A hybrid algorithm for detecting contour of moving object based on merging Mean Shift and GVF Snake model
Gu-quan Li, Zhongze Chen · 2011
In this paper, a new algorithm for extracting contour of moving objects through video sequences based on merging Mean Shift Algorithm and GVF Snake model is proposed. Firstly, object region (i.e. image region a moving object covers) is determined, thus actual contour searching activity is restricted to a small area; and then an initial position, which is usually within the extracted object region, of a Snake curve is given; Finally, the contour of a moving object is obtained by using a GVF Snake model. Experimental results show that the number of iterations as well as computation complexity for extracting contour are greatly reduced than that of by using the GVF Snake model alone, and also that it holds the advantage of extracting actual contour of a moving object effectively.