单边干涉数据光谱复原的关键问题分析

Key Problem Analysis of Spectral Reconstruction of Single-sided Interferential Data

针对干涉式傅里叶变换光谱仪中干涉信号的单边过零采样和实时快速数据处理要求,该文对光谱复原中的实时数据处理关键问题进行分析。在实验中以迈克尔逊干涉仪为核心,以632.8nm的激光为参考光源,以3.5～5.0μm的大气窗口为被测对象,产生干涉图;通过HgCdTe点探测器对干涉图进行光电转换,获取干涉数据。基于Matlab软件对干涉数据进行预处理、最小采样数据点确定、切趾、相位校正、频谱标定等光谱复原中数据处理的关键技术和问题进行分析和仿真分析。实验表明,大气窗口的4 096点的干涉数据采用小双边干涉数据量为512点,所有干涉数据采用非对称三角切趾的Mertz法相位校正,傅里叶变换、线性标定后能快速复原目标光谱,且光谱分辨率达到4cm-1。

For the requirements of the discrete sampling of the single-side interferogram beyond zero path difference in Fourier spectroscopy and the need of the fast real-time data processing,the key problem of real-time data processing in the spectrical reconstruction has been analyzed in this paper.In the experiment,Michelson interferometer is applied to acquire the interference fringes,the 3.5-5 μm atmospheric window is the measured object and the 632.8 nm laser is used as the referenced light source.The interferential signals are obtained by the dot HgCdTe detector.The interferential data is predisposed,apodized,phase corrected and calibrated to rebuild the spectrums based on the Matlab software.The experiment indicates that the interferential data of the atmospheric window of 4 096 points is phase corrected by the method of Mertz in which the amount of the little double-sided data is 512 points and apodized by the symmetrical triangle function,the total data is apodized by the asymmetrical triangle function,Fourier transformed and calibrated to quickly and accurately rebuild the spectrum which is high resolution and the spectrial resolution is about 4 cm^-1.