An Effective Trespasser Detection System using Video Surveillance Data

Abstract

Today video surveillance applications are not only enough for object monitoring and recording but also necessary for object detection and decision making to prevent and alert unauthorized cases. The main objective of the paper is to detect a person who is passing through into a restricted area. In this paper, there are two main stages, the first stages is human detection and classification. The second stage is the detection of the passing across in a restricted area. In this proposed technique, human motion detection and tracking is implemented using Gaussian Mixture Model. Arbitrary restricted region can be drawn as polygon shape on input surveillance video clip. By comparing the foreground motion object region and the restricted region area, the pass cross event of the motion object can be detected. The accuracy and performance of the proposed technique is proved with various experimental results.