Geo-Correction Of High-Resolution Satellite Imagery And Its Accuracy Measurement

Dealing with various spatial information, the extraction of spatial information from High-Resolution Satellite Imagery(HRSI) is a hot topic of Digital Photogrammetry, GIS and other fields. Aiming at the linear CCD push-broom of high resolution satellite imagery, theory and application on imagery geometry correction is discussed in the paper. The main work are as follows,1. Because ephemeris data of commercial High-Resolution Satellites are withheld from users, alternative empirical models should be used. The Point Feature Based Empirical Transformation Models are widely applied for imagery rectification, such as Rational Function Models and 3D-Affine transformation model.2. Compared with point features, linear features have plenty of advantage. As control features, linear features are successfully exploited for image rectification in Photogrammetry. And Generalized Point Photogrammetry theory is applied for linear features on CCD push-broom imagery. Therefore, a new mode (named the Line Based Transformation Model LBTM) is developed for HRSI rectification.3. There is two main factor of HRSI geometry correction. Besides mathematical model, Ground Control Points (GCPs) has played a big role on rectification. The paper analyses several aspects of GCPs, such as its resource, precision, quantity and distribution. Furthermore, the paper has deduced the relation between the accuracy of an arbitrary point and the accuracy of control points to show the effect of GCPs.4. According to such relation, the inverse distance power method is used to give the spatial interpolation to reveal the image rectification accuracy. Based on above, the visualizations of accuracy are used to give the detail description of the quality of image rectification.