Towards Mobile Palmprint Biometric System with the New Palmprint Database

In this paper, we address the challenge of palmprint-based human verification in the mobile scenario. We propose a novel way of palmprint acquisition that will be used in order to create a new palmprint benchmark dataset. The palmprints are acquired by the handheld devices using an application dedicated to the Android operating system. The application provides the graphical assistant that improves the hand positioning and ROI extraction. It enables acquiring samples of both hands, in unconstrained illumination conditions, with different backgrounds. After the acquisition, each sample is evaluated according to color and rejected if it does not meet the requirements. So far, there are not many mobile biometrics databases acquired in an unconstrained environment and meeting privacy and legal requirements.