Do multispectral palmprint images be reliable for person identification?
Do multispectral palmprint images be reliable for person identification?
复制标题
DOI：
10.1007/s11042-013-1706-3
复制
发表时间：
2015-02
影响因子：
3.6
通讯作者：
A. Meraoumia;S. Chitroub;A. Bouridane
中科院分区：
文献类型：
--
作者：
A. Meraoumia;S. Chitroub;A. Bouridane
本文主要研究多光谱掌纹图像在身份识别中的应用，并回答了这样一个问题：多光谱掌纹图像在身份识别中是否可靠？两个生物识别系统，然后提出。在第一个系统中，每个光谱图像对齐，然后使用1D Log-Gabor滤波器用于特征提取。对特征进行编码，并使用汉明距离进行匹配。在决策前进行匹配评分级的融合。第二个系统是基于多分辨率分析的特征提取。光谱图像被分解成具有不同分解级别的频率子图像。提取的系数被用作特征。MGPDF用于建模的功能和Log-Likestry分数用于匹配。在决策之前使用匹配分数水平的融合。然后对这两种系统进行了比较研究。实验结果表明，使用理大多光谱数据库和结果表明，这两个系统是更有效的多光谱图像时，比他们的单光谱对应图像。
This paper is concerned with an investigation of multispectral palmprint images for improving person identification by replying to the question: can multispectral palmprint images be reliable for such purpose? Two biometric systems are then proposed. In the first system, each spectral image is aligned and then used for feature extraction using 1D Log-Gabor filter. The features are encoded and Hamming distance is used for matching. The fusion at matching score level is used before the decision making. The second system is based on multiresolution analysis for feature extraction. The spectral images are decomposed into frequency sub-images with different levels of decomposition. The extracted coefficients are used as features. TheMGPDFis used for modeling the features and Log-Likelihood scores are used for matching. Fusion at the matching score level is used before decision making. A comparative study between the two systems is then developed. The experimental results are demonstrated using the PolyU multispectral database and the results show that the two proposed systems are more effective when using multispectral images than their monospectral counterpart images.