BioTwist: overcoming severe distortions in ridge-based biometrics for successful identification

This thesis focuses on ridge-based and highly distorted biometrics, the different chal-lenges involved in a verification of identity scenario, and how to overcome them. More specifically, we work on ridge-based biometrics in two different contexts: (i) newborn and infant biometrics and (ii) quality estimation for forensic investigations. Finally, we merge both applications and demonstrate that techniques from and optimised for the forensic context can be transferred and applied to newborn and infant biometrics and clearly improve results. The thesis is situated in-between the research areas of engineering, applied mathematics and the forensic and medical sciences.