SIFTBCS: scale invariant feature transform based fuzzy vault scheme in biometric cryptosystem

Biometric cryptosystem (BCS) is a challenging field of study where data is secured using biometrics features and cryptographic encryption. The necessity for retaining login credentials has been eliminated by the incorporation of biometrics in encrypted technology. The integration of biometrics into cryptographic systems eliminates the necessity of recalling passwords. This manuscript puts forward an innovative approach referred to as Scale Invariant Feature Transform (SIFT) based Biometric Cryptosystem (SIFTBCS) for the construction of cryptographic keys from the obtained features. The foundation of this technique rests upon the established Fuzzy Vault scheme utilized for safeguarding data. Themotive behind employing the SIFT scheme in biometric cryptosystems is that the extracted features remain unaltered by changes in orientation and scaling, thereby enhancing the accuracy and reliability of authentication. SIFTBCS method is comprised of three distinct stages, namely Key generation, Enrollment and Authentication. Comprehensive experiments were performed on five different ear databases, namely Mathematical Analysis of Images, Carreira-Perpinan, Indian Institute of Technology Delhi (version 1 & 2) and University of Science & Technology Beijing (version 2). Statistical and non-numeric data are used to draw meaningful conclusions from the study. The proposed technique achieves an accuracy of approximate to 94% on USTB-v2, IITD-v1 and CP databases. Our approach demonstrated superior performance using both kinds of performance evaluations.