Based On The Dual Method For Multiphase Image Segmentation

Image segmentation that is a key technology in Computer vision has been widely studied and applied in digital image processing.Active contour model is a very successful variation model based on the level set method in Image segmentation, which is driven under the energy functional to make the contour lines gradual deformation until the contour lines arrival at the boundaries of the segmentation Region, but its fatal drawback is difficult to deal with variable Buddhists and only has a local optimal solution, the problem can be solved through the Global Minimization of the Active Contour Model.Traditional level set methods use a signed distance function as the level set function, its difference scheme is very cumbersome.In the time of carrying out numerical calculation usually need to add a small variable in order to prevent the denominator of zero, which will inevitably have an impact on the results of segmentation. In order to solve the above problems, the introduction of dual methods to the active contour model,dual method that introduction of a variable to become dual variable by means of known variable, systematically establish the type of model calculation of the dual method, it is not only greatly improves the computing speed, but also to overcome the total variation of the traditional model of calculation of parameters of the introduction of the rules to bring a small decline in the precision. Multiphase image segmentation based on Dual Method are introduced in this paper as follows:Firstly, the basic theory on curve evolution, dual method and Active contour model are systematically researched and their application in image segmentation field are discussed; Secondly, Global Minimization of the Active Contour Model is realized based on the ROF model and the M-S model,including the two-dimensional images segmentation of the piecewise constant case and the piecewise smooth case; Furthermore, all the models mentioned above are implemented and applied to artificial images and real medical images, Many experiments denote that we get the satisfying results using the corresponding image segmentation models.Lastly, future research directions are put forward according to the problems and limitation occurred during my research work.