Liveness detection of dorsal hand vein based on the analysis of fourier spectral

Abstract

Liveness detection of dorsal hand vein is a necessary step towards higher reliability of identification and is attracting increasing attention of researchers. However, there's only few published research in this area. This paper proposes a novel method for liveness detection of dorsal hand vein. First, by applying the Fourier Transform, a feature is extracted as a statistical value of spectral energy derived from every blocked spectrum of single wavelength infrared images. Second, regarding the principle of blocking, massive experiments have been performed to find the optimum feature with the maxmin criterion. Furthermore, an SVM classifier is employed for clustering. The experimental results have verified the effectiveness of our proposed method. © Springer International Publishing 2013.