A Feature Extraction Method for Iris Recognition System Based on CNN(Transfer Learning
Loading...
Date
2023-09-20
Journal Title
Journal ISSN
Volume Title
Publisher
University of M'sila
Abstract
Iris recognition refers to the automated process of recognising individuals based
on their iris patterns. The seemingly stochastic nature of the iris stroma makes
it a distinctive cue for biometric recognition. This textural descriptor has been
observed to be a robust feature descriptor with very low false match rates
and low computational complexity. However, recent advancements in deep
learning and computer vision indicate that generic descriptors extracted using
Convolutional Neural Networks (CNNs) are able to represent complex image
characteristics. Deep CNN is a powerful visual model of machine learning.
We tend to present robustness and an effective structure for the iris recognition
system. The image first pass through these stages: enhancing the image
quality, determining the iris and pupil centre and radius for iris segmentation,
and converting the image from the Cartesian coordinates to the polar coordinates
to reduce the time of processing. The proposed system is named IRISNet
which extracts the feature and classifies them automatically without any domain
knowledge. The architecture of IRISNet consists of CNN layers to extract
features and a softmax layer to classify them into N classes for training CNN,
the back-propagation algorithm and Adam optimisation method are used for
updating the weights and the learning rate, respectively. The performance of
the proposed system was evaluated using the Sdumla iris database. The results
obtained from the proposed system outperform the supervised classification
model (VGG16, MobileNet, Inception, and Xception). The identification rate
is 97.32% and 96.43% for original and normalised images, respectively. The
recognition time per person is less than 1s. Experimental results conclude that
the proposed work obtained good performance compared to existing methods
Description
Keywords
Iris, CNN, Training, Convolution, Deep Learning, Image Recognition, Testing, Feature Extraction.