Segmenting Non-Ideal Irises Using Geodesic Active Contours

Abstract

The richness and the apparent stability of the iris texture makes it a robust biometric trait for personal authentication. The performance of an automated iris recognition system is affected by the accuracy of the segmentation process used to isolate the iris structure from the other components in its vicinity, viz., the sclera, pupil, eyelids and eyelashes. Most segmentation models in the literature assume that the pupillary, the limbic and the eyelid boundaries are circular or elliptical in shape. Hence, they focus on determining model parameters that best fit these hypotheses. In this paper, we describe a novel iris segmentation scheme that employs Geodesic Active Contours to extract the iris from the surrounding structures. The proposed scheme elicits the iris texture in an iterative fashion depending upon both the local and global conditions in the image. The performance of an iris recognition system based on multiple Gabor filters is observed to improve upon application of the proposed segmentation algorithm. Experimental results on the WVU and CASIA v1.0 iris databases indicate the efficacy of the proposed technique.