A Robust Iris Recognition Approach Using Fuzzy Edge Processing Technique

Iris recognition is considered as the most secure biometric application. A robust iris recognition system involves many crucial steps in implementation. Edge estimation of iris biometric image is the initial critical step for any iris recognition technique for iris localization. Many algorithms have already been suggested for iris localization; but for better accuracy, robustness and freedom to extract the desired edges quickly, still requires novel iris recognition approaches. This paper presents a distinct approach for iris recognition by extraction of edges using fuzzy logic approach followed by iris localization using Circular Hough Transform (CHT). The proposed approach results into improved robustness of iris detection due to the efficient edge detection. It also improves localization due to ability of CHT to detect even partially visible iris. Computational speed is also improved using the simpler iris feature extraction and template matching. The proposed approach was experimented on MMU1, IITD, and UTIRIS iris biometric databases. Results show iris localization is achieved with greater accuracy at lower computational time.