A descriptive geometry based method for total and common cameras fields of view optimization

Abstract

The presented work is conducted in the framework of the ANR-VTT PANsafer project (Towards a safer level crossing). One of the objectives of the project is to develop a video surveillance system that will be able to detect and recognize potential dangerous situation around level crossings. This paper addresses the problem of cameras positioning and orientation in order to view optimally monitored scenes. In general, adjusting cameras position and orientation is achieved experimentally and empirically by considering geometrical different configurations. This step requires a lot of time to adjust approximately the total and common fields of view of the cameras, especially when constrained environments, like level crossing environments, are considered. In order to simplify this task and to get more precise cameras positioning and orientation, we propose in this paper a method that optimizes automatically the total and common cameras fields with respect to the desired scene. Based on descriptive geometry, the method estimates the best cameras position and orientation by optimizing surfaces of 2D domains that are obtained by projecting/intersecting the field of view of each camera on/with horizontal and vertical planes. The proposed method is evaluated and tested to demonstrate its effectiveness.