Design On Moving Target Detection And Tracking System

As the visual detection and tracking technology development whichprogresses by leaps and bounds in recent years. Traditional computer cable typevideo monitoring system is gradually to the intelligent visual monitoring systemin the direction of the development. Aim at traditional video monitoring systemof defects and shortcomings in now, this paper designed and implemented movingtarget detection and tracking system. The system implements the real-time videoimage processing, detect moving objects in still images, Moreover, according tothe position of the moving object control the motor rotation, make the movingobject at the center of video image all the time, finally realize the intelligent ofthe system.This article mainly discusses the moving objects video image targetdetection and tracking algorithm for implementation. Use background differencemethod and the method of frame differential method, to achieve the goal ofmobile quickly complete pattern identification. Studied in Kalman filter based onthis, the application of the mean shift tracking algorithm, which overcomes thedrawback of two algorithms, make their complementary advantages, and improvethe tracking performance of the system. To the moving target detection andtracking system hardware platform design, this paper uses the video head forreal-time video image acquisition, yuntai driven in video head tracking movingtarget, speakers sound alarm, processor computing data and processinginstructions.Simulation results show that by the final experiment, verify the validity ofthe moving object detection and tracking algorithm. When the system on movingtarget detection and tracking, less susceptible to a moving object target speed,moving object obscured, background changes of peripheral factors such as light,system did it for a long time for moving object tracking has tracking robustness, accuracy and quickness, etc. Solve the traditional moving object detection andtracking algorithm of defects and deficiencies, applicable to unmannedmonitoring environment.