ABSTRACT
The biometric person identification technique based on the pattern of the human iris is a very
costly top secure application. This paper describes a personal identification system based on
the human iris imaging through providing a set of algorithms that describes image acquisition,
image segmentation, feature extraction and pattern forming. For image acquisition, we present
an image enhancement algorithm in order to get more accurate image feature results. In
addition, we propose a boundary localization algorithm, which used to find the pupil
boundary. A new iris recognition method that analyzes local variations of the iris is used to
construct a Feature Selection Vector (FSV) that can be used to extract features of any iris
image size. Extensive experimental results using Pearson's correlation coefficient to verify
one's identity on CASIA iris images database shows that the proposed system is effective and
encouraging.