辐射度定标是时间调制型FTIR数据处理中非常关键的一个环节,定标的好坏直接影响着其在应用中性能的优劣。根据光谱仪响应函数(线性或非线性)的不同,辐射度定标方法可分为线性定标和非线性定标;根据定标中采用的点数的不同可分为两点定标和多点定标。首先用MATLAB对光谱仪采集的数据进行线性度分析与仿真,然后用C++编程分别实现线性定标和非线性定标。实验结果为两点法的误差为0.1118,抛物线法的误差为0.1684,四点线性的误差为0.0599。结果表明多点线性的定标方法效果最好。采用四点线性的方法进行定标将大大提升光谱的准确度,为后面的光谱识别工作打好基础。 Radiometric calibration is a key step in the time-modulated FTIR data processing. The calibration has a direct impact on the performance of the application. According to the spectrometer response function (linear or nonlinear), the radiometric calibration method can be divided into linear calibration and nonlinear calibration; according to the calibration points used in the different points can be divided into two-point calibration and multi-point set Standard. Firstly, the linearity analysis and simulation of the data collected by the spectrometer are carried out by MATLAB, and then the linear calibration and the nonlinear calibration are respectively implemented in C ++ programming. The experimental result shows that the error of two-point method is 0.1118, the error of parabola method is 0.1684, and the error of four-point linearity is 0.0599. The results show that the multi-point linear calibration method works best. Using the four-point linear method for scaling will greatly enhance the accuracy of the spectrum and lay a good foundation for the subsequent spectral recognition.