Statistical shape influence in geodesic active contours：（统计形状影响测地线主动轮廓）.pdfVIP

Statistical Shape Influence in Geodesic Active Contours Michael E. Leventon1 W. Eric L. Grimson1 Olivier Faugeras 1;2 1 MIT AI Lab 2 INRIA Cambridge, MA 02139 Sophia Antipolis, France leventon, welg@ai.mit.edu Olivier.Faugeras@sophia.inria.fr Abstract u = 1 A novel method of incorporating shape information into u = 0 the image segmentation process is presented. We introduce C a representation for deformable shapes and define a prob- u = -1 ability distribution over the variances of a set of training shapes. The segmentation process embeds an initial curve as the zero level set of a higher dimensional surface, and evolves the surface such that the zero level set converges on Figure 1. Level sets of an embedding function u for a the boundary of the object to be segmented. At each step closed curve 2 C in . of the surface evolution, we estimate the maximum a poste- riori (MAP) position and shape of the object in the image, based on the prior shape information and the image infor- Boundary finding segmentation methods such as Snakes mation. We then evolve the surface globally, towards the [8],