The richness and stability of the iris patterns make it a robust biometric attribute for recognition of individuals. Segmentation of the noisy and non-circular iris images is a challenging task now a day. In this paper, anew approach has been proposed to isolate iris from the unwanted portions of eye images. An evolutionary algorithm Improved Differential Search (IDS) based Otsu multilevel thresholding (OMT)has been introducedas a pre-segmentation process in the iris recognition framework. The resultant images are then segmented using Geodesic Active Contours (GAC) incorporated by a novel stopping function. The experimental results are validated by comparing the proposed method with the well existing methods.The proposed method has been tested on the databases that are available publicly such as CASIA v3 Interval, UBIRISv1, and MMU1.
Volume 11 | 02-Special Issue
Pages: 1010-1023