基于MATLAB的图像压缩感知算法的研究

Research of Image Compression Sensing Algorithm Based on the MATLAB

由于压缩感知技术突破了奈奎斯特准则的局限性,文章对压缩感知算法中的关键技术稀疏变换基、观测矩阵和重构算法进行了研究,并在MATLAB中对不同的稀疏变换基和不同观测矩阵及重构算法进行仿真。通过不同的组合,仿真结果表明在其他条件相同下,采用离散小波基、部分傅里叶变换矩阵和子空间匹配追踪具有良好的图像重构效果。

Since compressed sensing technology breaks through the limitations of Nyquist’s criterion.In this paper, the sparse transform basis, observation matrix and reconstruction algorithm of key technologies in compressed sensing algorithm are studied. Different sparse transformation basis and different observation matrix and reconstruction algorithm are simulated in MATLAB.Through different combinations,the simulation results show that the discrete wavelet basis,partial Fourier transform matrix and subspace matching tracking have good effect on image reconstruction under the same conditions.