Biometric authentication is based on the physiological and behavioral features of a human being, and these are used more and more to produce secure and reliable methods of authentication. There is very little information available, if any, about the effects of eye disease and aging of the human eye on biometric authentication - namely iris and retina recognition. Two of the few things that are absolute certain in this world: people get older, and people get sick, so these are important factors to look at.
The master thesis topic is eye disease and aging of the eye, and how this affects iris and retina recognition as means of biometric authentication.
In the course of this MSc project, several diseases and aging-related effects are to be selected and investigated. These shall include but not be limited to glaucoma, macular degeneration (both wet and dry), cataracts, and pathological angiogenesis.
For each of these pathologies, a simulation of the typical disease pattern is to be designed and applied to a set of iris and retina images in such a way that a progression of the respective pathology can be applied and presented to sensors.
To ensure that the simulations are realistic, these shall be reviewed by subject matter experts (i.e. ophthalmologists) and a comparison with actual instances of the simulated pathology shall be effected.
To assess the impact of the pathology on recognition accuracy, an experimental protocol is to be designed in which a selected number of algorithms (including the Daugman algorithm for iris recognition) are presented with the results of the simulation. This experiment shall then be conducted, evaluating the recognition performance of the selected algorithms when confronted with the simulated disease and aging patterns.
