Generation of Secure Fingerprint Template using DFT for Consumer Electronics Devices

Abstract

Biometric authentication systems are being extensively used to secure various applications. Consumer electronics devices also utilize biometrics as one of the authentication methods. However, this growth in the use of biometric based authentication may lead to the permanent loss of the biometric identity of a person if it is compromised, as biometric data cannot be changed. Hence, this poses the requirement of securing user templates in biometric based systems. In this paper, we propose a secure, alignment-free, and non-invertible fingerprint based user template protection technique which utilizes the mapping of pair-polar structures of minutiae in a 3D-grid and Discrete Fourier Transform (DFT). The proposed technique has been analyzed in terms of revocability, diversity, security, and performance by using four publicly available fingerprint databases of Fingerprint Verification Competition (FVC) 2002 and 2004. In addition to these databases, a partial fingerprint database generated from the FVC2002 DB1 is also utilized. The Equal Error Rate (EER) values of the proposed technique for all the databases are compared with that of the state-of-the-art techniques, and the superiority of the results depicts the robustness and effectiveness of the proposed technique. IEEE