Robust Face Recognition under Varying Light Based on 3D Recovery

This paper addresses face recognition under varying light via 3D reconstruction based on the techniques of shape from shading (SFS). First, we improve the geometric-based SFS by introducing the integrability constraint as one of the regular terms. This operation preserves the local curvedness of the recovered surface. Second, we propose a novel method to investigate human face recognition in the illumination varying case using local topographic information, such as curvedness and shape index extracted from intensity images by SFS algorithms. The experimental results have shown that the curvedness and shape index are suitable for representing 3D local features, and also it is insensitive to light variations since only 3D information is involved. Compared with typical face recognition approaches based on principal component analysis (PCA) plus linear discriminant analysis (LDA), the proposed method has demonstrated a better performance. This implies local topological properties are effective attributes for face recognition under light variations