Path clustering using Dynamic Time Warping technique

In order to monitor the safety and security of an area, video surveillance system is deployed and implemented in the said area. Such video surveillance system usually relies on the detection of suspicious behavior that is captured by the surveillance camera. In this paper, we present a novel method for clustering similar trajectories in video surveillance system. The purpose of performing trajectories clustering is to build a path model which can be used to detect any suspicious activity in the monitored scene. Path models are learnt from the accumulation of trajectory data over long time periods, and can be used to augment the classification of subsequent track data. This approach does not employ the traditional way of constructing path model, yet it simplifies the computation in the process of clustering similar trajectories by calculating average path for each detected and matched trajectory. The results demonstrate the efficiency of the proposed approach in clustering the path for detecting deviant walking paths.