摘要
形变模型技术作为一种既能结合底层图像信息，又可结合高层先验知识的图像分割技术，以其独特的魅力吸引着该领域越来越多的研究者，但至今尚未有统一的分类方法。提出一种新的分类方法，把形变模型技术大致分为自由形态和受限形态两类模型，从新的视角对形变模型技术进行了系统分类和知识梳理，回顾了形变模型技术的主要工作及其在医学图像分割领域中的应用，为研究者对这一领域的总体把握提供了有益的参考。
Deformable models used in segmentation techniques, with its combination of low-level image information and high-level existent knowledge, attract more and more attentions of researchers in this area. There has not been a consistent taxonomy proposed until now. In this article a new taxonomy of these techniques was proposed： Deformable models are classified as free-form and constrained-form models. Systematical classifying and trimming of the knowledge were done from a new angle. Major works and their application in medical image segmentation area are reviewed. Helpful references are hopefully to be provided to researchers in this area.
出处
《国际生物医学工程杂志》
CAS
2007年第4期222-226,共5页
International Journal of Biomedical Engineering
关键词
图像分割
形变模型
主动轮廓模型
形变模板
image segmentation
deformable models
active contour models
deformable templates
