Macquarie University
01whole.pdf (13.21 MB)

Improved PCA-based techniques for face and finger-vein recognition systems

Download (13.21 MB)
posted on 2022-03-28, 12:29 authored by Sepehr Damavandinejadmonfared
Biometrics refers to metrics related to human characteristics and biometric technologies use physical or behavioural characteristics of people to identify and recognize an individual. With the ever increasing growth in cybercrimes, biometric technologies have become increasingly important in this digital information age. The focus of this thesisis on the design ans application of new techniques for image based biometrics recognition systems using faces and finger veins. Principal Component Analysis (PCA) is a way of identifying patterns in data, and expressing the data in such a way as to highlight their similarities and differences. The PCA technique plays an important role in feature extraction and dimensionality reduction. In this thesis, we propose a variety of extensions of PCA technique using kernel and entropy based extensions. We have proposed new algorithms to face and finger vein recognition systems and show that these extensions can yield better results in terms of accuracy and performance.


Table of Contents

1 Introduction -- 2 Background -- 3 Databases -- 4 Kernel Principal Component Analysis -- 5 Kernel Entropy Component Analysis in Face Recognition -- 6 Kernel Entropy Component Analysis in Finger-Vein Recognition -- 7 Proposed New Dimensionality-Reduction Method -- 8 New Image Transformation Method (FDGDA) -- 9 Feature Dependent Kernel Entropy Component Analysis -- 10 Finger-Vein Extraction -- 11 Overall Conclusion and Future Work.


Theoretical thesis. Bibliography: pages 285-295

Awarding Institution

Macquarie University

Degree Type

Thesis PhD


PhD, Macquarie University, Faculty of Science and Engineering, Department of Computing

Department, Centre or School

Department of Computing

Year of Award


Principal Supervisor

Vijay Varadharajan


Copyright Sepehr Damavandinejadmonfared 2017. Copyright disclaimer:




1 online resource (xxvi, 295 pages) illustrations (some colour)

Former Identifiers