Fingerprint Quality Estimation Using Self-Organizing Maps

Anton Makarov

AbstractFingerprint quality assessment is a crucial task of fingerprint image acquisition process. Modern fingerprint quality assessment algorithms are required to possess high predictive accuracy along with high computational efficiency in order to be used in mobile devices with limited computational resources. Although several quality estimation methods have been proposed, a high computational complexity is attributable to most of them.
The goal of the thesis is to present a novel approach for fingerprint quality assessment based on Self-Organizing Maps as well as to analyze the proposed approach in terms of predictive performance, speed and computational complexity. In the thesis most important aspects of fingerprint quality assessment are covered and an overview of existing fingerprint quality estimation approaches is given. Experimental results presented in this thesis show that SOM network trained on a large data set of feature vectors derived from fingerprint images is capable to distinguish a large number of quality classes and predict utility values for fingerprint samples.
Results of comparative analysis of the proposed and existing fingerprint quality estimation approaches show superiority of proposed approach in terms of computational complexity and speed and show promising results to achieve superiority in the accuracy of quality predictions. This makes the proposed method very attractive to be used in mobile devices based fingerprint identification systems.
TypeMaster's thesis [Academic thesis]
PublisherTechnical University of Denmark, DTU Informatics, E-mail:
AddressAsmussens Alle, Building 305, DK-2800 Kgs. Lyngby, Denmark
NoteSupervised by Professor Rasmus Larsen,, DTU Informatics.
Electronic version(s)[pdf]
Publication link
BibTeX data [bibtex]
IMM Group(s)Image Analysis & Computer Graphics