Enhanced Iris Recognition based on Image Match and Hamming Distance

Abstract

Abstract Iris recognition and favor because of its high recognition rate, noninvasive and simple algorithm and other advantages, in a variety of biometric identification technology is very prominent. The iris texture feature extraction is the core of the iris recognition algorithm. Fractal geometry theory provides new ideas and methods to express nonlinear image information, the fractal dimension is an important parameter of fractal geometry, is a measure of complexity of irregular change, covering blanket dimension can better reflect the graphics changes in different resolution characteristics; missing is the fractal dimension and independent statistics, is a supplement to the fractal dimension, overcome the different texture characteristics may have the same fractal dimension of the problem. This paper presents a blanket and missing items based on the combination of texture feature extraction algorithm, can make full use of radiation in different resolution iris texture information and texture, classification ability make feature matrix has a better. Iris matching is the key of iris recognition. How to effectively carry out the matching of the iris code matrix is the decisive step in iris recognition. In this paper based on the normalized correlation classifier, the matching method of cyclic shift, eliminated from the same eyes of different iris image due to differences in rotation caused, improve matching accuracy.