I. Introduction
Biometric-based authentication systems are used for a gamut of applications such as law enforcement, border security, surveillance, etc. Among all the different biometric modalities which may require expensive sensing device or may not work reliably in uncontrolled settings, fingerprint is one of the most widely used modality. One of the key component which attributes to the robustness of a fingerprint matching system is the fingerprint region of interest (roi) segmentation module. A fingerprint roi segmentation module is dedicated towards segmenting foreground fingerprint region with clear ridge patterns from the background noise. Noise in a fingerprint can originate due to presence of oil, grease or dirt on surface of the fingerprint sensor used to acquire the fingerprints. A fingerprint roi segmentation module serves dual purpose. Firstly, it minimizes spurious minutiae (feature) detection which translates to improved matching performance. Secondly, it limits the matching to only foreground which reduces computational time for matching.
Schematic diagram showcasing the benefit of proposed feature alignment method in improving segmentation performance on a sensor whose ground truth annotations are not available for training. For better understanding, roi marked fingerprints are presented instead of binary roi mask.