高光谱遥感在轨替代定标模型与方法研究

项目名称： 高光谱遥感在轨替代定标模型与方法研究

项目编号： No.41271370

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 天文学、地球科学

项目作者： 陈正超

作者单位： 中国科学院遥感与数字地球研究所

项目金额： 75万元

中文摘要： 高光谱遥感技术是光学遥感发展的前沿，是推动遥感定量化应用的主力军。高光谱遥感器的光谱和辐射定标是其数据定量化应用的前提和基础。本项目面向高光谱遥感定量应用需求，针对高光谱遥感器光谱分辨率高、波段范围宽、连续成像等特点，系统研究高光谱遥感器在轨替代定标模型和方法，包括深入分析高光谱遥感定标特殊性、定标过程影响因子敏感性、以及光谱和辐射定标误差传递机制；研究高光谱遥感光谱-辐射一体化综合定标模型、特殊波段定标异常修正模型、光谱维整体定标以及最终定标结果真实性检验方法等；解决高光谱遥感定标过程中光谱与辐射缠绕问题、特殊波段定标异常问题、定标系数在光谱维离散和非一致等关键问题，并建立高光谱遥感在轨替代定标实验和观测方法规范。本项目针对高光谱遥感定标的特殊性展开系统性研究，从高光谱遥感定标机理、定标模型、定标实验等方面全面提高其定标精度和可靠性，这在国内尚属首次，具有较高学术创新价值和应用前景。

中文关键词： 高光谱遥感；光谱定标；辐射定标；替代定标；真实性检验

英文摘要： Hyper-spectral remote sensing technology is the scientific front of optical remote sensing, and the main promoting force for quantificational remote sensing and its application, which premise and foundation is the spectral and radiometric calibrations for hyper-spectral remote sensing sensors. According to the application requirements of the quantificational remote sensing and the characteristics of hyper-spectral remote sensing sensors (such as the high spectral resolution, wide spectral range, continuous imaging, and so on), the model and methods of vicarious calibration will be studied systematically, including the particularities of the calibration for hyper-spectral remote sensing sensors, sensitivities of the influencing factors during this calibration, and the transfer mechanism of errors between from radiometric and spectral calibration. In addition, we propose the models and methods, including the integration of radiometer and spectrum, the modification of the abnormal values derived from the calibration for the characteristic bands, the calibration over the whole spectral range, and the validation for all those results. We plan to solve some key problems during the calibration for hyper-spectral remote sensing sensors, such as the confusion between radiometer and spectrum, the abnormal values derived f

英文关键词： Hyperspectral remote sensing；Spectral calibration；Radiometric calibration；Vicarious calibration；validation