Dense Surface Models from Airborne and Spaceborne (Multi-)Stereo Images

Abstract

Digital surface models representing the heights of an area can be derived from two or more (multi-)stereo images of airborne or spaceborne sensors. A satellite stereo image of a current very-high-resolution satellite like WorldView or GeoEye with ground pixel sizes of about half a meter allows the derivation of surface models in the range of the same resolution. Such surface models are the Basis of many applications like the three-dimensional representation of the area, 3D change detection, calculation of volumes, detection of sight lines, or water flow and flooding. Satellite imagery covers large areas of about 400 square kilometers with ground resolution of about 1 meter, while airborne images from planes or drones usually cover only small areas but with higher resolution. In this chapter the basics of digital surface models are shown, and the actually best method for deriving dense digital surface models from airborne and spaceborne images is described. Some examples finally show the possible results.