A Sparsity Adaptive Greedy Iterative Algorithm for Compressed Sensing

Aiming at the problem of signal reconstruction with unknown sparsity in compressed sensing, an improved iterative greedy algorithm is proposed to signal reconstruction from low sampling rate. Firstly, a sparsity estimation strategy is used to estimate the sparsity and the true support set of target signal, and then the residual is initialized with the estimated value. In order to optimize the estimated value, it is necessary to determine the direction of the iteration before the iteration begins. The proposed algorithm is associated with the ideas of adaptive, backtracking and greedy choice to iteratively refine. The simulation results show that the proposed algorithm performs better than that of OMP, CoSaMP and SAMP algorithms for both one-dimensional signal and two-dimensional image signal, and the algorithm has better practical application value.