Palm vein authentication and verification system

Abstract

In this work, both hardware and software of a palm vein capture system have been designed and a prototype is produced. By using this system, a genuine database was created with age and gender information of volunteered test subjects. Widely used pre-processing, acquisition of feature vectors, and classification methods were tested on the database and results are compared with speed versus accuracy. All the modules used in hardware and software were determined as open source. Python has been chosen as a software module programming on a single board computer with Linux kernel. As a result, an appropriately running system has been built with both secure software and hardware interfaces having maximum accuracy at 95.33% while specificity is at 100%.