基于MATLAB视网膜视锥细胞的图像处理

背景：视网膜自适应光学成像系统所获取的视锥细胞图像具有灰度分布相对集中、亮点边缘比较模糊、存在伪轮廓及边缘相粘连的特点，寻找一种适合视锥细胞图像的处理算法来获取视网膜视锥细胞清晰轮廓成为今后工作的重点内容。目的：采用MATLAB图像处理工具箱对视锥细胞图像进行边缘提取，获取其清晰的边缘轮廓。方法：对30例正常受试者不同部位视锥细胞图像进行预处理、边缘提取和形态学处理，获取视锥细胞图像清晰的边缘轮廓；对处理后的图像进行细胞计数，并分析视锥细胞密度分布特性，进而来验证图像处理算法应用于视锥细胞分布特性研究的可行性。结果与结论：获取了清晰的视网膜视锥细胞图像轮廓；从结果来看，随着远离黄斑中心凹，细胞密度呈现出降低的趋势，从黄斑中心凹偏0.5°到1°范围内，视网膜视锥细胞密度下降最快。结果提示：实验所设计的图像处理算法在研究视网膜视锥细胞分布特性方面是可行的。

Processing of the retinal cone image based on the MATLAB

BACKGROUND: The cone cell for the image retina adaptive optical imaging system is characterized by the relatively concentrated gray distribution, vague window edge, the existing pseudo outline, and the adhesion of edge photograph. Looking for a suitable processing algorithm of the cone image to get the clear outline is the key content for the future work.OBJECTIVE:To perform edge extraction of the retina cone cell image and get clear edge profile of the cone cell based on the MATLAB.   METHODS:According to the retinal image of the different parts of 30 normal subjects, based on Mat lab image processing toolbox, we could preprocess the cone cell image, edge extraction and process morphological, and get the clear edge profile of the cone cell image. The cell numbers of processed image was counted, the cone distribution characteristics were analyzed, and then the feasibility of this study which designed image processing algorithm applied to the study of the distribution characteristics of cone cells were verified.RESULTS AND CONCLUSION:A clear outline of the retinal cones image was obtained successfully. From the results of cone cell’s density, as far away from the center fovea of macula, the cell density presented a trend of reduction; from the center fovea of macula partial range 0.5° to 1°, the retinal cones density decreased rapidly. The findings indicated that the algorithm of the image processing designed in this study can be used to study the retinal characteristics of the cones distribution.