Fingerprint quality assurance using image processing

Marek Dusio

AbstractThe subject of this dissertation is measuring fingerprint sample quality using image analysis methods that are fast to compute. Of interest are analysing the impact of fingertip skin moisture on quality and providing feedback if the finger is too dry or too wet. The goal is to propose methods that could be incorporated in NIST Finger Image Quality version 2.0 or in ISO/IEC standards.
A dataset of 6600 samples is collected from 33 subjects using 5 sensors with objective fingertip moisture measurement and varying skin moisture conditions.
The impact of skin moisture on fingerprint sample quality is analysed and a Moisture Indication method is proposed and used with thresholds to provide binary indication on skin dryness or wetness.
Three fingerprint Quality Measurement Algorithms are proposed - Ridge Valley Difference, Ridge Line Count and Contrast; their performance is assessed in terms of execution time, output quality correlation with observed utility, and using Error versus Reject Curves. The methods are compared to current state of the art: NIST Finger Image Quality, Orientation Certainty Level, Ridge Valley Uniformity, Local Clarity Score and Gabor Shen.
All proposed methods work and offer acceptable performance - all are fast to compute and provide quality that predicts samples' performance. Some proposed methods are better than state of the art in terms of either execution time or in performance prediction. The Moisture Indication method is successfully used to classify samples as acquired from dry or wet skin with reasonable detection error rates. All proposed methods can possibly be incorporated in the ISO/IEC standards or in NFIQ 2.0.
TypeMaster's thesis [Academic thesis]
PublisherTechnical University of Denmark, DTU Compute, E-mail:
AddressMatematiktorvet, Building 303-B, DK-2800 Kgs. Lyngby, Denmark
NoteDTU supervisor: Rasmus Larsen,, DTU Compute
Electronic version(s)[pdf]
Publication link
BibTeX data [bibtex]
IMM Group(s)Image Analysis & Computer Graphics