Hyperspectral Band Selection Based on Matrix CUR Decomposition

Band selection is an important technique for eliminating spectral redundancy of hyperspectral imagery (HSI) while preserving critical information. Recently, correlations among neighboring bands or pixels have been exploited in the form of graph regularizations to reduce the data dimensionality efficiently. However, manipulation of graph regularizations typically causes computational bottlenecks. In this work, we propose a robust method for hyperspectral band selection based on spatial/spectral graph Laplacians and matrix CUR decomposition. The efficiency of the proposed method has been shown on two real data sets by comparing with several other state-of-the-art band selection methods.