Two Dimensional Discriminative Projection Based on Nearest Orthogonal Matrix and its Application to Face Recognition

To solve the interferential problem of illumination in face recognition,this paper proposes a method called two dimensional discriminative projection based on nearest orthogonal matrix（2DDP-NOM）.It obtains nearest orthogonal matrix representation of face image matrix by singular value decomposition firstly;then constructs the intra-class scatter based on nearest orthogonal matrix and the inter-class scatter based on nearest orthogonal matrix by nearest orthogonal matrix;lastly,obtains two dimensional discriminative projection by maximizing the inter-class scatter and minimizing the intra-class scatter simultaneously and gets the low dimensional features by this projection.Experiments are performed on Yale,CMU-PIE and AR databases and the experimental results demonstrate the effectiveness of 2DDP-NOM.