Literature Review on Segmentation in IRIS Recognition

Abstract

IRIS recognition is one of the most accurate identify verification system. A Biometric System identifies a human being based on the unique feature like finger prints, facial features, plam print etc. Generally algorithm is divided into four steps like localization, normalization, feature extraction and matching. The performance of IRIS recognition systems depends heavily on segmentation and normalization techniques. Segmentation involves various techniques in IRIS reorganization. No single method has been proposed to detect all four kinds of noises viz., eyelashes, eyelids, reflections, and pupil in a single algorithm. The Iris recognition is not widely accepted because of its high cost. The various segmentation methods are included in this paper.