摘要
掌纹识别是生物特征识别中的一种,由于其分辨率要求低、设备成本低、用户易接受等优点受到众多研究者的关注。同其他生物特征识别一样也包括几何特征和数学特征识别两类方法,在数学特征中PCA、ICA、FLD等特征都可以用于掌纹识别。PCA是一种基于二阶统计的最小均方误差意义上的最优维数据压缩技术,FLD是一种较为普遍的用于特征提取的线性分类方法。将PCA和FLD结合起来进行掌纹识别,在识别阶段进一步利用了以前仅用于降维的PCA特征和FLD特征相融合进行识别。实验结果证明该方法比未改进的FLD方法在识别率上得到明显的提高。
Palmprint recognition has received many researchers＇ attention because of its low resolution and cheap devices. Like other feature recognition,palmprint recognition includes two recognition methods, geometric features and algebraic features. PCA, ICA, FLD features belong to the second class, and they all can be used for palmprint recognition. PCA （principal component analysis） is the optimal dimension compression technique based on second-order information in the sense of mean-square error. FLD is one of the most popular linear classification tech- niques for feature extraction. An improved palmprint recognition method based on traditional PCA and FLD is proposed. In this method, PCA features are used again in recognition stage. The experimental result shows that the novel method is effective.
出处
《计算机应用与软件》
CSCD
北大核心
2008年第1期204-206,共3页
Computer Applications and Software
