The Research On Moving Objects Detection Algorithm In Video Sequence

With the rapid development of science and technology,the video surveillance system has been applied to various fields of national economy and national defense construction,and the video surveillance system plays an important role in human life.Detecting moving object quickly and correctly in the video surveillance system is important to object tracking,object recognition and object analysis.It is also a popular subject that many scholars have studied in recent years.However,because of complexity,noise and so on,it is difficult to design and applicate moving object detection algorithm in video sequences.Many scholars at home and abroad have proposed many algorithm,this article explores the moving object detection algorithm in video sequences on the basis of predecessors' research.The main researches are as follows:1.Several moving object detection algorithms are introduced,including optical flow method,frame difference method and background subtraction method.And this paper focuses on the background modeling in background subtraction,including multi-frame average method,mixed Gaussian modeling and codebook algorithm,Vi Be.Thus lay the theoretical foundation for follow-up research.2.Concerning the problem that noise and local motion have seriously effect on the accuracy of moving object detection,a moving object detection method based on background subtraction for video sequences was proposed.The background subtraction was combined with frame difference to estimate the motion state of current frame pixels.The related pixels in the static and motion region were replaced and updated respectively.The Otsu method was used to extract moving object and the mathematical morphological operation was finally used to eliminate the noise and redundant information in the objects.The proposed algorithm can overcome the problems such as local movement and noise and has good accuracy for detecting moving objects in video sequences.3.Concerning the problem that illumination changing,noise and local motion have seriously effect on the accuracy of moving object detection,a moving object detection method based on ?-? background estimation and Kalman filter background model for video sequences is proposed.The motion state of current frame pixels is estimated by using the ?-? filter with Kalman filter background estimation model,the background pixels has been selected roughly.Then after the stable pixels are selected to build the background image which is similar to the real background image.And the absolute difference image has been calculated.Finally,?-? filter background model was used to extract moving object from the absolute difference image.Through the experimental simulation,the effectiveness of the algorithm was verified and analyzed by using quality evaluation indicators.The results show that the proposed algorithm can overcome the problems which caused by local movement and adapt the light which changed slowly.The proposed algorithm has higher precision in comparison other algorithm,which has strong robustness.