Increasing error tolerance in biometric systems
Proceedings of the 8th International Conference on Advances in Mobile Computing and Multimedia
Hi-index | 0.00 |
Accurate segmentation is a crucial phase in theimplementation of an iris recognition system. In thispaper we investigate a novel technique for irissegmentation. Morphological operations and areacomputation are applied together with other irissegmentation techniques in order to increase thespeed and accuracy of the preprocessing phase. Arough approximation of the pupil’s location is firstdetermined in the initial stage, followed by edgedetection and circular Hough transform for accurateiris segmentation. The edge image used to localizethe outer iris border is modified increasing the speedand accuracy of the process. Finally, we investigatethe effect of eyelids detection using a parabolic curvefitting technique. Two data sets of eye images areused to evaluate the proposed techniques.Experimental results show that the proposedsegmentation technique is efficient and performs wellon both data sets of images.