Iris Segmentation Using Geodesic Active Contours and GrabCut

Abstract

Iris segmentation is an important step in iris recognition as inaccurate segmentation often leads to faulty recognition. We propose an unsupervised, intensity based iris segmentation algorithm in this paper. The algorithm is fully automatic and can work for varied levels of occlusion, illumination and different shapes of the iris. A near central point inside the pupil is first detected using intensity based profiling of the eye image. Using that point as the center, we estimate the outer contour of the iris and the contour of the pupil using geodesic active contours, an iterative energy minimization algorithm based on the gradient of intensities. The iris region is then segmented out using both these estimations by applying an automatic version of GrabCut, an energy minimization algorithm from the graph cut family, representing the image as a Markov random field. The final result is refined using an ellipse-fitting algorithm based on the geometry of the GrabCut segmentation. To test our method, experiments were performed on 600 near infra-red eye images from the GFI database. The following features of the iris image are estimated: center and radius of the pupil and the iris. In order to evaluate the performance, we compare the features obtained by our method and the segmentation modules of three popular iris recognition systems with manual segmentation ground truth. The results show that the proposed method performs as good as, in many cases better, when compared with these systems.