Research and Design of Moving Target Detection and Tracking System Based on Web Server

摘要

We present an algorithm to improve the detection and tracking of moving target. Combined with the ORB feature detection operator, the center of the target is estimated, and a kind of anti-shadow interference and good matching target detection algorithm is designed. Combined the adaptive filter, a motion object tracking algorithm that resists background color interference and resists short-term occlusion is designed. This paper implements target tracking on embedded ARM-Linux system, completes the design of image acquisition module, image preprocessing design, transplantation of moving target detection and tracking algorithm, and design of video transmission module based on RTP protocol. Finally, web client is designed to view the system’s tracking effect. The experimental results show that the system algorithm has a certain ability of anti-illumination and background color interference, and it has adaptability to the short-term occlusion of the target and improves the algorithm detection accuracy, increasing processing time accordingly, decreasing tracking error, with good real-time and accuracy.