Analysis and Application of Biometric Recognition Methods Using Iris Scanning Techniques

In this thesis work, the focus is on exploring the use of iris recognition as a reliable method for biometric authentication, highlighting its advantages over traditional knowledge-based and token-based methods. In the first chapter, an introduction is provided on these conventional authentication methods, their limitations, and why biometric authentication is increasingly necessary. Various biometric systems are discussed, with emphasis on the superiority of iris recognition in terms of security and accuracy. The second chapter describes the methodology, detailing each step of the developed iris recognition system, from image acquisition to pupil and iris segmentation (using Daugman's integrodifferential operator), iris normalization (using Daugman's rubber sheet model), feature extraction (using 1D Log-Gabor filters), and feature encoding to generate the final iris code template. The process culminates in the calculation of Hamming distance for feature matching step. The final chapter presents experimental results, analyzing key metrics such as FRR, FAR, TRR, and TAR, and assessing how different parameter settings affect security and user experience. Trade-offs between these metrics are evaluated, and future improvements are suggested.