A shape-perceived object tracking algorithm for intelligent surveillance systems

Abstract

In this paper, we present an effective algorithm design for object tracking in intelligent surveillance applications. The proposed shape-perceived algorithm uses the building-block-based matching method. This method is fast due to its use of one-pass scanning and its low-cost of object classification. Two kinds of cameras are designed for different goals, a wide-angle camera and a PTZ camera, to construct the intelligent surveillance system and to achieve behavior analysis and body recognition. Furthermore, the experimental results show that this low-cost shape-perceived object tracking architecture is feasible for intelligent surveillance systems.