Optimal Placement of Cameras in a Distributed Video Surveillance System

Abstract

The relevance of the topic is determined by the widespread development of distributed video surveillance systems characterized by a large number of video cameras and located on a large territory. The efficiency of distributed video surveillance systems depends on a variety of factors, such as the recognition algorithms used, the technical characteristics of cameras and data processing facilities. Also, the quality of the video surveillance system largely depends on the topology of the placement of video cameras. The objective function minimizes the number of cameras while providing the required level of veracity of event recognition. It is believed that the quality of recognition mainly depends on the spatial resolution of the recognized object on the video frame. This parameter is determined by the technical characteristics of the camera and the distance to the recognized object. When crossing the field of view of several cameras and the possibility of combining their results increases an estimation of veracity of object recognition. Thus, when locating cameras in a controlled area, it is necessary to take into account a number of parameters: coordinates, viewing angles, areas of intersection of the field of view of cameras. This problem is characterized by high computational complexity. Therefore, to obtain the result for an acceptable time, an algorithm for optimizing the topology based on the method of swarm particles is proposed. The presented approach to optimize camera placement topology will reduce the time of designing a video surveillance system.