Moving Objects Detecting and Tracking for Fighting Shadow and Occlusion
Jianyi Li · Infrared Technology · 2011
Moving shadow interference and obstacles occlusion problem is very universal in the actual scene,and it seriously influenced the effect of object detecting and tracking.To solve the problems,a detecting and tracking method based on shadow detecting and occlusion environment is presented.Firstly,the classification floating refresh strategy is performed to lighten operation burden.Then,a shadow detecting module based on texture gradient difference and color feature invariant is introduced to more precise distinguish foreground.While tracking movind objects,the method based on meanshift combines with maximum posterior probability is used.Meanwhile,the mechanism of obstacles occlusion tracking is introduced to reduce the possibility of moving objects missing in the occlusion environment.Experiments show,the proposed algorithm can effectively detect the shadow of moving objects,and it also can precisely track the objects occluded.