A Model Based Radiative Transfer Algorithm to Correct Remotely Sensed Image in Mountainous Area
基于模型的山地遥感图象辐射订正方法

Topographic effects are the main obstacles to further application of remotely sensed images. Due to the fact of that different slopes and aspects show very different brightness in images even if the ground cover is the same type, the general classification methods may give wrong results. How to remove the topographic effects is a necessary step in quantitative remote sensing. However, topographic effects are very complex and difficult to be removed successfully. Based on detailed analysis of illumination sources that include solar direct irradiance, sky diffuse irradiance and adjacent reflectance, this paper presents a reflectance retrieval model to remove the topographic effects in satellite images and retrieve the information contained in shadow areas. From the corrected TM images, it is found that the topographic effects are removed effectively. This just illustrates that this method can provide more information for post processing of the remotely sensed images.