Enhanced Security in Biometrics: A Cancelable Multi-Instance Iris Authentication Utilizing Quotient Filter

Biometric-based authentication systems (BAS) can provide a strong security guarantee regarding the identity of users over traditional authentication systems.The iris of the eye is widely acknowledged as one of the most robust biometrics due to its exceptional performance.Despite this, templates used in traditional iris recognition systems remain unprotected, rendering them highly susceptible to various security and privacy breaches.However, several cancelable biometric schemes being introduced but at the expense of substantially decreased accuracy performance and increased computational time.To address this, we propose a cancelable multi-instance iris authentication system utilizing a quotient filter (CMAQF).The purpose of the quotient filter in CMAQF is to distort the biometric information without compromising the accuracy.Modified local random projection is applied on the fused iris template to generate the reduced template results in less authentication time.Experiments have been conducted on publicly available iris databases to assess the efficiency of CMAQF.The experimental results conclude that CMAQF achieves reasonable performance compared to existing methods, satisfying the properties of irreversibility, diversity, and revocability.