The Study of Motion Target Detection Algorithm Based on LBP Texture Feature
Gu Wei;Xuzhou Higher Normal School;
In the process of moving target detection,if the light of two images changes,the test results will be affected.In order to solve this problem,this paper proposes a new texture feature based on local binary pattern background modeling method,to calculate to background frames and LBP texture features of the current frame diagram,and the background frame and current frame LBP texture characteristics graph corresponding to pixels of different prospects or generate probability graph,probability graph describes the prospect of the current pixel is the prospect of probability,according to the probability of future figure adaptive update background,can make the background quickly close to the real background,target is obtained using background subtraction division again. The experiment shows that the proposed algorithm can effectively deal with the sudden change of illumination,the background is updated quickly,and the detection target is close to the real target.
【CateGory Index】： TP391.41