S2DPCA with DM and FM in Face Recognition

Face symmetrical feature can be applied to two-dimensional principal component analysis (2DPCA) for face image feature extraction, and this procedure can be called symmetrical 2DPCA (S2DPCA). Now, these S2DPCA-based face recognition algorithms almost pay much attention to the feature extraction, and the classification measures have been little investigated. In this paper, the typical similarity measure used in 2DPCA is applied to S2DPCA, which is the sum of the Euclidean distance between two feature vectors in feature matrix, called distance measure (DM). The similarity measure based on Frobenius-norm is also developed to classify face images for S2DPCA. Furthermore, the relative theories on S2DCPA are proofed. The experimental results on ORL and FERET face databases show that S2DPCA has the potential to outperform traditional 2DPCA, especially on condition that DM is used for S2DPCA.