Assessment of geo-positioning capability of high resolution satellite imagery for densely populated high buildings in metropolitan areas

This paper analyzes the geo-positioning capability of highresolution satellite images in a very special metropolitan environment where within a relatively small region there are a large number of densely populated skyscrapers and high buildings. A pair of QuickBird stereo images covering downtown Shanghai were used for analysis. Multi-source data including GPS survey data, aerial images, and lidar data collected in the same area were used for Ground Control Points (GCPs) and Independent Check Points (ICPs) measurements and accuracy analysis. The vendor-provided Rational Function Model followed by a translation and scale correction model in image space was used to introduce the ground control and improve geo-positioning accuracy. The experimental results revealed that there is a clear dependency of geo-positioning accuracy on elevation where the GCPs are placed. Using the vendor-provided RFM and several GCPs across all the elevation ranges improves the overall geo-positioning accuracy.