Near-Infrared Illumination Add-On for Mobile Hand-Vein Acquisition

Abstract

There is a growing need for mobile authentication solutions. Biometric recognition systems provide several advantages over conventional knowledge and token based solutions. Especially the use of vascular patterns as a biometric trait gains more and more attention. We present a near-infrared illumination add-on for smartphone devices which allows to capture the vascular pattern of the hands (hand-veins). This device is connected and controlled via Bluetooth and customised for the Nexus 5 smartphone but can be easily adopted to fit other models too. Due to the inherent risk of fraudulent authentication attempts on a nontrusted platform like a smartphone, we propose a challenge-response approach to ensure the authenticity of the captured hand-vein images. A hand-vein data set comprising of 31 subjects and 920 images in total is acquired with the presented device. A performance evaluation utilising different hand-vein recognition schemes is conducted to show the applicability of our device and the proposed challenge-response approach.