Resumen
The article presents the results of the development of a motion detection method taking into account the filtering of noise in the frame. Based on the analysis of literature sources, a combined approach to motion detection based on classical and heuristic algorithms was proposed. In the course of the work, both classical approaches to motion detection in video sequences and existing ready-to-use solutions in the field of intelligent video surveillance systems were considered. As a result, a motion detection algorithm was developed that solves the problem of a large number of false positives. Distinctive features of the developed algorithm are resistance to noise and minor movements in frames of video sequences, as well as the ability to easily adapt to various changes in the motion detection environment. The main advantage of the proposed modification of the motion detection algorithm is a sufficiently high processing speed of each frame for a video analytics system in real time.