Optimization of filter parameters for iris detection

Biometric system furnishes automatic recognition of a person based on the unique characteristics. Iris identification is one of the most authentic methods of biometric identification. Many iris realization system algorithms are developed which are capable to get good identification rates. But the outputs are obtained under favorable conditions only like input images are captured by same camera and having same characteristics like fixed zoom factor with same distance and fixed illumination. The present work is based on sensing, feature extraction and matching. The parameters (wavelength and angle θ) of the filter, used for feature extraction are optimized by using Genetic Algorithm so that the presented signal to noise ratio in an image reduce. The proposed method is examined on several images using CASIA database version 1.0.The experimental results demonstrate the effectiveness of proposed approach as the overall success is higher than the previous methods.