基于局部邻域四值模式的掌纹掌脉融合识别

为了解决掌纹掌脉识别技术中稳定性差和识别率低的问题，提出一种基于局部邻域四值模式的掌纹掌脉融合识别算法。对掌纹掌脉图像利用非下采样轮廓波变换(non-subsampled contourlet transform，NSCT)进行分解，将得到的低频和高频子图像分别利用区域能量和图像自相似原理进行融合；利用局部邻域四值模式(local neighbor quaternary pattern，LNQP)获取掌纹掌脉融合图像的纹理特征向量，并用主成分分析(principal component analysis,PCA)算法对其进行降维；根据特征向量间的汉明距离实现匹配识别，并在PolyU图库和SUT图库上完成仿真验证。实验结果表明，算法的最低等误率分别为0.17%和0.75%，与其他传统及最新算法相比，算法能够有效地提取掌纹掌脉图像的纹理特征，具有良好的识别性能，并且掌纹掌脉特征的融合增强了系统的安全性。

Palmprint and palm vein fusion recognition based on local neighbor quaternary pattern

To solve the problem of poor stability and low recognition rate in palmprint and palm vein recognition technology, we propose a fusion recognition algorithm based on local neighbor quaternary pattern. Firstly, the palmprint and palm vein images are decomposed by non-subsampled contourlet transform (NSCT), and the low-frequency with high-frequency sub-images are fused by the principle of region energy and image self-similarity respectively. Then, the texture feature vectors of palmprint and palm vein fusion image are obtained by local neighbor quaternary pattern (LNQP), and the principal component analysis (PCA) algorithm is used to reduce the dimension. Finally, the matching recognition is realized according to the Hamming distance between the feature vectors, and the simulation verification is completed on the PolyU and SUT image databases. The experimental results show that the minimum equal error rates of the article algorithm are 0.17% and 0.75% respectively. Compared with other traditional and state-of-the-art algorithms, the article algorithm can effectively extract palmprint and palm vein texture features, and has good recognition performance. In addition, the fusion of palmprint and palm vein features can enhance the security of the system.