摘要
为了对存在异常值的图像构建低维线性子空间的描述,提出用鲁棒主元分析(RPCA)的新方法进行掌纹识别。运用图像下抽样方法降低掌纹空间的维数,在低维图像上应用RPCA提取低维的投影向量,然后将训练图像和待识别图像向投影向量上投影得到鲁棒主元特征,计算特征向量间的余弦距离进行掌纹匹配。运用PolyU掌纹图像库进行测试,结果表明,与主元分析(PCA)、独立元分析(ICA)和核主元分析(KPCA)相比,RPCA算法的识别率最高为99%,特征提取和匹配总时间0.032 s,满足了实时系统的要求。
In order to construct low-dimensional linear-subspace representations from the data containing outliers, a new palmprint recognition method based on Robust Principal Component Analysis （RPCA） is proposed. The image down-sample is firstly used to reduce palmprint space dimensionality. The RPCA is applied to extract the low projec- tion vectors. Then the training images and test images are projected onto the projection vectors to get the robust prin- cipal component feature vectors. Finally, the cosine distance between two feature vectors is calculated to match palmprint. The new algorithm is tested in PolyU plmprint database. The results show that compared with Principal Component Analysis （PCA）, Independent Component Analysis （ICA） and Kernel Principal Component Analysis （KPCA）, the recognition rate of the new RPCA algorithm is the highest up to 99%, and all the time for feature extraction and matching is 0.032 s, so it meets the real-time system specification.
出处
《计算机工程与应用》
CSCD
2012年第19期8-10,60,共4页
Computer Engineering and Applications
基金
国家自然科学基金(No.61174119)
关键词
掌纹识别
主元分析
独立元分析
核主元分析
鲁棒主元分析
palmprint recognition
Principal Component Analysis （PCA）
Independent Component Analysis （ICA）
Kernel Principal Component Analysis （ KPCA ）
Robust Principal Component Analysis （RPCA）
