Research On Moving Object Detection And Tracking Algorithm In Surveillance Video

At present,the application of video surveillance systems has become more and more extensive,and it has gradually developed toward the direction of intelligence,which brings great convenience to people's lives.Moving target detection and tracking are two core technologies in intelligent video surveillance systems.Many excellent algorithms have emerged,but there are still many challenges that need to be further improved.In the aspect of moving target detection,a target detection algorithm based on mixed Gaussian model and average background method is proposed.When the background image is extracted,the algorithm uses the idea of average background to deal with the background images extracted from the mixed Gauss model to make the acquired background image more pure.The method of contour screening is used to remove background interference and can effectively remove the tiny isolated and slightly larger noise points existing in the background.Experimental results show that the proposed algorithm has a high degree of completeness of the target when the target stays in a short pause and rapid movement,and when the target scene exists many disturbances,the effect of moving the target interference point is also good.At the same time,the algorithm has higher accuracy rate,overall detection rate and comprehensive evaluation index.In the aspect of moving target tracking,a target tracking algorithm that combines feature point detection and spatiotemporal context information is proposed.The algorithm can judge whether the target is blocked by the result of feature point matching.It can also correct the errors in the process of tracking the spatio-temporal context and overcome the shortcoming that the spatio-temporal context method can't self-correct the error tracking result.The experimental results show that the proposed algorithm improves the robustness of target tracking for occlusion and solved the problem of easily causing target tracking offset or even failure when the target is moving fast.In addition,the algorithm has smaller center position error,higher distance accuracy and overlap accuracy when tracking target.The improved algorithm mentioned above solves some problems in detection and tracking,and through the analysis of the performance of the algorithm,the algorithm can improve the performance while solving these problems.Compared with the detection and tracking algorithms used in the comparison,the performance of this algorithm is better.