A system of transmission line tower anti-theft monitoring based on video difference analysis is designed in this paper. Gaussian Mixture Model is firstly used to create dynamic background model. Then the moving objects can be extracted effectively with the background subtraction. SVM is used to detect moving targets on the quick identification. Finally, Based on regions and characteristics method,the targets of moving are tracked so as to achieve the goal of monitoring and guarding. Experimental results show that this method is accurate in Detecting, Tracking and Recognition of pedestrians about the Video Monitoring, which will timely warn the monitoring center for the suspected pedestrian to provide security for the power systems.
展开▼