Removing shadows from Google Earth satellite images

Abstract

June 2005, Google has released its geographic search tool "Google earth", a new application that combines local search with satellite images and maps from around the globe. It is designed to make every person owned a computer easily "fly" to aerial views of many locations on the planet. However, just as ordinary satellite images, there inevitably exist shadows in it, made some ground objects obscure, even unidentifiable. According to the basic thinking of Radiative Transfer Theory, this paper built a image shadow removal model, which using the Radiative Transfer Theory combined with preknowledge to compensate the lost shadow area information. The results shows: shadows in images were successfully removed and the target objects were returned to their original scenes.