Research On Algorithm Object Detection At Different Situation And Moving Object Tracking

With the development of the technology，the field of computer vision has obtained rapidadvances. People pay a lot of attention on object detection and object tracking. Its value isgradually increasing.The object detection at the motive-state background is applied by the backgroundsubtraction algorithm，the key to get a demanded result is a better background. in order to achievea better background，a adaptive background algorithm is proposed. Intro the third chapter，theexperiment result can prove the advantages of the proposed algorithm..Texture feature plays a vital role in target detection at the set-state background，intro thefourth chapter，a improved morphology object detection algorithm is proposed，meanwhile，we dosome research on removing noise and propose a new algorithm on removing noise. People canfind that the proposed algorithm is successful and satisfy the project’s demand according to theexperiment.Moving object tracking is a hot issue in the field of computer vision，its necessary to solvethe problem about complicated background，such as the background target ‘s color feature issimilar as the target，so is the texture feature. intro the sixth chapter，a feature fusion algorithm isproposed，we improve the classic mean shift algorithm，as a result of the experiment，it can beseen that the proposed algorithm can solve the problem and satisfy the demand.