Iris Biometric Technique for Person Authentication Based on Fusion of Radon and 2D Multi-Wavelet Transform

Iris Biometrics is a recent biometric technique being extensively used for user authentication in banking applications, for access control in high security buildings etc. It has covered almost each and every area where security is an important concern. The system takes an iris image as an input and declares whether a person is genuine or imposter. The goal of this research is to give high accuracy at the same time it should fulfill functional, performance and reliability requirements. CHT based iris segmentation, radon and 2DMWT based feature extraction is used in our experiment. Iris segmentation is more challenging task as it affects further stages and at the last giving less accuracy. Most times acquired iris images are affected by poor illumination, off angles, motion blur, large standoff distances, improper focus and occlusion by eyelashes and eyelids etc. Euclidean distance formula have been used for matching between enrolled iris template and input iris template. The proposed system gives recognition rate of 86.052%