Experimental Results on the Feature-level Fusion of Multiple Fingerprint Liveness Detection Algorithms
GHIANI, LUCA;MARCIALIS, GIAN LUCA;ROLI, FABIO
2012-01-01
Abstract
The aim of fingerprint liveness detection is to detect if a fingerprint image, sensed by an electronic device, belongs to an alive fingertip or to an artificial replica of it. It is well-known that a fingerprint can be replicated and standard electronic sensors cannot distinguish between a replica and an alive fingerprint image. Accordingly, several countermeasures in terms of fingerprint liveness detection algorithms have been proposed, but their performance is not yet acceptable. However, no works studied the possibility of combining different feature sets, thus exploiting the eventual complementarity among them. In this paper, we show some preliminary experiments on feature-level fusion of several algorithms, including a novel feature set proposed by the authors. Experiments are carried out on the datasets available at Second International Fingerprint Liveness Detection Competition (LivDet 2011). Reported results clearly show that multiple feature sets allow improving the liveness detection performance.File | Dimensione | Formato | |
---|---|---|---|
Experimental results on the feature-level fusion of multiple fingerprint liveness detection algorithms.pdf Solo gestori archivio
Dimensione 940.07 kB
Formato Adobe PDF
|
940.07 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.