The Affection of Gabor Parameters to Iris Recognition and Their Optimization

In this paper, a Gabor filter optimization method based on real-coded genetic algorithm is presented for iris recognition. First, we list Gabor filter parameters and analyzed the validity of the expression for texture features. Then, since Gabor parameters has a great influence in Correct Recognition Rate, we took Gabor kernel parameters as chromosomes and Discriminative Index as fitness to on the CASIA V3 and JLUBR-IRIS for optimization. Moreover, the optimized Gabor filters are adopted to extract features for corresponding iris databases, which can obtain excellent results.