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.
