The Recognition And Tracking Of Moving Targets

With the development of computer and video monitoring technology unceasingdevelopment, the motion object tracking and recognition has become an importantresearch. This paper which is based on this resech, made a series of research work in thethree of moving object background extraction, moving object recognition and motionobject tracking.In the first, this paper made a series of image preprocessing, including images of gray,inmage enhancement, image flilter and noise removal, and image mathematic morphologyetc. In addition, this paper has made simple introduction in the video image sequence andimage color space knowledge.In the second, this paper made a large number of meticulous research works on thebackground extraction of the video image. This paper made a variety experiments forevery background extraction method and made a comparative analysis for the experimentresults. Finaly we selected the improved averaging method as the final method. Then thispaper analyzed the various motion objects detection algorithm, such as backgrounddifference method, symmetric difference and optical flow field method.Finally, this paper has put forward two kinds of moving target tracking algorithm bas,object tracking algorithm based on region and Kalman filtering object tracking algorithm,and through experimental analysis on the algorithm’s reliability and feasibility, the effectis proved good.This paper has made some improvement on background extraction and update, andthe analysis of experiment shows that the improved algorithm is better than the originalalgorithm. This paper also has done a large number of experimental analyses and has putforward some solving ways on the moving targets detection and tracking. The resultsshow that these methods are feasible and have good application value.