基于光谱采样率的反卷积算法分析

Analysis of Deconvolution Algorithm Based on Spectral Sampling Rate

反卷积方法是提高光谱仪分辨率的重要手段。采用空域迭代反卷积和频域维纳滤波对多纵模激光器光谱进行数据仿真,并在不同光谱仪采样率条件下,比较了迭代反卷积和维纳滤波结果。仿真结果表明,迭代反卷积和维纳滤波可以有效消除光谱仪仪器响应函数引起的光谱展宽,提高光谱仪分辨率。在光谱仪采样率低的情况下,迭代反卷积的分辨率增强效果优于维纳滤波。随着采样率的增加,维纳滤波的误差小于迭代反卷积。实验分别测量了单纵模和多纵模632.8nm He-Ne激光器光谱,并对测量结果进行反卷积处理。结果表明,低分辨率光谱仪测量的激光器光谱经反卷积处理后与高分辨率光谱仪直接测量结果一致。

Deconvolution technique is currently available for application in spectral analysis to enhance the resolution of spectra,especially the spectra with multi-longitude modes.Iterative deconvolution and Wiener filtering techniques are discussed.The root-mean-square errors of iterative deconvolution and Wiener filtering with different spectral sampling rates are compared via computer simulation.The results indicate that the resolution of the observed spectrum is improved by using iterative deconvolution and Wiener filtering.The iterative deconvolution works better than Wiener filter at low spectral sampling rate.Experimentally,the spectra of two He-Ne lasers centered at 632.8 nm with single mode and multi-mode are deconvolved.The number of longitude modes,longitude spaces and spectral bandwidths of the He-Ne lasers are obtained by using iterative deconvolution.The deconvolved spectra of He-Ne lasers measured with a low-resolution spectrometer are in agreement with the spectra measured directly with a high-resolution spectrometer.