Motion Detection Algorithm for Agent Interaction Surveillance Systems

Motion detection is the essential process in the extraction of information from moving objects in automatic video surveillance system. In this work, a novel and accurate approach is used to detect moving object from video. It contains two phase background matching techniques such as rapid matching and accurate matching for background modelling. The rapid matching technique is used to quickly find out difference between the current and previous incoming video frames. The accurate matching is used to compare the current incoming frame and current background candidates. The background module is used to produce optimum background pixels for the background model. Next, our proposed AT module eliminates the unnecessary examination of the entire background region allowing the subsequent OE module to only process block containing moving objects. Finally, the OE module forms the binary object detection mask in order to achieve highly complete detection of moving objects. The detection results produced by our proposed method were both qualitatively and quantitatively analyzed through visual inspection and for accuracy, along with comparisons to the results produced by other state of-the-art methods. Index: Motion Detection, Video Surveillance System, Alarm trigger, Object Extraction Model.