间接型光谱成像仪相对光谱响应函数标定方法

Calibration method of relative spectral response function of indirect imaging spectrometer

间接型光谱成像仪通常采用面阵探测器作为光电转换器件,探测器像元间光谱响应的不一致性会导致采集到的目标光谱失真,所以标定和修正像元间光谱响应的不一致性是提升间接型光谱成像仪光谱辐射测量精度的重要手段.本文以干涉光谱成像仪为例,分析了系统像元间相对光谱响应不一致对目标光谱辐射测量准确性的影响,提出了基于傅里叶变换调制定标源的间接型光谱成像仪全系统相对光谱响应函数测量方法,并建立了相对光谱响应函数标定的数理模型.仿真分析结果表明,理想无噪声时,像元间1%的相对光谱响应不一致性会对复原光谱造成1.02%的相对误差,经过相对光谱响应校正后,不同行复原光谱的相对误差降至0.08%.最后,仿真分析了相对光谱响应不一致性在不同光谱信噪比下的校正效果.该方法可提高间接型光谱成像仪光谱测量的准确性和一致性.

In imaging spectrometers,area array detectors are usually used as photoelectric conversion devices,but the inconsistency of the spectral response among pixels can distort the collected target spectra.To improve the spectral radiometric accuracy of imaging spectrometers,calibrating and correcting the inconsistency of the spectral response among pixels is essential.The signal received by each pixel of area array detector of the indirect imaging spectrometer is usually the superposition of the target multi-spectral radiation signals or fullspectral radiation signals.Therefore,its relative spectral radiometric calibration requires measuring the spectral response of each pixel at different wavelengths on the array detector.Under the ideal conditions,the response values of each pixel in the area array detector are different,so the indirect imaging spectrometer cannot simply calibrate the relative spectral response(RSR)function between pixels by using the method of“monochromator+integrating sphere”.In this work,taking the interferometric imaging spectrometer for example,we analyze the influence of the inconsistency of the RSR among pixels on the target spectral radiation measurement accuracy,and propose a system-level RSR function measurement method for the indirect imaging spectrometer based on the Fourier transform modulation calibration source.In addition,we establish a mathematical model for calibrating the RSR function,and provide guidelines for selecting calibration system parameters such as light source,spectral resolution,and OPD sampling interval.The simulation results show that under the ideal noisefree condition,the 1% spectral response inconsistency among pixels results in a relative error of 1.02% to the recovered spectra.After RSR correction,the relative error of the recovered spectra of different rows decreases to 0.08%.Furthermore,in this work we simulate and analyse the influence of spectral signal-to-noise ratio on the calibration accuracy of the RSR function,and point out that increasing the brightness of the calibration light source,extending exposure time,and combining multi-frame interferograms can enhance RSR function calibration accuracy in practical applications.The research result can provide a theoretical basis for realizing the relative spectral radiometric calibration of indirect imaging spectrometer,which is of great significance in promoting quantitative spectral remote sensing.