Journal ID : TRKU-08-10-2020-11221
[This article belongs to Volume - 62, Issue - 10]
Total View : 377

Title : Fuzzy Logic Approach to Local Binary Pattern for Extracting Iris Information with Support Vector Machines Classifier for Classification

Abstract :

Local structures within an image generally represents a challenge task to discriminate an accurate image texture specifically for some degree of uncertainty due to contrast and other noises on iris images. Feature encoding is a crucial stage so that robust feature descriptors can be produced for representing personal statistical information as failing to tackle it may degrade the performance. In achieving promising result, an evaluation on observing the capability of the fuzzy logic local binary pattern (FLLBP) to encode the iris texture is studied in this paper. An experimental setup based on the degree of fuzziness and the number of samples in the training data is performed and evaluated using iris images from CASIA V3 and IITD database. The results demonstrate a good accuracy score and show that the encoded descriptors from the FLLBP method are able to achieve a significant performance

Full article