Geometric Modelling and Photogrammetric Processing of High- Resolution Satellite Imagery

The latest generation of high-resolution commercial imaging satellites, such as IKONOS and QuickBird, has opened a new era of earth observation and digital mapping. This paper presents the geometric modeling principles and photogrammetric processing methods involved in high-precision mapping using stereo IKONOS and QuickBird images. First, the imaging geometry and systematic errors in the Rational Function-based sensor model are described. Then the results of a comparison study of IKONOS and QuickBird geopositioning accuracy improvement in which different adjustment models, as well as different number and configuration of ground control points, are presented. Results indicate that a simple adjustment model (e.g., Affine or Scale & Translation) is effective for elimination of the systematic errors found in vendor-provided RFCs (Rational Function Coefficients) and for improvement of 3D geopositioning accuracies to a 1-2m level for IKONOS images and a 0.6-1m level for QuickBird images. For coastal mapping purposes, a semi-automatic 3D shoreline extraction method is proposed. In this method, a 2D shoreline is extracted by manual digitizing on one QuickBird image; then corresponding shoreline points on the other image of the stereo pair are automatically extracted by image matching. The 3D shoreline is computed using photogrammetric triangulation with the improved geometric model.