A shadow processing method of high spatial resolution remote sensing image

The shadows on the High Spatial Resolution Remote Sensing Image (HSRRSI) have negative effects and building shadows of large area on HSRRSI (up to 30% in some cases) are one of the biggest obstructions. This paper introduces a semi-automated method for shadow removal in high spatial resolution satellite image. At first, the image consists of three parts: shadow regions, non-shadowed regions and transition regions through semi-automated human-computer interaction. Then the shadow regions are equalized to the non-shadowed regions through histogram transformation. At last, post-processing on the shadow regions edges gets the result of a smooth image without shadows. The experiment results show that this method is valid.