A Generalized Subspace Projection Approach for Sparse Representation Classification

摘要

In this paper, we propose a subspace projection approach for sparse representation classification (SRC), which is based on Principal Component Analysis (PCA) and Maximal Linearly Independent Set (MLIS). In the projected subspace, each new vector of this space can be represented by a linear combination of MLIS. Substantial experiments on Scenel5 and CalTechlOl image datasets have been conducted to investigate the performance of proposed approach in multi-class image classification. The statistical results show that using proposed subspace projection approach in SRC can reach higher efficiency and accuracy.