Camera shielding detection method

Abstract

The invention relates to the technical field of digital image processing, in particular to a camera shielding detection method. The camera shielding detection method includes the steps of 1, building a camera resource configuration module; 2, enabling a monitoring center to judge whether a camera is in operation or not according to operation state information, and to acquire a current-frame digital image from the camera in the operation state; 3, performing image preprocessing; 4, building an image brightness background, smoothing the image to a smooth brightness distribution diagram by an average filtering method, and extracting possibly shielded sectors Si of the image by a dynamic threshold method; 5, subjecting the possibly shielded sectors Si to connected component analysis, setting a sector threshold value ST, selecting the possibly shielded sectors Si not less than the sector threshold value ST, and defining the selected possibly shielded sectors Si as high-probability shielded sectors S2; 6, computing shielded probability P of the high-probability shielded sectors S2; 7, judging the relation between the shielded probability P and preset shielded probability PT, if P>PT, judging the high-probability shielded sectors as final shielded sectors, and if not, judging the high-probability shielded sectors as unshielded sectors. The camera shielding detection method is high in detection efficiency.

Description

technical field [0001] The invention relates to the technical field of digital image processing, in particular to a camera occlusion detection method. Background technique [0002] With the widespread use of the monitoring system network, a large number of cameras are deployed at thousands of monitoring points in the monitoring system network, and in terms of the number of cameras, there are as many as tens of thousands. Due to the large number of cameras and the fact that most of the cameras at monitoring points are installed in high places and hidden places, if no one maintains them for a long time, it will lead to power failure, blurring, blocking, etc., which will affect the normal use of monitoring points. How to intelligently detect a large number of cameras has become the goal pursued by the industry. In the existing technology, image processing is performed on the data uploaded by the cameras through the monitoring center, and based on the image processing detection ...

AI Technical Summary

Problems solved by technology

[0003] The detection technology for occlusions in the prior art mainly has the following defects: because there are many kinds of occlusions, such as chewing gum, paper and cloth strips, etc., which are deliberately made by some people, and also naturally generated by the environment, such as dust, mud and stains, etc. The size and shape are different, the size of the occluder will affect the detection effect, and the occlusion area of ​​certain shapes and sizes cannot be detected

And the change of ambient light will also interfere with the detection results.

The existing detection technology will also misjudge some non-occluded objects

Images