一种基于奇对称2D Log-Gabor的虹膜特征提取算法  

Iris feature extraction algorithm based on odd symmetry 2D Log-Gabor

作　　者：吕林涛[1] 杨涛[1] 

机构地区：[1]西安理工大学计算机科学与工程学院,西安710048

出　　处：《计算机应用》2009年第4期976-978,共3页journal of Computer Applications

基　　金：西安市2008年科学技术计划项目(CXY08017)

摘　　要：针对2D Log-Gabor滤波器提取虹膜特征存在部分非有效特征点等问题,提出一种基于奇对称2D Log-Gabor的虹膜特征提取算法。采用奇对称2D Log-Gabor滤波器对虹膜纹理的相位和幅度信息在不同频率和方向进行分析,运用特征融合技术消除冗余信息,较好地解决了2D Log-Gabor滤波器提取虹膜特征存在的问题。实验表明,该算法在抗噪声和抗干扰中具有强鲁棒性,有较高的理论和实用价值。When 2D Log-Gabor extracted iris features, part of them are non-effective. Aiming at solving this problem, a new algorithm using odd symmetry 2D Log-Gabor filter to extract iris feature was proposed. Based on the study of odd symmetry 2D Log-Gabor filter, this paper analyzed the phase position and amplitude of iris texture in different frequency and directions, and eliminated redundant feature by feature integration. This algorithm solves the defects of 2D Log-Gabor wavelet extract iris feature problem. The experimental results show that the odd symmetry 2D Log-Gabor filters have strong robustness with noise and interferences. This algorithm has high theoretical and practical value.

关 键 词：虹膜识别 特征提取 2D LOG-GABOR滤波器 2D Log-Gabor奇对称滤波器 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]