3D Fast Level Set Image Segmentation Based on Chan-Vese Model
Jianyuan Dong, Chongyang Hao · 2009
In this paper, we propose 3D fast level set segmentation base on Chan-Vese Model. Define the signed distance as integers without computation, and creative construct the evolution speed field of surface by the partial differential equation for Chan-Vese model (only the sign is velocity is used), we implement 3D fast level set segmentation without solving partial differential equation. It preserve the advantages of Chan-Vese model that detect objects whose boundaries are not necessarily defined by gradient or without edges. Finally we present some segmentation experiment for 2D and 3D medical image.