A new unconstrained iris image analysis and segmentation method in biometrics

Abstract

Iris recognition remains the most viable and reliable biometric method for security purposes. With the increasing demands in public safety and security, iris identification is becoming a requisite that seeks high accuracy with a fast and reliable outcome. Thus an effective iris recognition method is one that should overcome the rigid constraints imposed during image acquisition, and offer near real time processing. This research introduces a robust and fast (near real time) iris segmentation approach towards less constrained iris recognition. The accuracy is estimated to be slightly over 98% when using 500 randomly selected images from the UBIRIS.v2 partial database.