Sparse representation based classification by using PCA-SIFT descriptors
Feng‐Xiang Ge, Yishu Shi, Bo Sun, Feng Xu, Victor O. K. Li · 2014
Sparse representation based classification (SRC) is an efficient method with high recognition rate in many pattern recognition applications. Unfortunately, the original SRC method generally requires rigid alignment. In this paper, the feature-based SRC method is addressed by using PCA-SIFT descriptors. The presented method is not only efficient for alignment-free, face recognition, but also robust for the image illumination and affine, where the image processing is moved from pixel-domain into the feature-domain, i.e. PCA-SIFT descriptors. Experimental results show the presented method in this paper has higher recognition rate, more robustness, and lower computational complexity than MKD-SRC and SRC in the above scenarios.