Sherlock Supports The Geocoding Of SAR Images

The geocoding of synthetic aperture radar (SAR) image data is performed by a transformation of a digital SAR im­ age to a georeferenced system. A way is to search for con­ trol points in the image and a corresponding digital elevation model (DEM). In the presented approach we avoid this time­ consuming search for control points by applying different fea­ ture detectors in DEM and SAR image in order to find features with the same characteristics. This process is supported by the knowledge based system SHERLOCK which manage the various methods applicable to the special circumstances. The predominant part of SHER­ LOCK is the knowledge base which controls the data driven processing under the (interactive) guidance of the human op­