An iris recognition based robust intrusion detection system

Iris Recognition technology the most sought after biometric technology these days uses statistically independent patterns generated from human iris that are unique to every human being, reliable and has greater user acceptance. A robust and foolproof iris recognition system is expected to have zero FAR (False Acceptance Rate) and FRR (False Rejection Rate). In this work the impact of number of outsiders on the performance of the proposed system has been investigated. Then the system has been made robust by tuning threshold to bring down FAR and combination of two feature extraction techniques is proposed i.e. Gabor and Haar to improve FRR thus increasing recognition accuracy of the system.