Iris matching by means of Machine Learning paradigms: A new approach to dissimilarity computation

This paper presents a novel approach for iris dissimilarity computation based on Computer Vision and Machine Learning. First, iris images are processed using well-known image processing algorithms. Pixels of the output image are considered the input of the previously trained classifiers, obtaining the a posteriori probability for each of the considered class values. The main novelty of the presented work remains in the computation of the dissimilarity value of two iris images as the distance between the aforementioned a posteriori probabilities. Experimental results, based on the testing dataset given by the MICHE II Challenge organizers, indicate the appropriateness of the deployed method for the iris recognition task. Best results show a precision score above 90% even for iris images of new individuals.