A comparison of fingerprint image quality and matching performance between healthcare and general populations

Abstract

Research has shown for some age groups, quality of fingerprints can impact the performance of biometric systems. A desirable feature of biometrics is that they are suitable for use across the population. This applied study examines the performance of a fingerprint recognition system in a healthcare environment. Anecdotal evidence suggested front line healthcare workers may have lower image quality due to continued hand washing which may remove oils from their skin. During training, individuals are told to add oil to their fingers by wiping oil from their foreheads to improve the resulting quality of the fingerprints. In the healthcare population the authors tested, compared to two general populations (collected on optical and capacitance sensors) there was a significant difference in skin oiliness, but not in image quality. There was a difference across healthcare and non-healthcare groups in the performance of the fingerprint algorithm when compared against the capacitance dataset.