欢迎访问中国科学院大学学报,今天是

中国科学院大学学报 ›› 2012, Vol. ›› Issue (6): 750-756.DOI: 10.7523/j.issn.2095-6134.2012.6.005

• 数学与物理学 • 上一篇    下一篇

分段常值多相图像分割的变分水平集方法

李忠伟, 倪明玖   

  1. 中国科学院研究生院物理学院, 北京 100049
  • 收稿日期:2011-03-30 修回日期:2011-10-18 发布日期:2012-11-15
  • 通讯作者: 李忠伟
  • 基金资助:
    国家自然科学基金(50936006)资助

Variational level set methods for multiphase image segmentation based on piecewise constant

LI Zhong-Wei, NI Ming-Jiu   

  1. College of Physics, Graduate University, Chinese Academy of Sciences, Beijing 100049, China
  • Received:2011-03-30 Revised:2011-10-18 Published:2012-11-15

摘要: 基于变分水平集方法,建立一种具有噪声去除能力的分段常值多相图像分割模型. 该模型能在完成噪声去除的同时完成图像分割,从而缩短针对噪声图像的分割时间;其次,基于十进制数和二进制数转换,构造区域划分的特征函数表达式,从而建立一个新的多相图像分割模型;最后把该多相图像分割模型应用于合成图像和医学图像的分割.实验结果表明,与Chan-Vese模型相比,该模型能更快地提取噪声图像中的目标.

关键词: 噪声去除, 多相图像分割, 水平集, 特征函数

Abstract: A multiphase image segmentation model with the noise removal ability is proposed based on the variational level set mehod. The new model can complete the segmentation and the denoising process simultaneously. Therefore it can reduce the execution time of image segmentation. On the other hand, we construct the characteristic function of the regional division based on the decimal-binary number conversion. Finally, the proposed model is applied to the segmentation of the synthetic images and medical images. The experimental results show that the proposed model can extract the targets from the noise image faster than Chan-Vese model.

Key words: denoising, multiphase image segmentation, level set, characteristic functions

中图分类号: