在压缩传感技术应用中,根据稀疏基底选择抽样模型对重构结果影响很大。在傅里叶空间中,极坐标星形抽样和随机抽样的重构效果差异巨大,应用傅里叶光学理论对傅里叶空间的频谱分布进行分析,从理论上解释了原因,并且据此提出稀疏基底和抽样模型的匹配情况会影响重构效果。在小波空间中,进行了均匀抽样和随机抽样的对比重构实验,发现后者的重构效果更好,并确定了根据稀疏基底选择合适抽样模型的可行性,为实际应用中降低抽样率,提高重构效果提供了方法依据。 In the application of compression sensing technology, the selection of sampling model based on sparse base has a great impact on the reconstruction results. In Fourier space, there is a huge difference in reconstruction effect between polar star sampling and random sampling. The Fourier spectral theory is used to analyze the spectral distribution of Fourier space, which explains the reason theoretically and proposes accordingly The matching of sparse base and sampling model will affect the reconstruction effect. In the wavelet space, the uniform reconstruction and random sampling experiments were carried out. The reconstruction effect of the latter is better and the feasibility of selecting a suitable sampling model based on the sparse substrate is determined. In order to reduce the sampling rate in practice, Improve the reconstruction results provide a method basis.