the proportion of non-black pixels in the overall summation due to the calculation of the square. Compared with the first scheme, its detection stability is higher, but because the pixel value of the black point is not the theoretical zero value, if When the pixel value is zero, it is regarded as pure black, which will cause a large number of square values ​​to be calculated for black pixels. In this way, when the pixel value of black pixels fluctuates, the sum of the square values ​​will also fluctuate greatly. Therefore, the threshold value of the interpolation method is not stable enough, making the detection accuracy rate not high

[0005] It can be seen that the above-mentioned black edge detection method of the video image that only relies on the threshold value is difficult to reasonably determine the threshold value due to the difference between the pixel value of the black point in theory and practice, which in turn leads to a low detection accuracy rate.

Images