ABSTRACT
This chapter introduces an efficient multiresolution level set image segmentation method with multiple regions. It describes statistical formulation for image segmentation and two-region level set models, including representation of a partition, multiregion segmentation functional, curve evolution equations, level set implement, and description of our proposed method. The chapter shows a representative sample of the tests with a real medical image. Chan and Vese proposed an active contour approach to the Mumford–Shah problem for a two-phase case, which is based on a piecewise constant function. The resulting segmentation is a fixed but arbitrary number of regions, which is guaranteed to be a partition of the image domain. The experimental results show that the final segmentation remains a partition of the image domain and that the method is efficient. The resulting segmentation is a number of regions, which is guaranteed to be a partition of the image domain.
