摘要
为了克服传统的梯度矢量流(GVF)模型对细长拓扑结构、噪声及弱边界敏感的缺陷,提出一种基于区域信息的各向异性GVF模型.首先由模糊C均值(FCM)聚类算法得到聚类信息并将其融入到GVF模型中,以降低弱边界和噪声的影响;然后利用图像结构信息改进GVF模型,使其具有各向异性,克服了细长拓扑结构的影响;最后把得到的各向异性GVF模型融入到Snake方程中引导曲线的演化,得到目标边界.实验结果表明,该模型具有较好的分割结果.
In order to overcome the shortcomings that the traditional gradient vector flow（GVF）model is sensitive to structures with slender topology,noise and weak borders,an anisotropic GVF model based on regional information is proposed in this paper.First,the clustering information calculated by the fuzzy C-means（FCM）is applied to the GVF model in order to reduce the impact of weak borders and noise.Second,the image structure information is used to improve the GVF model,which could make the GVF model anisotropic and reduce the impact of slender topology.Finally,the anisotropic GVF model is included into the Snake equation which could guide the curves evolution to get the target border.Experiments show that the new model has better performance in segmentation.
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2010年第11期1887-1891,共5页
Journal of Computer-Aided Design & Computer Graphics
基金
国家自然科学基金(60973157)
关键词
图像分割
梯度矢量流模型
模糊C均值模型
各向异性
image segmentation
gradient vector flow model
fuzzy C-means（FCM）model
anisotropic
