This letter proposes a novel compressive sensing reconstruction method for correlated images by using joint regularization,<\/strong><\/p>\n\n\n\n where<\/strong> a compensation-based adaptive total variation (CATV) regularization and a multi-image nonlocal low-rank<\/strong><\/p>\n\n\n\n (MNLR) regularization are included. In CATV, local weights are assigned to the residual values in the gradient domain so as to<\/strong><\/p>\n\n\n\n constrain<\/strong> the regularization strength at each pixel. In MNLR, the search of similar patches goes across different images so that both<\/strong><\/p>\n\n\n\n self-<\/strong>similarity and inter-image similarity are explored. Afterward, an efficient algorithm is proposed to solve the joint formulation,<\/strong><\/p>\n\n\n\n using<\/strong> a Split-Bregman-based technique. The effectiveness of the proposed approach is demonstrated with experiments on both<\/strong><\/p>\n\n\n\n multiview<\/strong> images and video sequences.<\/strong><\/p>\n\n\n\n Index Terms\u2014Compressive sensing, motion estimation\/disparity estimation (ME\/DE), nonlocal low-rank regularization (NLR), total variation.<\/strong><\/p>\n\n\n\nSource Code<\/h3>\n\n\n\n