Paper Details: Downloads: 1245
Serial Number: P1151420328
Title: Improved Iris Verification System
Authors: Basma M.Almezgagi and M. A. Wahby Shala and Hesham N. Elmahdy
Abstract: Most computer based systems need to be secured. Human iris is a unique identifier for each person which is used widely. Iris recognition is preferable for person verification. Iris verification divided in two phases: the enrollment phase and the testing phase. This paper tackles developing an improved iris verification system using Chinese Academy of Science Institute of Automation (CASIA) iris database. Basma Mohamed Hesham System (BMHS) consists of four main steps, they are; iris and pupil localization/segmentation using a Canny Edge Detection scheme and circular Hough Transform with a novel dynamic-based technique in determining iris radius range, iris normalization (using Daugman’s rubber sheet model), feature extraction/ encoding (using the convolution with log Gabor filter wavelets) and pattern matching (using hamming distance as a matching technique). This proposed dynamic-based technique in determining the iris radius led to improve the iris localization/segmentation level. We compared our system with other iris verification systems. BMHS is evaluated based upon False Acceptance Rate (FAR), False Rejection Rate (FRR) and computational cost required in both open and close set. A comprehensive experimental work shows an improvement in the false rejection rate and reduction in the system computational cost on different types of iris image.
Keywords: Localization; biometric identification; pupil
Journal/Conference: International Journal of Graphics, Vision and Image Processing
Volume: 14
Issue: 1
Submission Date: 5/13/2014 12:00:00 AM
Review Date: 5/17/2014 12:00:00 AM
Publishing Date: 5/20/2014 12:00:00 AM
Article Downloads: 1245
Download:

Facebook