Wireless webcam based car burglar detection system

摘要

This paper presents a wireless webcam based smart visual surveillance system with real-time moving object detection, classification and tracking capabilities for car burglar detection. This image processing based car detection system can handle object detection, classification and tracking in both indoor and outdoor environments. The classification algorithm makes use of the temporal tracking results and the shape of the detected objects to accurately classify objects into classes such as human, human group and vehicle. The purpose of designing a smart surveillance system that is based on the integration of motion detection and visual tracking is to achieve a better performance in helping the security officers to prevent car burglar crimes from escalating. Experimental results show that the proposed car burglar detection algorithm is able to achieve accuracy as high as 96.2% in optimum setting with routine time of 1.28 seconds.