Iris Recognition system using Principal Components of Texture Characteristics
Patnala S. R. Chandra Murty, Yenumula B. Reddy · 2009
In this paper, a new algorithm for iris recognition based on texture characteristics is presented. Phasor information obtained from no rmalized iris templates is convolved with logical operators which have texture discrimination properties such as energy, contrast, randomness and size, to obtain the feature sets. The responses are further applied to PCA to identify th e patterns and reduce the feature space. Parametric and nonparametric classifiers are used to classify various iris templates extracted from different open source databases. The algorithm has low computational complexity and high classification accuracy which makes the recognition systems more reliable and fast.