基于自适应Gabor滤波与支持向量机的虹膜识别方法  

Iris recognition algorithm based on adaptive Gabor filtering and SVM

作　　者：马晓峰 高玮玮 MA Xiaofeng;GAO Weiwei(School of Mechanical and Automotive Engineering,Shanghai University of Engineering Science,Shanghai 201620,China)

机构地区：[1]上海工程技术大学机械与汽车工程学院,上海201620

出　　处：《智能计算机与应用》2020年第6期212-217,共6页Intelligent Computer and Applications

基　　金：海高校青年教师培养资助计划(ZZGCD15081);上海工程技术大学校科研启动项目(E1-0501-15-0185)。

摘　　要：本文针对传统Gabor滤波器只能手动修改滤波参数,通用性低的缺点,提出了基于自适应Gabor滤波与支持向量机的虹膜识别算法。首先,结合虹膜图像的灰度特征完成虹膜的定位与归一化;其次,利用Gabor滤波器提取虹膜特征,并通过粒子群算法寻找最优参数,根据最优参数提取最优虹膜特征;最后,通过支持向量机进行虹膜识别,同时利用该方法对CASIA V1和Lamp虹膜库进行识别,得到系统识别率分别为99.23%和99.11%。与传统的虹膜识别方法相比,基于自适应Gabor滤波与支持向量机的虹膜识别能对不同的虹膜库自动优化参数,克服了传统方法中的人工调参问题,且能显著提高系统的识别性能,具有更强的实用性。This paper proposes an iris recognition algorithm based on adaptive Gabor filter and support vector machine,aiming at the disadvantage that the traditional Gabor filter can only modify the filtering parameters manually and has low universality.Firstly,iris localization and normalization are accomplished by combining with the gray feature of iris image.Secondly,the iris features were extracted with Gabor filter,and the optimal parameters were searched by particle swarm optimization algorithm,and the optimal iris features were extracted according to the optimal parameters.Finally,iris recognition was carried out by support vector machine.Meanwhile,CASIA V1 and Lamp iris library were identified by this method,and the system recognition rates were 99.23%and 99.11%,respectively.Compared with the traditional iris recognition method,iris recognition based on adaptive Gabor filtering and support vector machine can automatically optimize the parameters of different iris libraries,overcome the problem of manual parameter adjustment in the traditional method,and can significantly improve the recognition performance of the system,with stronger practicability.

关 键 词：虹膜识别 GABOR滤波 特征提取 粒子群算法 支持向量机 

分 类 号：R318[医药卫生—生物医学工程]