讨论一种分解遥感数字图像的解析方法。其目的是将组成原始遥感的各分量(太阳直射光、天空散射光和大气亮度)逐点分离,并各自生成新的数字图像。由于这些图像是自然光各分量分别与地物和大气之间相互作用的产物,它们在地表辐射能量和大气环境状况的遥感定量反演研究及遥感图像的模式识别中有着具其他传统遥感图像无法具有的独特作用。

This paper discusses the analytical way of decomposing the remote sensing digital imagery. The method canseparate all the components (including the straight light, the sky scattering light and the brightness of the atmosphere)which form the original image information. Every component will generate a new digital image. These images are theresults of the interaction between all the components and the atmosphere, and objects, so, they have distinctive effect inthe remote sensing quantitative inversion searching for the earth's surface radiant energy and the state of environmentand pattern recognition of remote sensing pictures.