Date of Graduation


Document Type


Degree Type



Statler College of Engineering and Mineral Resources


Lane Department of Computer Science and Electrical Engineering

Committee Chair

Afzel Noore

Committee Co-Chair

Mayank Vatsa

Committee Member

Mayank Vatsa

Committee Member

Richa Singh


The widespread use of smartphones has spurred the research in mobile iris devices. Due to their convenience, these mobile devices are also utilized in unconstrained outdoor conditions. At the same time, iris recognition in the visible spectrum has developed into an active area of research. These scenarios have necessitated the development of reliable iris recognition algorithms for such an uncontrolled environment. Additionally, iris presentation attacks such as textured contact lens pose a major challenge to current iris recognition systems.

Motivated by these factors, in this thesis, a detailed analysis of the effect of textured contact lenses on iris recognition in an uncontrolled environment is presented by creating Mobile Uncontrolled Iris Presentation Attack Database. It consists of more than 10,000 iris images of subjects wearing textured contact lens and without wearing contact lenses captured in an indoor and outdoor environment using a mobile iris sensor. The first contact lens database in the visible spectrum is also introduced, named as UVCLI database. Using this database, a detailed analysis of the effect of textured contact lenses on iris recognition in the visible spectrum is presented. It is observed that textured contact lenses degrade the unconstrained iris recognition performance by over 25% and thus, may be utilized intentionally or unintentionally to attack existing iris recognition systems. Also, the first study to investigate the impact of textured contact lenses on identity impersonation is performed. The results demonstrate that a perpetrator can impersonate an enrolled subject by wearing the same textured contact lens.

This thesis also presents a novel algorithm, DensePAD, which utilizes DenseNet based convolutional neural network architecture for the near-infrared spectrum based iris presentation attack detection. The proposed algorithm is trained on a combined iris database of more than 270, 000 real and attack iris images. In-depth experimental evaluation of this algorithm reveals its superior performance in detecting iris presentation attack images on the combined iris database as well as the proposed MUIPAD.

DY_MS_Form.pdf (834 kB)
Electronic Thesis and Dissertation Committee Signature Form