Illumination-Invariant Face Recognition in Hyperspectral Images

Illumination-invariant face recognition remains a challenging problem. Previous studies use either spatial or spectral information to address this problem. In this paper, we propose an algorithm that uses spatial and spectral information simultaneously. We first learn a basis in the spectral domain. We then extract spatial features using 2D Gabor filters. Finally, we use the basis and the spatial features to classify face images. We demonstrate the effectiveness of the algorithm on a database of 200 subjects