Palm-print recognition based on image quality and texture features with neural network

Biometric is the science of validating the integrity of human’s based on their physiological or behavioural attributes. Different biometric traits like retina, fingerprint, ear, face, palm-prints are broadly utilized for person authentication and user access. Palm-print as a biometric have attracted much research attention in various security applications. This paper presents the use of the convolutional neural network (CNN) combined with Gabor filter that extract highly discriminative features. An image quality module is applied to get the good quality images. Gabor filter having various scales and orientations is employed to extract the texture information of ROIs. The use of texture descriptor with CNN strengthens the learning of texture information. Experiments are conducted on the CASIA and IIT-Delhi touchless palm-print databases. The method yields an accuracy of 98.69% and Equal Error Rate (EER) of 0.62% on CASIA database. The experimental result demonstrates the superiority of the proposed method over the current methods.