Anti-spoofing study on palm biometric features

Abstract

Spoofing attacks severely threaten the security of biometric recognition systems. Two concerns are raised regarding palm biometrics. First, the anti-spoofing ability of the palm vein as a subcutaneous feature needs further exploration. In-depth research is in need. Second, the capability of the palmprint and palm vein in liveness detection has been underestimated. Extracting extra liveness information for spoofing detection is urgently needed. This study proposes a dual-wavelength synchronous acquisition system for palm biometrics. Static biometrics of palmprint and palm vein and dynamic biometrics of SpO2 and pulse are captured and extracted. The proposed system maintains the recognition ability of the palm biometric system and achieves high anti-spoofing ability without additional hardware requirements. Under the proposed scheme, most hand feature measurement systems can be upgraded to obtain liveness information with minimal adjustment. A three-layer anti-spoofing strategy is also proposed based on the dynamic features. Based on the analysis of the human skin, various artificial palmprints and palm veins are fabricated using a variety of materials. Experiments investigated the anti-spoofing capabilities of palm biometrics. The results show that the proposed system can accurately extract SpO2 and pulse from palmprint and palm vein images, and SpO2 and pulse can significantly increase the anti-spoofing capability of palm biometrics.