A Liveness Detection Method For Palmprint Authentication

Many smartphones are equipped with a biometric authentication function to prevent their unauthorized use. An authentication method using palmprint as a physical feature has been proposed. Palmprints can be authenticated without special devices because the features of palmprints can be acquired using a smartphone camera. However, impersonation may be performed using user photographs. In this study, we examine a liveness detection method against a presentation attack in palmprint authentication using a smartphone. Assuming impersonation using familiar media, we consider attacks using printed and displayed images. In the attack image of the printed and displayed palms, we focus on the characteristics of resolution degradation caused by ink bleeding and the moire (interference fringe) caused when capturing the attack palm.