Physical object authentication with correlated camera noise.
Fokko Beekhof, Slava Voloshynovskiy, Maurits Diephuis, Farzad Farhadzadeh · 2013
Abstract: In this paper, we address the problem of physical object authentication based on surface microstructure images. The authentication is based on digital con-tent fingerprints computed from the microstructure images. We analyze the impact of camera distortions which follow additive correlated Gaussian noise. The optimal decision rule is derived and the performance is analyzed in the direct and transformed domain. The theoretical derivations are supported by experimental results obtained for the FAMOS dataset. 1