A Fast Algorithm for Image Segmentation Based on Local Chan Vese Model

Image segmentation plays a very important pole in image processing and computer vision field. Most of the energy minimization of level set methods are based on the steepest descent method and finite difference scheme. In this paper, we propose a sweeping algorithm to minimize Local Chan Vese (LCV) model. We calculate the energy change when a pixel is moved from the outside region to the inside region of evolving curves and vice versa, instead of directly solving the Euler-Lagrange equation. The algorithm is fast and robust to initial level set contour and can avoid solving partial differential equation. There is no need for the re-initialization step, any stability conditions and the distance regularization term. The experiments have shown the effectiveness of the proposed algorithm.