Font Size: a A A

A Research Of Moving Target Detection And Tracking Method Under Complex Background

Posted on:2014-01-03Degree:MasterType:Thesis
Country:ChinaCandidate:W WangFull Text:PDF
GTID:2268330425956854Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
With the gradual improvement of the people security awareness and the development ofcomputer related technology, video surveillance can be seen everywhere in the life. Usually, thepurpose of video surveillance is to realize the detection and tracking of moving target in thesurveillance area. Currently, the moving target detection and tracking algorithms used in videosurveillance system are mostly aim at ordinary life envrionment, which environment is relativelysimple. In terms of surveillance under complex natural background, there is still no a goodsolution. Through research, an algorithm in view of moving target detection and tracking undercomplicated nature background is proposed in this paper, which is suitable for the videosurveillance system. Main work in this paper are as follows:In terms of moving target detection, firstly, propose an edge detection algorithm based onfractal error to extract the target edge feature from the video sequences; second, propose amoving target detection algorithm based on fractal feature. This algorithm detects the targetthrough the differences of fractal dimension between target and natural background. Verified thatthe algorithm can detect a moving target under complex background, and has fast speed andaccurate result.In terms of moving target tracking, improve the conventional particle filter trackingalgorithm based on color, adopt the adaptive filter to update the moving target model, whichenhanced the robustness of target tracking. In addition, propose a target tracking algorithm ofparticle filter based on multiple features fusing adaptively. This algorithm can effectivelyovercome the problems of tracking not ideal and tracking failure due to target features changeduring the process target tracking. Verified that the algorithm is more accurate and robustness inmoving target tracking under complex background.For verifing the feasibility of target detection and tracking method proposed in this paper,we design and set up a video surveillance system based on B/S structure. Verified that themethod has a strong applicability, and can achieve good running effect in the video surveillancesystem.
Keywords/Search Tags:video surveillance, target detection, target tracking, fractal theory, particle filter
PDF Full Text Request
Related items