The Research of The Method of Moving Object Detecting in Video Surveillance System

The intelligent video surveillance system based on computer vision has more and more widespread application, moving target detection is very important in the in-telligent monitoring system, the image quality of moving target detected will af-fect its post-processing such as identification and classification of moving object, in the video surveillance system, the traditional detection methods of moving tar-get includes optical flow method, the inter-frame difference and background dif-ference. The independent moving targets can be detected by optical flow method, but it has high complexity of computation, big disturbance from noise and poor real-time. The inter-frame difference has low complexity and the shorter time be-tween adjacent frames, therefore, the method is better suited detection in a dynam-ic environment, however, it is also easy to wrongly taken shaded background as target. The background difference is the most common method in the video sur-veillance system, but it is vulnerable to the impact of the three aspects. such as changing light, noise, the consistency of the updated background, the real back-ground and the accuracy of the being selected threshold[1][2]. Therefore, the pape