Texture Descriptor Extraction for Iris Classification using Fuzzy Logic Local Binary Pattern

• Published in: 2020 10th IEEE International Conference on Control System, Computing and Engineering (ICCSCE) pp. 29 - 32
• Main Author: Ali, Ahmad N.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.08.2020
• Subjects: Feature extraction; Fuzzy logic; Image segmentation; Iris; iris classification; Iris recognition; Kernel; local binary pattern; Support vector machines; vector support machines
• DOI: 10.1109/ICCSCE50387.2020.9204949

Feature extraction in the iris classification field is an essential stage that is applied to produce a discriminative descriptor for each subject. Uncertainty due to contrast and other noises may degrade the exactness of descriptors between subject. In achieving to tackle the problem, fuzzy logic local binary pattern concept that uses a specific rule to decide the uncertainty intensity values between central and neighboring pixels has been applied for feature extraction. An experimental setup has been performed to evaluate the concept using two freely available databases which are CASIA V3 and IITD. A promising result with more than 98% of the classification score success rate has achieved by using Support Vector Machines classifier.