Random matrix-based nonnegative sparse representation for hyperspectral image classification

With a consideration of the limitations of regular classification model using sparse representation (SR), an innovative model named Random Matrix-Nonnegative Sparse Representation (RM-NSR) is proposed to improve the classification results of hyperspectral imagery. The RM-NSR model introduces a random matrix inspired by random projection to improve the restricted isometry property (RIP) condition of measurement matrix in the regular SR model. The new model also considers the non-negativity of reconstructed sparse coefficient vectors. Based on Urban and PaviaU hyperspectral datasets, three different schemes in the RM-NSR model are utilized to recover the sparse coefficient and the classification results are compared with those of the regular SR model. Experimental results show that the RM-NSR model obviously outperforms the regular SR model in the average classification accuracies (ACAs). Furthermore, the relationship between the projected dimension of random matrix and the ACAs shows that a greater projected dimension guarantees the improvement of ACAs by the RM-NSR model.