一种基于Gabor小波和2DPCA的掌纹识别改进算法  

AN IMPROVED PALMPRINT RECOGNITION METHOD USING GABOR WAVELETS AND 2DPCA

作　　者：苏滨[1] 姜威[1] 

机构地区：[1]山东大学信息科学与工程学院,山东济南250100

出　　处：《计算机应用与软件》2011年第1期242-245,共4页Computer Applications and Software

摘　　要：提出一种改进的基于Gabor小波变换和二维主分量分析2DPCA(2-Dimensional Principal component analysis)的掌纹识别。2DPCA克服了传统Gabor小波变换后直接进行主分量分析PCA(Principal component analysis)遇到的维数灾难问题,并且将PCA与Fisher线性判别FLD(Fisher Linear Discriminate)结合起来,利用了以前仅用于降维的PCA特征和FLD特征相融合进行掌纹识别。基于PolyU掌纹库的实验结果表明,该方法不仅有更高的识别率,而且维数更低。This paper presents an improved palmprint recognition method using Gabor wavelets transform and 2-dimentional principal component analysis （2DPCA）. 2DPCA in this method overcomes the curse of dimensionality encountered in traditional way of making principal component analysis （ PCA ） directly after Gabor wavelets transform, moreover, PCA is integrated with Fisher linear discrimination （ FLD ）, and the PCA features, of which only being used to reduce dimensions in the past, and the FLD feature are fused to carry out palmprint recognition. Simulation results based on PolyU Palmprint Database show that the proposed method can get higher recognition rate with lower dimensions.

关 键 词：掌纹识别 GABOR小波变换 二维主分量分析 主分量分析 FISHER线性判别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]