Intrusion detection method applied to intelligent video surveillance of long-distance track lines

Abstract

The invention relates to an intrusion detection method applied in the intelligent video monitoring of long-distance railway lines, which belongs to the field of video monitoring. The method is implemented by the following steps: (1) initialization: by a perspective modeling step and a background leaning step which are parallel, perspective zooming factors corresponding to the different positions of a railway line in video frames and relatively stable monitoring field background images are obtained, and if the perspective zooming factors and the monitoring field background images are already obtained by initialization, then initialization does not need to be repeated; (2) intrusion detection: intrusion detection includes a detection alarm step and a background updating step which are parallel, wherein detection alarm is composed of four substeps, i.e. foreground detection, form correction, perspective restoration and threshold determination, and the perspective restoration can restore moving objects intruding at different positions on the railway line into a unified scale, so that the affection of the perspective phenomenon is eliminated. Compared with the prior art, the intrusion detection method can eliminate the affection of the perspective phenomenon, so that all the moving objects appearing at different positions on the railway line can submit to the constraint of a unified threshold range, consequently, false alarm and missing alarm caused by the perspective phenomenon can be greatly reduced, and thereby the long-distance railway line can be effectively protected.

Description

technical field [0001] The invention belongs to the field of video monitoring, in particular to an intrusion detection method applied to intelligent video monitoring of long-distance track lines. Background technique [0002] With the wide application of video surveillance systems, video surveillance has become an important technical means of security protection for state organs, enterprises and institutions, and residential buildings. Traditional video surveillance mainly relies on manual monitoring, and it is difficult to avoid false alarms or missed alarms caused by fatigue or off-duty personnel on duty. With the improvement of computer performance and the development of image processing technology, intelligent video surveillance shows outstanding advantages such as strong alarm initiative and long-term work, which has gradually become a development trend. [0003] The core application of intelligent video surveillance is intrusion detection. Its function is to detect in...

AI Technical Summary

Problems solved by technology

However, this approach significantly increases the overall system investment

Images