一种改进的光谱图像反演地表目标遥感反射率的方法

Improved Method for Retrieving Remote Sensing Reflectance of Ground Targets from a Spectral Image

提出一种改进的反演地表目标遥感反射率的方法.利用光谱图像中的太阳耀光估计当时气象条件下的大气透过率,分析了天空辐射和色散现象对反演的影响,据此构建遥感反演模型,计算出地表目标的遥感反射率.由于消除了天空漫射辐射带来的额外影响使得可见光波段的反演准确度大为提高,此外将菲涅尔反射率表示为波长的函数进行计算使反演光谱在整个波段范围内能更好地匹配实验室光谱.实验结果表明,改进算法计算的反射率能更好地匹配实验室光谱,具有更高的反演准确度.

An improved ground targets remote sensing reflectance retrieve method was presented. According to the sun-glint from a spectral image, the atmospheric transmittance under the weather condition was estimated. The effects of the sky radiance and dispersion phenomenon were discussed, then a remote sensing retrieval model was developed,and the ground reflectance was computed. The retrieval accuracy in visible bands was improved greatly because of the eliminating of extra effect which was brought by sky radiance. And,a better match with laboratory spectra was gained by retrieval spectral in the whole bands using the method of expressing the Fresnel reflectance as the function of wavelength. The experimental results prove that the reflectance of the improved algorithm matches with the laboratory spectra remarkably well ,and acquires higher accuracy.