Logo des Repositoriums

Auflistung nach Autor:in "Beslay, Laurent"

1 - 2 von 2
  • Konferenzbeitrag
    Fingermark Quality Assessment: An Open-Source Toolbox
    (BIOSIG 2021 - Proceedings of the 20th International Conference of the Biometrics Special Interest Group, 2021) Oblak, Tim; Haraksim, Rudolf; Beslay, Laurent; Peer, Peter
    Fingermark quality assessment is an important step in a forensic fingerprint identification process. Often done in the scope of criminal investigation, it is performed by trained fingerprint examiners whose quality assessment can be rather subjective. The goal of this work is to develop an automated fingermark quality assessment tool, which would assist the fingermark examiners in their work. In this paper, we present a fast, open-source, and well documented fingermark quality assessment toolbox, which contains more than 20 algorithms for feature extraction, segmentation, and enhancement of fingermark images. We demonstrate the utility of the toolbox by assembling a feature vector and training various baseline machine learning models, capable of predicting the quality of fingermark images with high accuracy.
  • Konferenzbeitrag
    Fingerprint Quality: a Lifetime Story
    (BIOSIG 2018 - Proceedings of the 17th International Conference of the Biometrics Special Interest Group, 2018) Galbally, Javier; Haraksim, Rudolf; Beslay, Laurent
    Currently, it is a largely accepted fact that biometric sample quality is the most determinant factor to achieve high recognition accuracy in biometric systems. However, even in extensively researched characteristics such as fingerprints, there is still a lack of evidence on how quality evolves throughout the life of an individual. For instance, how does the quality of children fingerprints compare to that of adults or elders? Do these changes imply any age limits for the use of fingerprints with current technology? The present paper addresses this key problem based on a database of over 400K fingerprints coming from more than 250K different fingers. The database was acquired under real operational conditions and contains fingerprints from subjects aged between 0 and 98 years. Such a unique set of data has allowed us to analyse for the first time how fingerprint quality changes through life.