Extraction of complex object contour
Famao Ye, Lin Su, Shukai Li, Xiangjun Li · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2005
In this paper an approach based on particle filtering to extract complex object contour from imagery is presented. To deal with sharp tips of complex object, we take an adaptive state transition model with adaptive change in contour direction and adaptive number of particles. We not only use the information of luminance gradient but also use the speed and the direct of the move to compute the likelihood. Experimental results show that the proposed approach is capable of locating the target object contour of sharp tips accurately in the interactive way.