Konferenzbeitrag
Improved Liveness Detection in Dorsal Hand Vein Videos using Photoplethysmography
Lade...
Volltext URI
Dokumententyp
Text/Conference Paper
Zusatzinformation
Datum
2020
Autor:innen
Zeitschriftentitel
ISSN der Zeitschrift
Bandtitel
Verlag
Gesellschaft für Informatik e.V.
Zusammenfassung
In this study, a previously published infrared finger vein liveness detection scheme is
tested for its applicability on dorsal hand vein videos. A custom database consisting of five different
types of presentation attacks recorded with transillumination as well as reflected light illumination is
examined. Additionally, two different methods for liveness detection are presented in this work. All
methods described employ the concept of generating a signal through the change in average pixel
illumination, which is referred to as Photoplethysmography. Feature vectors in order to classify a
given video sequence are generated using spectral analysis of the time series. Experimental results
show the effectiveness of the proposed methods.