A Dynamic Iris Authentication System with One-Time Encoding

Abstract

Biometric features are extensively utilized in real-time applications for unique human identification. Among these, iris recognition stands out as a highly reliable physiological biometric modality. However, traditional iris recognition systems are prone to security vulnerabilities due to their reliance on static encoding techniques. This paper proposes a dynamic iris authentication system that generates a unique One-Time Encoded Iris Code (EIC) per session by combining Gabor filters for texture enhancement and Haar wavelets for compact feature representation. The proposed (EIC) uses matrix rotation, XOR operations, random numbers, and timestamp embedding to ensure security. Formal evaluations, including FAR vs. FRR and Real-or-Random analysis, confirm the system’s resilience, with results showing 7% FAR and 11% FRR. Performance metrics indicate an average processing time of 0.067 seconds and memory usage of 2309.36 KB, demonstrating superior efficiency compared to existing schemes. This work contributes a scalable and secure solution for modern authentication needs. The proposed system offers a secure, scalable solution for applications requiring reliable user authentication, such as smart home environments and secure access systems.