Hyperspectral Image Classification Using Compressive Sampling Measurements

In this paper, we develop a new approach for hyperspectral image classification directly from the compressive sensing measurements without reconstructing the original hyperspectral image first. The proposed method is based on the fact that each pixel in the hyperspectral image lies in a low-dimensional subspace, and thus it can be represented as a sparse linear combination of vectors in a dictionary obtained from training samples. In compressive sensing theory, with the sparsity prior, we can reconstruct the original signal from the random sampling measurements using appropriate algorithms. And finally the recovered sparse vector is used to determine the class label of the test pixel by the nearest neighbor classifier. The proposed method can fulfil the classification task and reconstruction at the same time.