| In the research and application of image, people are usually interested in certain parts of the image. Image segmentation is the technology and process that segmenting the image into many areas with different characteristics and extracting the goals interested in. These features can be gray, color, texture. Image segmentation is one of the most important kind of technology in image processing. It's also a kind of basic computer vision technology. For the advantage of integration in various models and the level set method of complex aspects of topological structure expression, variational level set method has become the basis of multiphase image segmentation method. This paper studied the level set model convex optimization problem based on piecewise constant multiphase image segmentation variational level set method. The work this paper studied includes the following aspects: Firstly, the application of level set method in image segmentation was systematically studied. Secondly, two-phase image segmentation convex optimization problem and Split Bregman algorithm were systematically researched. Thirdly, based on the competition strategy of n level set functions divided n regional, convex optimization treatment of multiphase image segmentation variational level set model on plane was designed, and the corresponding Split Bregman algorithm of it was also designed. Fourthly, alternating convex optimization of multiphase image segmentation model and Split Bregman algorithm were realized by programming, including the multiphase image segmentation on plane and 3d image segmentation, and they were applied on the artificially images and real medical images. Experimental results and comparison show that the proposed model makes the ideal segmentation effect,and in the level set evolution speed it is superior to the traditional model, and. Finally, this paper summarized the problems and defects, and pointed out the future direction of the research work. |
PDF Full Text Request