Privacy Preserving Cancellable Template Generation for Crypto-Biometric Authentication System

In the recent era, the cancellable approach of template protection is a widely applicable and secure method in fingerprint authentication systems. It has been proven that biometric traits offer promising security and enhanced confidentiality. Integrating cryptographic techniques with biometric modality has presented a dual-layered security mechanism, significantly mitigating the risk of biometric data breaches. In this paper, we propose a biometric template preservation scheme that employs minutiae transformation and encryption of biometric features for secure biometric template generation. The proposed methodology not only ensures the protection of biometric data in storage but also facilitates secure data retrieval for authentication purposes. Integrating cryptographic algorithms with biometric minutiae features effectively addresses the limitations of cancellable biometric templates, which can be susceptible to inversion attacks and template reconstruction. Our main contribution includes: (a) Secure template generation by Mobius transformation using user keyset (p, q), (b) Securing the user template using permutation and encryption operation. The analysis of security and matching performance of the proposed preservation scheme has been evaluated against publicly available fingerprint database FVC2000 DB1, FVC2000 DB2, FVC2002 DB1, FVC2002 DB3, FVC2004 DB1 and FVC2004 DB3 containing 800 images of 100 persons, 8 images per person in each database using FMR, FAR, GAR, ROC Curve, and EER parameters. The comparative performance has shown that the proposed technique performed favourably against the existing state-of-the-art techniques and can be used feasibly.