Palm vein technology for biometrics

Abstract

The use and implementation of biometrics for identification and authentication has become more important in the past decade. This is because there has been an increased risk associated with textual passwords such as dictionary attacks, eavesdropping, shoulder surfing etc. We have worked on Palm vein recognition for detecting palm veins in applications for biometric security using near infrared absorption phenomena. The goal is to produce a software prototype that is capable of identifying a person by the vein structures of the hand. The images used for the same were taken from the CASIA-MS-PalmprintV1 database collected by the Chinese Academy of Sciences' Institute of Automation (CASIA). After pre-processing, LDR and DCT have been used for feature extraction and Euclidian Distance is calculated for generating matching score. Acceptance/rejection is based on this matching score. The efficiency obtained was 93.2% when compared with other systems.