Abstract:
This paper proposes a palmprint recognition algorithm by integrating Contourlet transform and Non-negative Matrix Factorization (NMF). Three steps are involved in the algorithm: a source palmprint image is convolved by the Contourlet transform; NMF is used to extract features from the low frequency components, and the nearest neighbor classifier is used for classification. Experimental results show that the recognition rate is significantly improved than using NMF, 2DPCA and other recognition methods.
Key words:
Contourlet transform,
Non-negative Matrix Factorization(NMF),
palmprint recognition,
fusion,
feature extraction,
2DPCA algorithm
摘要： 提出一种基于Contourlet变换和非负矩阵分解(NMF)的掌纹识别算法。通过对源图像Contourlet进行小波变换，将提取出的低频分量用NMF法提取特征值，用最近邻方法进行分类。实验结果表明，该算法较单纯的NMF和2DPCA等算法识别性能有较大提高，能较好地捕捉图像的边缘信息。
关键词:
Contourlet变换,
非负矩阵分解,
掌纹识别,
融合,
特征提取,
2DPCA算法
CLC Number:
LIU Xiang, LI Yan-Hua, BO Xin, CHE Hua-Qiong, SU Jing. Palmprint Recognition Algorithm Based on Contourlet Transform and NMF[J]. Computer Engineering, 2012, 38(13): 175-177.
刘洋, 李燕华, 潘新, 多化琼, 苏静. 基于Contourlet变换和NMF的掌纹识别算法[J]. 计算机工程, 2012, 38(13): 175-177.