Research Of Video Object Tracking And Implementation Based On Mean Shift And Particle Filter Technology

Video tracking technology is currently emerging as a research area, which coversmachine vision, pattern recognition, artificial intelligence and many other academiccontent. Not only widely used in video surveillance, human-computer interaction, videocompression and retrieval, medical image processing and other fields has also been widelyused. Target detection, tracking, video surveillance and behavior recognition as theprimary research content, a machine vision research focus.In this paper, research and innovation are as follows:1. frame difference, the meanshift and particle filter video object tracking technology to conduct adetailed comparative analysis of the strengths and weaknesses of eachmethod to makecertain improvements strategy.2. Lack of commonly used video tracking technology toanalyze, and in a specific experiment, for a variety of mathematical models to improve inorder to improve the timeliness and robustness of the algorithm.3. Based on Vc++platform, combined with opencv and the gsl technology we development a video objecttracking system, to carry out specific research on the different tracking.The actual target tracking technology is significant, driving other research areas incomputer vision forward. Video object tracking technology research with the ultimate goalis to make the machine function similar to human visual perception, to find the target inthe image sequence, tracking goals, and objectives to identify and understand the behavior.In this area, there are still many problems to be solved. Future studies needto beconstantly improved and perfected.