|Author:||Kong, Wai-kin Adams|
|Title:||Using texture analysis on biometric technology for personal identification|
|Subject:||Hong Kong Polytechnic University -- Dissertations|
Identification -- Automation
Pattern recognition systems
|Department:||Department of Computing|
|Pages:||vi, 108 p. : ill. ; 30 cm|
|Abstract:||Using texture analysis as a tool to extract biometric features for personal identification is the main goal of this thesis. In this study, three biometrics, the iris, paimprint and ear are investigated. Each of them shows a different level of achievement. As far as iris recognition is concerned, we propose a new noise detection model for accurate segmentation of an iris. Eyelashes, the eyelids and reflection are the three main sources of noise. The eyelid issue has been solved by the traditional eye model; however, eyelashes and reflection have yet to be addressed. To determinate a pixel belonging to an eyelash, our model follows three criteria: 1) separable eyelash condition, 2) non-informative condition and 3) connective criterion. For reflection, strong reflection points are detected by a threshold and the weak reflection areas around the strong points are determined by a connective criterion and a statistical test. Using Boles's [47-49] texture-based iris recognition approach to evaluate the accuracy and usefulness of our detection model, we find the experimental results encouraging. For palmprint identification, we develop a novel textured feature extraction technique, in which a 2-D Gabor filter is used to obtain the texture information and two palmprint images are compared by their hamming distance. The experiments give impressive results and show that our method is effective and comparable with fingerprint (FingerCode), iris (IrisCode) and 3-D hand geometry. For ear recognition, we consider two issues: 1) image acquisition and 2) textured feature extraction technique. We have developed a special device for image acquisition. We also propose a novel feature extraction for ear recognition that measures two ear features by a simple vector norm. The experimental results show that ear recognition can provide middle level security.|
|Rights:||All rights reserved|
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item: