International Journal of Engineering and Technology

As a reliable and robust biological characteristic, the vein pattern increases more and more the progress in biometric researches. Generally, it was shown that single biometric modality recognition is not able to meet high performances. In this paper, we propose a new multimodal biometric system based on fusion of both hand vein and finger vein modalities. For finger vein recognition, we employ the Monogenic Local Binary Pattern (MLBP), and for hand vein recognition an Improved Gaussian Matched Filter (IGMF). Experimental results confirm that the proposed multimodal biometric process achieves excellent recognition performance compared to unimodal biometric system. The Area Under Curve (AUC) of the proposed approach is very close to unity (0.98).