A general variational framework considering cast shadows for the topographic correction of remote sensing imagery

Topographic shadows are inevitable obstacles for the interpretation of remote sensing images covering rugged terrain. A general variational topographic correction (TC) framework is proposed in this paper by considering not only self shadows but also cast shadows. Cast shadows are first detected by integrating the radiometric and topographic features of the observed region. The cosine values of the incidence angles for the cast shadows are then corrected by the variational framework. The corrected incidence angles can be used in any traditional TC model to compensate for the shadows in mountainous regions. The proposed variational framework was utilized in eight different traditional TC models, and the results were compared with the traditional results. Images from two different regions were employed to test the framework. The results suggest that the proposed framework can raise the accuracy of shadow correction by both subjective and objective evaluations, owing to the correction of the cast shadows.