Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (31): 209-211.DOI: 10.3778/j.issn.1002-8331.2010.31.058

• 图形、图像、模式识别 • Previous Articles     Next Articles

Approach of novel level set for medical image segmentation based on local region information

ZHENG Wei,CHEN Yan-jiang   

  1. College of Electronic and Information Engineering,Hebei University,Baoding,Hebei 071002,China
  • Received:2009-03-24 Revised:2009-05-26 Online:2010-11-01 Published:2010-11-01
  • Contact: ZHENG Wei

基于局部区域信息的水平集医学图像分割方法

郑 伟,陈彦江   

  1. 河北大学 电子信息工程学院,河北 保定 071002
  • 通讯作者: 郑 伟

Abstract: Because medical image has the characteristic of intensity inhomogeneity,this paper presents a novel level set based on the simplified Mumford-Shah model for image segmentation proposed by Chan-Vese.Local region information is crucial for accurate segmentation of images with intensity inhomogeneity,however,the traditional level set method don’t utilize it.This paper proposes a novel level set medical image segmentation base on local region information.This mothed costs less,and the experimental result verifies the effectives and robustness of this segmentation method.

Key words: image segmentation, level set method, C-V model, local region information

摘要: 针对医学图像中存在的亮度分布不均匀(intensity inhomogeneity)的特点,对Chan-Vese提出的基于Mumford-Shah模型的水平集分割图像的算法进行了改进。局部区域信息是对亮度分布不均匀图像进行准确分割的关键,但是传统的基于区域信息的C-V模型没有利用到这种局部区域的图像信息,因此无法正确分割强度分布不均匀图像。利用局部区域信息构造能量函数,提出了一种基于局部区域信息的改进C-V模型。该模型无需大量计算,水平集函数可快速收敛。MR图像、血管造影图像和X线骨折图像的实验结果证明了该方法的高效性。

关键词: 图像分割, 水平集方法, C-V模型, 局部区域信息

CLC Number: