结合形状统计的水平集图像分割

Statistical shape based approach to image segmentation using level sets

给出了一种结合先验形状统计信息的Mumford-Shah模型的水平集实现方法。结合形状统计的水平集图像分割主要包括先验形状模型的构造和形状能量项的构造,针对这两个主要方面做了如下两点工作:(1)提出了一种简单可行的先验形状模型构造方法;(2)重新构造了形状能量项,它综合考虑了全局和局部形状信息,且不含形状姿态参量,使曲面演化稳定可靠。带标记线左心室核磁共振(MR)长轴图像的实验结果和合成图像的分割结果证明了该方法的有效性。

This paper proposes a level set implementation of the Mumford-Shah model integrating prior shape statistical knowledge.The statistical shape based approach to the image segmentation using level sets mainly consists of the constructions of the prior shape model and the shape energy term.Aiming at these two parts,two pieces of work are mainly done： （1）A simple and feasible construction method of the prior shape model is proposed,which is based on binary images; （2）A new construction method of shape energy term is presented,which considers the global and local shape information at the same time,and without introducing pose parameters makes evolving surface stable.Promising experimental results are demonstrated on left ventricle Magnetic Resonance（MR） images of long axis with tag lines and a synthetic image.