基于相位一致和核模糊C均值的眼底血管分割  

Vasculature segmentation of fundus image based on phase congruency and kernel fussy C-mean algorithm

作　　者：谭春林[1] 曹鹏[1] 黄华[1] 

机构地区：[1]四川大学电气信息学院,四川成都610065

出　　处：《计算机工程与设计》2015年第6期1551-1554,1570,共5页Computer Engineering and Design

摘　　要：为提取反映心脑血管疾病的眼底视网膜血管,提出一种结合相位一致性和核模糊C均值的分割算法。利用基于Log Gabor小波的相位一致性算法在频域做血管边缘检测,利用核模糊C均值聚类检测结果,去噪声并二值化得到血管图像。基于STARE数据库的实验结果表明,以专家手工分割结果为参考标准,使用该算法的平均分割正确率高达92.65%。To extract retinal vessels which demonstrate the condition of cardiovascular and cerebrovascular diseases efficiently ,an algorithm was proposed based on phase congruency （PC） and kernel fussy C‐mean （KFCM） .Various vessel edges were detected and localized utilizing PC based on Log Gabor wavelet in frequency domain .The KFCM was applied to PC’s result to cluster ves‐sel pixels .The binary retinal vessel image was denoised and got .Results of experiments based on STARE database show refer‐ring to results gained by specialist ,the average accuracy of the proposed method reaches 92.65% .

关 键 词：眼底图像 聚类算法 视网膜血管分割 相位一致性 核模糊C均值 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]