Iris recognition for person identification

Abstract

Security has always been an important issue. It has developed drastically over the years and today there is a sector of security, known as 'Biometrics', which deals with identifying a person though unique personal traits, through methods such as: Facial Recognition, Fingerprint Recognition, Retina Scanning, Hand Geometry, Signature Dynamics, Speaker Verification, Palm print and Hand Vein. All of the methods are good, but some perform better than others in certain situations. The reliability and effectiveness of a system depends on how it is implemented and for which reason. Iris recognition for example has the lowest error rate when compared to all other biometric systems due to the large number of degrees of freedom it has. Iris recognition systems implemented in the market all use the method John G. Daugman developed and his algorithms. This involves a process of obtaining the image, image segmentation, image normalization, data extraction, template comparisons. This thesis will aim to emulate Daugman's Algorithms through MATLAB, by using some of Daugman's algorithms together with other more time efficient methods and comparing results at the end.