Multi-Biometric Iris Recognition System Using Consensus Between Convolutional Neural Network Architectures

Multi-Biometric Iris Recognition System Using Consensus Between Convolutional Neural Network Architectures

Djalal Rafik Hammou, Sidi Ahmed Mahmoudi, Reda Adjoudj
Copyright: © 2022 |Pages: 30
DOI: 10.4018/IJOCI.305210
OnDemand:
(Individual Articles)
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

The field of biometrics represents an important domain and is currently undergoing extensive research to improve data security. Among the most used biometric modalities, we find the iris which, is specific and adequate. It is unique for each human being. Iris identification is used at international airports, sea, and land borders. In this article, we propose to use convolutional neural networks (CNN) with a consensus between the architectures (VGG, DenseNet, Resnet, Inception, MobileNet, and Xception). The goal is to classify and combine CNN architectures with the Softmax classifier focusing on consensus orientation using a voting system (Hard voting). In addition, the use of optimization techniques such as Transfer learning, Batch-normalization, Dropout, Data augmentation, Global average pooling, and Cross-validation will accelerate the learning process and avoid overfitting. Using the GPU with its parallelization will significantly speed up the processing of our program. The experiments of our approach were carried out on the MMU1 iris database with an accuracy reaching 100%.
Article Preview
Top

Introduction

Biometrics has become an essential tool in information security and personal data protection. A biometric recognition system allows the identification and authentication of a person using a specific template. This data can be morphological or behavioral. Among the morphological biometric modalities, we find the fingerprint, face recognition, the veins of the apple of the hand, the iris, and the retina. The existing behavioral biometric modalities include the way of walking, voice recognition, the mechanics of a keystroke, the method of a signature, the style of writing, etc. The purpose is to use the modalities in legal and administrative applications, commercial and banking applications, and public service applications.

Anatomy of the Eye

The human eye is shaped like a spherical ball. It is not circular, but elliptical. It is very similar to the earth's structure in everything at the north and south poles (see Figure 1). The diameter of the eye is between 2.2 cm and 2.4 cm. It has a diameter of 24 mm on the anteroposterior side, 23.5 mm on the longitudinal side, and 23 mm on the vertical side. The eye has a circumference of 77 mm. Its volume is 6.5 ml, and its weight varies between 7 gm and 8 gm (Khurana et al., 2015).

The organic structure of the human eye consists of two main compartments: the anterior and the posterior segment. The anterior segment is composed of the anterior and posterior chambers. The anterior room contains the components that surround it. On the right, we find the rear surface of the pupil and the anterior part of the iris. Above and below the right side, we can see a part of the ciliary body. The end of the right side is the lens, which forms a direct border with the posterior chamber. On the left is the cornea, and finally, the anterior chamber contains 0.25 mL of aqueous humor. The posterior chamber is delimited by a large part of the posterior surface of the iris. On the left, there are the border cristalline and zonules. Above that is the ciliary body. The space in this room contains approximately 0.06 mL of aqueous humor. Finally, the posterior segment encompasses many elements (the retina, vitreous humor, choroid, right lateral muscles, medial rectal muscles, and the structures posterior to this segment) (Khurana et al., 2015).

Complete Article List

Search this Journal:
Reset
Volume 14: 1 Issue (2024): Forthcoming, Available for Pre-Order
Volume 13: 1 Issue (2023)
Volume 12: 4 Issues (2022)
Volume 11: 4 Issues (2021)
Volume 10: 4 Issues (2020)
Volume 9: 4 Issues (2019)
Volume 8: 4 Issues (2018)
Volume 7: 4 Issues (2017)
Volume 6: 4 Issues (2016)
Volume 5: 4 Issues (2015)
Volume 4: 4 Issues (2014)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing