A Secure Template Protection Technique for Robust Biometric Systems

Biometric features are advantageous for human authentication systems over traditional methods. Fingerprints are a widespread biometric feature for user verification. Fingerprint-based systems usually rely on minutiae points as user templates. However, using minutiae points as user templates may not be reliable, as it may enable the original fingerprint to be recovered. Therefore, protecting fingerprint information from database attacks is crucial. This paper presents a novel technique for generating secure fingerprint templates using a polar coordinate system. The polar coordinates of fingerprint minutiae points are transformed by the Chebyshev polynomial and chaotic map function to generate secure cancelable templates. The proposed method is robust and has high revocability, diversity, security, and performance. The proposed approach is tested using the FVC2002 DB1, DB2 and DB3 fingerprint databases. In comparison to existing methods, the proposed approach demonstrates notable performance improvement.