Research On Image Segmentation Algorithms Based On Snake Model

Image segmentation plays an important role in the whole image processing. Research on the related issues has always been the hot issues at home and abroad, and a lot of mature methods have been existed. Among them active contour model(Snake) has been paid widely attention by researchers because it's combined with image's lower visual nature and object's priori knowledge. Firstly, in this paper the research status of segmentation algorithm based on active contour model is reviewed. Secondly, important improved methods such as Balloon Snake model, gradient vector flow(GVF) Snake and and so on are introduced, and then the research tendency of GVF Snake and its problems to be solved are given. On the basis, this thesis aims to research on image segmentation based on Snake model, and a better result is obtained. The main work is as follow:Aimed at the defects that the existing methods based on parametric active contour models(PACM) cannot accurately locate to corners, and discontinuous edges are easily affected by the surrounding irrelevant information, a new method for image segmentation based on PACM is proposed. In this method the edge preserving term is first constructed, which is introduced to the active contour model of image segmentation, and the tangent direction of Laplace diffusion term still persisted, and then two weight parameters are introduced to control tangential direction and normal direction so that the accuracy and efficiency for segmentation are improved. Experimental results showed that the proposed model can detect weak edges and locate accurately corners, meanwhile converges to the depth of concave boundary and reduce the impact of independent information on edge discontinuities. Furthermore, it overcomes the edge leakage and is very good for protecting image details. Both the efficiency and accuracy of segmentation are significantly improved in contrast with the existing similar models.In order to solve the bottleneck problem for segmenting the image with deep concaves, a segmentation method of the image with deep concaves based on GVF Snake is proposed. In this method, GVF Snake model is first used to segment initially the image and detect the curve that does not converge to object boundaries called the bottleneck curve. Then the direction of further convergence is determined by the geometrical feature of the bottleneck curve. Finally, the optimization area of the external force field is established based on the distribution characteristics of the external force field in the deep concave regions. The force field of the area is optimized and then the image segmentation is completed. Both theoretical analysis and experimental results show that without initializing curve entirely under interior or exterior of object boundary, the proposed method can not only guide curves toward concave regions, and adaptively converge into the concave and convex, but also conform to bend, long and narrow structure, avoid the gap of target boundaries as the bottleneck curve effectively, and prevent edge-leak and protect image details. Results of comparative experiments also show that the segmentation curve obtained by the proposed method is superior to that obtained by other similar algorithms in the fitting degree of depth structure.