基于C-V模型的医学图像分割方法

摘要

This paper focuses on two different GAC models：Li Chunming model based on gradient information and C-V model based on region information.In light of the drawbacks and advantages of these two models,we introduce penalty-function to C-V mode,and propose a C-V model without re-initialization.A comparison between the two models shows that the proposed model incorporates the segmentation speed of Li Chunming model and the segmentation performance of C-V model.%重点阐述了两种几何活动轮廓模型,基于梯度信息的李纯明模型和基于区域信息的C-V模型,在分析了两种模型的优缺点后,将李纯明模型中的罚函数项引入到C-V模型中,提出了无需初始化的C-V模型。实际结果表明,改进后模型具有李纯明模型的分割速度和C-V模型的效果。