Sparse Subspace Target Detection for Hyperspectral Imagery

摘要

In this paper, we propose a new sparsity-based algorithm for automatic target detection in hyperspectral images (HSI). Thisalgorithm is based on the concept that a pixel in HSI lies in a low-dimensional subspace and thus can be represented bya sparse linear combination of the training samples. The sparse representation (a sparse vector representing the selectedtraining samples) of the test sample can be recovered by solving an to-norm minimization problem. With the recent de-velopment of Compressed Sensing theories, the minimization problem can be recast as a linear programming or solvedefficiently by a greedy pursuit algorithm. Once the sparse vector is obtained, the class of the test sample can be directlydetermined by the behavior of the vector on reconstruction. In addition to the constraints on sparsity and reconstructionaccuracy, we also exploit the fact that HSI are usually smooth in that neighboring pixels have a similar spectral charac-teristic. In our proposed algorithm, a smoothness constraint is also imposed by forcing the Laplacian of the reconstructedimage to be zero in the minimization process. The proposed sparsity-based algorithm is applied to several hyperspectralimages to detect targets of interest. Simulation results show that our algorithm outperforms the other HSI target detectionalgorithms, including the popular spectral matched filters, matched subspace detectors, and adaptive subspace detectors.