Research On The Technique Of Moving Target Detection And Tracking Based On Image

With the acceleration of information technology, electronic eye and surveillancecan be seen everywhere whether it is on the traffic road, in community housing, in themall or in the banks. They gradually come into people’s sight and protect the safety ofpeople with recording the daily life. The technologies cannot be made into truth ifthere is not the research on moving target detection and tracking.Although, many scholars are dedicated to the research work of moving targetdetection and tracking, and get a lot of great achievements, but there are still manyproblems to be solved on the study of the technology. The thesis studies on movingtarget detection and tracking based on image on the basis of the previous research.The main work is done as follows:Firstly, focuses on background subtraction by analyzing the optical flow, framedifference and background subtraction on the moving target detection. Improve thebackground method by plus an adaptive threshold on the original fixed one, reducingthe effect of light change on moving target detection. Then, experiments are taking toanalyze the impact of the learning rate, the number of Gaussian and the threshold ofbackground model for background modeling in a single Gaussian model and Gaussianmixture model, and an adaptive learning rate are used to improve the result.Secondly, analyze the performance of the filtering algorithm used on thedetection result with experiment, then, choose the morphological operations toremove noise. Make use of blob analysis to locate the target in the frame of currentvideo. This can make good effect.Finally, focus on the three typical algorithm used in target tracking on the basisof the basic theory of target tracking algorithm. On the basis of research analysis onthe theory, add the sub-modules matching and the coordinate predicting into particlefilter and use the improved particle filter to track moving target. Experiment confirmsthat the method can effectively deal with the occlusions problems happening in theprocess of tracking.