A Multi-feature Joint Distribution Camshift Target Tracking Algorithm
As the Camshift algorithm has the problem that the tracking object is lost and low robustness in the complex back-grounds such as texture and color similarity interference,target rotation,etc.A multi-feature joint distribution Camshift target track-ing algorithm is proposed.The new algorithm chooses uniform LBP texture,hue and saturation as multi-feature and converts RGB space in the picture to ULBP-H-S space.The moving target in the picture is selected as target template,the ULBP-H joint probabil-ity distribution and H-S joint probability of the target template are calculated,and the target joint probability distribution is calculat-ed by bit-and-operation of two joint probability distribution through adaptive coefficients.In each iterative search process,the adap-tive search window algorithm is used to predict the location and size of search window in the next frame,and the Camshift algorithm is used to continuously track the target in the predicted search window.The experiment results show that the improved algorithm has higher accuracy and robustness in tracking moving target under the complex environment of texture and color similar interference and target occlusion.
国家科技期刊平台
NSTL
万方数据
