PERSON DETECTION IN SURVEILLANCE VIDEOS ? A SURVEY

The main goal of this paper is to compare various person detection approaches used in video surveillance applications. Over the last decade, person detection has made remarkable progress. The paper gives an overview of some of the successful techniques for people detection. In this paper we discuss about the current methods involved in human detection in fixed cameras used for surveillance purposes. Sliding window approach scans the image at all relevant positions and scales to detect a person. There are two stages in part based methods. First uses low level classifiers to identify the parts and second connects based on specified topology.