The Improvement Of Image Segmentation Based On GAC Model And C-V Model

Image segmentation plays an important role between image processing and image analysis. Image segmentation is the most difficult in image processing, but Image segmentation is a technique which is very attractive. Image segmentation based on partial differential equations is a new technology which is only active in the recent three decades. This technology brings a new feeling in the image segmentation, because this technology has rigorous mathematical theory. This technology has a vigorous vitality in the development process. Image segmentation methods based on partial differential equations, these methods are a primarily used theoretical idea which is presented in active contour model. Now, the main methods are based on the ideas in the evolution of evolved.First of all, this paper reviewed the existing image segmentation method, and image segmentation method based on partial differential equation, and Computational geometry knowledge, variational method and the variational level set method. Then active contour model is mainly studied. The main contents include: (1) Research several classic image segmentation methods which are based on partial differential equation; (2) Presents a model based on the GAC model. And based on the original model introduced a binding energy term, makes the improved algorithm can not only avoid reinitializes the sign distance function, thus improving the distance functions in on image segmentation initialized again when the sign distance functions, making the numerical calculation process more stable; (3) Presents a new model based on the C-V model. And based on the original model introduced a binding energy term, makes the improved algorithm can not only avoid reinitializes the sign distance function.