关键词: 压缩感知, 稀疏度自适应匹配追踪, 稀疏度, 广义正交匹配追踪, 贪婪类重构算法

Abstract: The basic idea of compressed sensing theory is that the original signal is sparse in a transform domain or compressible, and the sampling process and the compression process in the Nyquist sampling theorem are combined into one. Sparse Adaptive Matching Pursuit（SAMP） algorithm can realize the reconstruction under unknown sparsity, and the generalized orthogonal matching pursuit algorithm selects multiple atoms at each iteration, which improves the convergence speed of the algorithm. This paper proposes a Generalized Sparse Adaptive Matching Pursuit（gSAMP） algorithm based on the advantages of the above two reconstruction algorithms, and then the peak signal to noise ratio, reconstruction time, relative error, etc. of the reconstructed image are proposed. Objective evaluation indicators and subjective visual comparisons of the proposed algorithm and the traditional greedy algorithm. When the compression ratio is fixed at 0.5, the reconstruction effect of the gSAMP algorithm is better than that of the traditional greedy reconstruction algorithms such as MP, OMP, ROMP, SAMP and gOMP.

Key words: compressed sensing, Sparsity Adaptive Matching Pursuit（SAMP）, sparsity, generalized orthogonal matching pursuit, greedy reconstruction algorithm