Face anti-spoofing based on Radon transform
Răzvan Daniel Albu · 2015
In this paper I will present a technique to generate a digital signature for an image, which will uniquely identify it, using Radon transform. Even if Radon based approaches are broadly applicable to tomography (the construction of an image from the projection data related with cross-sectional scans of it), in this research work I will show how it can be successfully utilized to classify images and detect face spoof attacks. This increased popularity of face recognition system has raised disquiets about biometric attacks, where a photo of an authorized person's face could be used by unauthorized individuals to access diverse services. The presented results prove this method offers similar performances to the state of the art and they were obtained using MATLAB 2013 environment and NUAA Photograph Imposter Database.