眼底血管图像的特征提取和匹配

Feature Detection and Matching of Eyeground Vein's Image

针对传统的指纹识别中存在的缺点,本文初步研究眼底血管的特征提取与识别.利用眼底血管照相机提取眼底血管原始图像,对其二值化、滤波以及增强处理,进而进行特征提取与匹配的算法研究.通过在MATLAB仿真实验,拒真率和认假率均达到实验要求.利用眼底血管图像进行个人身份识别是完全可行的,并有良好的应用前景.

In order to resolve the defects of the fingerprint identification, the paper do a pilot study in the identification of eyeground vein. We picked up the images of eyeground vein, which then were normalized, filtrated and enhanced to be extracted and matched. By the emulational experiment in MATLAB, the truth rejection rate and the false acceptance rate arrived at the requirement. The results show that it is feasible to identify the human identity by using the eyeground vein＇s images.