IRIS RECOGNITION AND VERIFICATION

Abstract

The use of human biometrics for automated identity verification has become widespread. The most commonly used human biometrics are the face, fingerprint, iris, gait, retina, voice, hand geometry, and so on. One ofthese is the outward-looking organ, whose unique pattern remains stable throughout life. These features are attractive to biometrically to identify individuals. We present a detailed study ofthe technique of iris recognition. This includes analysis ofthe reliability and accuracy ofthe iris as a biometric identification of the individual. The main stages of iris detection are segmentation, authentication, encoding and matching. In this job, automatic segmentation is done using a circular hog conversion process. Dogman rubber sheet model is used in the process of authentication. Four-step phase sizing based on 1D log Gabor filters is used to encode the special features ofiris into a binary template. Finally, the hamming distance is considered to evaluate the affinity at the machining point of the two models. We experimented with improved detection results for the CASIA-iris v4 database.