基于灰度和梯度的视网膜下积液分割  

Sub-retinal Fluid Segmentation Based on Gray Level and Gradient

作　　者：时佳佳[1] 陈强[1] 

机构地区：[1]南京理工大学计算机科学与工程学院,南京210094

出　　处：《科技通报》2014年第1期51-54,共4页Bulletin of Science and Technology

基　　金：国家自然科学基金项目(编号:60805003)

摘　　要：视网膜下积液的容积计算可以被用来评价不同类型的视网膜和黄斑紊乱,因此准确找到下积液的轮廓也就十分必要。但是,目前关于视网膜下积液边界分割的研究很少。本文的目的是根据图像的灰度和梯度特征来分割出视网膜下积液,该方法利用图论分割算法找到经常出现下积液的区域边界,限定搜索区域后利用自适应阈值和梯度阈值来达到分割的目的。实验表明,该方法对于分割视网膜下积液效果良好。The volume of sub-retinal fluid can be used to evaluate different types of retinal and macular disorders, so it is necessary to get the boundary of sub-retinal fluid. However, there are few studies about the segmentation of sub-retinal fluid. The purpose of this paper is to segment sub-retinal fluid by features of gray level and gradient, this method gets the search region boundary where often appears fluid by graph-cut algorithm, after limiting the search region, we uses adaptive threshold and gradient threshold to obtain the boundary. Experiments show that our method has good performance on the segmentation of sub-retinal fluid.

关 键 词：视网膜下积液 分割 图论分割算法 灰度 梯度 

分 类 号：R339.146[医药卫生—人体生理学]