Skip to Main Content
This paper describes a framework for the detection and segmentation of the body of human in image sequences captured from a single stationary camera. An effective connected components labeling algorithm called CCPL (connected chips linking) is proposed to generate blobs and then be compared with the classic CCL algorithm in. An estimation of blobs clustering based on mean-shift algorithm is implemented. Then blob-sets are classified into ldquohumanrdquo or ldquofragmentsrdquo. Our goal is to solve the fragment problem caused by a cluttered background, improve the elimination of interference, and generalize feature information more effectively. Experiments show that our method performs well.