An Approach to Retinal Image Segmentations Using Fuzzy Clustering in Combination with Morphological Filters
Retinal vessel appearance is an important feature for personal identification, information security and confidentiality. It also is a key indicator for many early diagnoses, such as diabetes and hypertension. In this approach, a extraction method of retinal vessels based on fuzzy clustering in combination with morphological filtering is proposed. By decomposing the green channel image into smooth and textured components, fuzzy clustering is firstly performed on the textured composite, then the morphological open operation with multiscale linear-like structure elements is applied to suppressing noise structure. Experimental results indicate that the method can automatically and effectively extract most of the vessel backbones and branches.
blood vessel segmentation image decomposition fuzzy clustering morphology filtering
Liang Ding Yongping Zhang Xueying Zhang
College of Information Engineering Taiyuan University of Technology Taiyuan,Shanxi,China School of E School of Electronic and Information Engineering,Ningbo University of Technology,Ningbo,China College of Information Engineering Taiyuan University of Technology Taiyuan,Shanxi,China
国际会议
太原
英文
411-414
2011-02-26（万方平台首次上网日期，不代表论文的发表时间）