Retina Identification Algorithm Based on Bifurcation Points and SURF Descriptor

Abstract

The recent advanced developments in the research have enabled the industries to find more complex methodologies for personal authentication. Biometric authentication has become more important because of the increasing activities of terrorists and hackers. Retina biometric security system is one of the more authoritative security systems, because no two people have the same retinal pattern. This research proposed an idea for human identification model instituted on retinal images. Despite the images of digital retina constantly sustain from distortion, the Speed Up Robust Features (SURF), that is famed for its disparateness and invariability for gauge and turn over that is inserted to retinal instituted consistency. To fix the hurdle, a new pre-processing technique instituted on detecting the Bifurcation Points (BP) that are detected with the help of structure process. By this BP, it is allocate the important points that are used in SURF descriptor for more similarity transformation. In accordance with more transference by the repeated of locative substance sleek technique, the issue of not useful SURF key points is reduced significantly. Experiments display the results of the proposed model is around 98% exact so it became outstanding to a mighty range against former systems.