On the vulnerability of palm vein recognition to spoofing attacks

The vulnerability of palm vein recognition to spoofing attacks is studied in this paper. A collection of spoofing palm vein images has been created from real palm vein samples. Palm vein images are printed using a commercial printer and then, presented at a contactless palm vein sensor. Experiments are carried out using an extensible framework, which allows fair and reproducible benchmarks. Results are presented comparing two automatic segmentations. Experimental results lead to a spoofing false accept rate of 65%, thus showing that palm vein biometrics is vulnerable to spoofing attacks, pointing out the importance to investigate countermeasures against this type of fraudulent actions. A study based on the number of the enrolment samples is also reported, demonstrating a relationship between the number of enrolment samples and the vulnerability of the system to spoofing.