Geometric Data from Images

Abstract

Chapter 10 showed how to geo-reference single images and how the position of features may be determined from those images. There are several limitations to these techniques if “high” precision and accuracy is required for the position of features. Implicit in single image methodologies for obtaining height is the need to have an elevation model. But we know that height information already exists in an aerial photograph in the form of relief displacement. This chapter introduces the concepts of photogrammetry — the extraction of geometric information from images — and shows how it is used to extract more reliable position data from overlapping images and blocks of images. Thus it will lead to the geo-referencing of stereo (and multiple) images.