摘要
活动轮廓模型已经成功应用于图像分割,它可以是基于边界的,也可以是基于区域的。在演化过程中,关键问题是如何使水平集函数逼近符号距离函数。CHAN-VESE(C-V)模型是基于Mumford-Shah分割模型和水平集的,它不依赖图像梯度而检测目标,但其距离保持能力较差。在对C-V模型研究的基础上,提出了一个解决其距离保持问题的办法。
Active contour model has been successfully applied in image segmentation. It can be edge-based or region-based. During the evolution, the key issue is to force the level set function to be close to a signed distance function. CHAN-VESE （C-V） model can detect objects whose boundaries are not necessarily defined by gradient. Based on Mumford-Shah functional for segmentation and level sets, it is weak in preserving distance. In this paper, we proposed a method to address the issue of the distance preserving based on a study of the C-V model.
出处
《淮海工学院学报（自然科学版）》
CAS
2009年第4期18-21,共4页
Journal of Huaihai Institute of Technology:Natural Sciences Edition
基金
江苏省高校自然科学基金资助项目(06KJB520005)
江苏省"六大人才高峰"项目(06-E-028)
关键词
活动轮廓
区域
图像分割
水平集
模型
active contour
region
image segmentation
level set
model
