Human body image segmentation based on wavelet analysis and active contour models

GVF Snake model is used widely in computer vision and image segmentation. However, there are problems in convergence processing to boundaries of human body image because of noise. This paper presents a new segmentation algorithm to human body image. First, rough edge is got by multi-scale algorithm of wavelet analysis, and then thinning method based on mathematical morphology is adopted to get edge map as foundation of GVF Snake model. This method solves the problem that the edge based on wavelets analysis is not consecutive. And it improves GVF Snake model´s anti-noise ability. Experiments indicate that the new algorithm can improve snake model´s ability to segment the complicated image.