Biometric analysis for remote examination system

Hockey, Josh (2011) Biometric analysis for remote examination system. [USQ Project] (Unpublished)

[img]
Preview
PDF
Hockey_2011.pdf

Download (2792Kb)

Abstract

Tertiary study is considered by many to be a necessary step in not only increasing knowledge, but also for employability and receiving promotions. Accessing Universities, particularly through on campus study can be problematic for those working full time or raising a family. The University of Southern Queensland promotes external studies and although they provide for the most part flexibility examinations are rigid. The ability to complete exams remotely may be an ideal solution; however security and exam integrity must be maintained. This dissertation considers the possibility of using biometric verification as a way of ensuring security. While many biometric techniques are considered only facial recognition adheres to all of the specifications of a remote examination system. Over the past twenty years facial verification has been a subject of interest, particularly in the area of security. Numerous algorithms have been proposed, stemming primarily from linear and multi-linear algebra. The Eigenface algorithm based on Principal Component Analysis was implemented to determine if accuracy rates required by a remote examination system could be achieved. The tests were conducted on the XM2VTS database as well as a small test database composed of images similar to those expected during a remote examination. The algorithm was implemented using MATLAB and results show that Eigenfaces was able to achieve results of 78.8% accuracy, with an imposter error as low as 19.36%. The tests conducted also conclude that pre-processing techniques, in particular face detection and cropping will increase verification rates of the algorithm. Facial verification is considered to be a feasible method of ensuring security and maintain examination integrity for a remote system.


Statistics for USQ ePrint 21949
Statistics for this ePrint Item
Item Type: USQ Project
Refereed: No
Item Status: Live Archive
Depositing User: epEditor USQ
Faculty / Department / School: Historic - Faculty of Engineering and Surveying - Department of Electrical, Electronic and Computer Engineering
Date Deposited: 17 Sep 2012 03:43
Last Modified: 03 Jul 2013 01:29
Uncontrolled Keywords: biometric analysis, remote examination system, facial verification
Fields of Research (FOR2008): 09 Engineering > 0906 Electrical and Electronic Engineering > 090699 Electrical and Electronic Engineering not elsewhere classified
URI: http://eprints.usq.edu.au/id/eprint/21949

Actions (login required)

View Item Archive Repository Staff Only