Research On Moving Object Detection And Tracking Algorithms Based On Video Sequence

The technology of video sequence based on object detection and tracking is one of the hotspots in the field of computer vision,which is also the basic in the applications of smart surveillance , human machine interface, mobile robots navigation, industrial robots hand—eye system and so on. Nowadays, Video surveillance system still depends on manual operation, which wastes resources and affects the efficiency. So studying the typical algorithms used in video surveillance and designing an intelligent video surveillance system is very important.The algorithms of moving object detection and object tracking are researched and a multi—objects detection and tracking software system based on video image sequence is realized in this paper.This dissertation introduces the history and current situation of video surveillance technology; describes the existing method of moving target detection and tracking of the basic classification and difficulties; analyzes the moving target detection and tracking algorithms, proposes a modified adaptive Kalman filter tracking algorithm based measurement covariance and completes the moving target detection and tracking software system based on video sequence. The experimental results demonstrated that the detection and tracking system which was put forward by this paper has the advantage of high accuracy, real-time performance and a better robustness.On moving object detection,main concern is consecutive frames difference,optical flow and background subtraction.We construct a moving target detection module, which is suitable for intelligent surveillance system based on the summarizing usual target detecting methods. A modified selective updating model is proposed as the reliable adaptive background updating method,and we use the background subtraction to detect the motion information. After the motion detection operation, morphologic filtering and connected region area measurement are introduced to suppress the noise and solve the background disturbance problem.Finally the moving objects are extracted reliably. On the basis of having fully studied all the tracking methods available, we propose a modified adaptive Kalman filter tracking algorithm based measurement covariance and implement it. It adjusts covariance parameters dynamically to predict the true state of the target accurately, and the prediction have has the advantage of real-time performance and robustness to achieve a fast moving target tracking.