Compressive Image Sensing: Turbo Fast Recovery With Lower-Frequency Measurement Sampling
Fast and accurate compressive sensing recovery is still a challenging issue, and has received considerable attention in the literature. In this paper, the authors do not follow the tradition of imposing certain sparsity patterns on a CS recovery algorithm. On the contrary, they propose to design a new and novel sampling matrix for the purpose of preserving important measurements. They have also studied 1D sensing and 2D separate sensing strategies for 1D signals and 2D images, respectively. Compared to 1D sensing, 2D separate sensing is found to be particularly feasible in compressive sensing of large-scale images in terms of storage and computation overhead reduction and reconstruction quality improvement.