针对GF-1遥感影像的基于影像与基于辐射传输模型的两种交叉定标方法比较

摘要

"高分一号"配置了4台16m分辨率多光谱宽幅(WFV)相机,组合观测幅宽达到800km.为了将其应用于定量遥感,需要对其进行精确的辐射定标.目前针对高分一号卫星有两种交叉定标方法,都在传统方法的基础上进行了改进.一种是基于影像的交叉定标方法(image-based),另一种是基于辐射传输模型和二向反射分布函数的交叉定标方法(RTM-BRDF).本文采用这两种方法对高分一号(GF-1)的4个相机进行辐射定标,并对这两种方法进行了对比分析,发现对于WFV2和WFV3这两个近似星下点成像相机,image-based方法可以得到精度较高的辐射定标系数,而对于WFV1和WFV4这两个非星下点成像相机来说,RTM-BRDF方法得到的定标系数精度较高.因此,最终将两种方法结合给出GF-1 4个相机最终的定标系数.%Four wide-field-of-view (WFV) instruments are on board the Gaofen-1 (or GF-1) satellite, providing a combined swath of ~800km.Before appling to quantitative remote sensing, precision radiometric calibration is needed.Currently, there are two cross-calibration methods.One is the traditional cross-calibration method (image-based), the other is based on radiative transfer model and bidirectional reflectance distribution function(RTM-BRDF).In this study, the two methods were used to cross-calibrate the WFVs of GF-1, and the comparisons were made at the same time.The verification based on satellite data and in situ measurements have shown that, for the two approximately nadir imaging camera WFV2 and WFV3, Image-based method could get higher precision radiometric calibration coefficients, while for two non-nadir imaging cameras WFV1 and WFV4, high precision calibration coefficients would be obtained by the RTM-BRDF method.Finally, the calibration coefficients of GF-1 was derived by the combination of the two methods.