Researches On Key Technology Of Intelligent Video Surveillance

Intelligent video surveillance technology will improve the degree of automation of traditional video surveillance system. It has significance for economy, national defense and public security. Detection and tracking of moving objects is the key technology of intelligent video surveillance. This paper focuses on the two aspects. Through analyzing the noise and the structure of the difference of adjacent images, a method of automatic estimation of noise parameters is brought forward. And it is put into the use of detecting moving object. The methods of analyzing static background are studied. The theory of estimating static background is improved and consummated. The result of the background estimation is also put into the use of objects detection. We have studied and improved bounding box tracking, Kalman tracking, and dynamic appearance model. We use these three methods together to complete the task of objects tracking.