Date of Award

5-2015

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Computer Science

Advisor

Dr. Sophie Jorg

Committee Member

Dr. Timothy Davis

Committee Member

Dr. D.E. Stevenson

Committee Member

Dr. Andrew Duchowski

Abstract

As biometrics become more prevalent in society, the research area is expected to address an ever widening field of problems and conditions. Traditional biometric modalities and approaches are reaching a state of maturity, and their limits are clearly defined. Since the needs of a biometric system administrator might extend beyond those limits, new modalities and techniques must address such concerns. The goal of the work presented here is to explore the periocular region, the region surrounding the eye, and evaluate its usability and limitations in addressing these concerns. First, a study of the periocular region was performed to examine its feasibility in addressing problems that affect traditional face- and iris-based biometric systems. Second, the physical structure of the periocular region was analyzed to determine the kinds of features found there and how they influence the performance of a biometric recognition system. Third, the use of local appearance based approaches in periocular recognition was explored. Lastly, the knowledge gained from the previous experiments was used to develop a novel feature representation technique that is specific to the periocular region. This work is significant because it provides a novel analysis of the features found in the periocular region and produces a feature extraction method that resulted in higher recognition performance over traditional techniques.

Share

COinS