Novel Algorithm to Classify Iris Image Based on Entropy by Using Neural Network
Ali Asghar Beheshti Shirazi, Leila Nasseri · 2008
A novel algorithm to classify iris images automatically is announced by this paper, which apply local entropy, and using neural network. First iris image is separated from other part of captured eye image then normalized. After that normalized iris images partition to sixteen boxes that size of per box is (32times32) eight blocks as upper group and remain eight blocks as lower group, then calculate entropy of each boxes, we have sixteen entropy of per image, after estimated entropy for each box, then normalized and uncorrelated them finally apply neural network to categorize. This classification method has been tested and evaluated on 716 iris images. Consequent of this method iris images classify to six categories with accuracy 93.58%.