基于Gabor滤波与数学形态学的视网膜图像增强方法  

Enhancement method for retinal images based on Gabor filter and morphology

作　　者：刘尚平[1] 陈骥[1] 

机构地区：[1]重庆大学生物工程学院,重庆400044

出　　处：《光电子．激光》2010年第2期318-322,共5页Journal of Optoelectronics·Laser

基　　金：重庆市"十一五"科技重大专项资助项目(CSTC2008AA5010)

摘　　要：提出一种双尺度多方向的Gabor滤波器与数学形态学结合的视网膜图像增强方法。首先使用双尺度Ga-bor滤波器提取各个方向的主要血管信息并弱化背景,然后使用数学形态学top-hat变换来增强图像微小的细节,最后结合两种方法实现增强的动态调节。实验表明,该方法能够很好地增强视网膜图像中的血管信息,并大范围地弱化了背景和噪声,对于改善视网膜图像的视觉效果有极大帮助。A new retinal enhancement method based on two-scale Gabor filter and morphology is proposed.Firstly,the two-scale Gabor filter is used to extract the major blood vessels and weaken the background.And then the mathematical morphology top-hat transform is used to enhance the image details.Finally,the two methods are integrated to realize enhanced dynamic regulation.The experiments demonstrate that this method can enhance the blood vessels of the retinal images well and weaken the background and noise.which can improve the visual effects of the retinal images greatly.

关 键 词：图像增强 GABOR滤波 数学形态学 视网膜 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]