28 IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE MARCH 2023 Factorization-Based Approaches Rank Minimization Approaches Nonlocal Similarity Total Variation Subspace Representation Tucker Decomposition Nonlocal Tensor Decomposition TR Representation Inpainting Tucker Decomposition Tucker Decomposition + Total Variation Destriping Tucker Decomposition + Nonlocal Similarity Anomaly Detection Total Variation + Tensor Nuclear Norm + Nonlocal Similarity + Tucker Decomposition Truncated Tensor Nuclear Norm TR Representation Tensor-Based Greedy Multidimensional Block Sparsity Kronecker Compressive Sensing Spectral Unmixing Block Term Decomposition Compressive Sensing CP Decomposition + Total Variation + Sparsity + Local Geometry + Nonnegative Matrix Factorization + Convolution Operation Tucker Decomposition FIGURE 1. The main tensor decomposition-based methods for HS data processing. PCA: principal component analysis. + Sparsity Tensor PCA HS Data Processing Tucker Decomposition + Unmixing + PCA + Clustering + Tensor Nuclear Norm + Endmember Extraction Deblurring Restoration Other Priors Regularized Tensor Decomposition HS-MS Fusion Block Term Decomposition TT Decomposition TR Decomposition Tensor SVD + Total Variation + Nonlocal Similarity CP Decomposition Tensor Decomposition + Interimage Variability Denoising Tucker Decomposition + Higher-Order Singular Value Decomposition + Nonlocal Sparsity + Total Variation + Graph + Graph + Matrix Factorization + Nonlocal Similarity