Geometric Deformable Model In Medical Image Segmentation In Color

Medical image segmentation is the essential step of medical image processing, and it plays a crucial role in both qualitative and quantitative medical image analyses. Because different medical images have different image characteristic, different image segmentation algorithms are needed. This paper aims to do some research on color medical image segmentation algorithms.After the summarization of the characters of medical image segmentation and the progress of medical image segmentation research, we discuss the problem and difficulty in this field. Then the paper analyzes the existing image segmentation techniques synthetically and list advantages and shortage of these techniques by comparing. Deformable models have many advantages over other classical segmentation algorithm. It has been applied to medical image segmentation.There are basically two types of deformable models: parametric deformable model and geometric deformable model. Researching and comparing on this two models, we find the latter is fit for the medical image segmentation. Because of the geometric deformable model, which is based on curve evolution theory and level set method, can treat with the change of topology naturally and void the parameterization of evolving curve.In order to use deformable model in image segmentation, we must select a segmentation model to form the energy field attracting the evolving curve to the boundary of ROI (Region of Interest). This paper introduces the theory of geodesic model, and then introduces level set method-a method of solving geometric deformable model. We discuss the fast algorithm of level set method: fast marching method and narrow band method and give the kernels of level set procedure in detail.Chan-Vese model is more and more popular in recent years. The scalar C-V model overcome the shortage of the classical segmentation model by using the global information of image to make curve stop at the edge of ROI, and can detect objects with very smooth boundary or even with discontinuous boundaries. The scalar C-V model can be extended to the vector case naturally, so it can segment color image.In this paper, we altered the speed function of vector C-V model and make model detect the edge of ROI faraway from origin curve by eliminating the suppression of Dirac function to none-zero level set. Besides, in order to further stabilize and fasten the level set evolution procedures, the paper addresses an improved Fast Sweeping Method to construction of the signed distance function(SDF). Experiments on real-world color medical images show the robustness and good performance of the method.