EVALUATION OF SPACEBORNE AND AIRBORNE LINE SCANNER IMAGES USING A GENERIC ORTHO IMAGE PROCESSOR

Kurzfassung

One of the main processing steps of evaluating remote sensing data is the production of ortho images from the acquired raw scanner data. Since in most applications of thematic analysis, a rectified data set is required, there is a need for an effective – regarding time and accuracy - and generic – regarding different sensor systems – processor for performing this rectification for any desired sensor imagery. This is especially true when using the image data in Geographic Information Systems (GIS) and for data fusion and analysis with data from different sources or seasons. The accuracy of this rectification result is crucial for overlaying the data with existing data sets or maps and using them for evaluations like change detection, map updating a.o.. Triggered by the demand of an automatic processor embedded in the Data Information and Management System DIMS of DLR a generic ortho image tool was developed. The generic ortho image processor supports the production of ortho images from airborne and spaceborne digital line scanner images, as well as images from frame cameras. It is based on the Direct Georeferencing model using measurements of the exterior orientation of the sensor platform or sensor itself, the interior orientation (sensor parameters) and a digital elevation model. For the interior orientation models for pushbroom, whiskbroom and frame cameras as well as sensor calibration tables are supported. For the exterior orientation local level co-ordinate frames (navigation frame, orbital frame), earth centred earth fixed (ECEF) co-ordinate frames and generic mapping frames are supported. An approximate processing in a map projection is provided for airborne scanner data. The boresight misalignment matrix and the lever arm values are part of the functional model. Map projections are included in the processor. It also includes a link to the atmospheric correction processor ATCOR, which is also part of the automatic processing chain within DIMS. The ortho image processor is applied for different sensors like the spaceborne line scanners SPOT5 and Quickbird, and the airborne line scanners HyMap, ROSIS, DAIS and Daedalus. For the calibration of the boresight misalignment angles or attitude offset angles ground control information is used. The accuracy of the ortho images with and without ground control information is shown. For SPOT5 images with an absolute location accuracy of 1 to 2 pixels using only the metadata delivered by the image provider, few ground control points (about 2-4) are sufficient to reach horizontal accuracy in the sub-pixel range. For Quickbird a comparison example between ortho images produced with RPC (Rational Polynomial Coefficients) and DG (Direct Georeferencing) is shown. A series of ground control points at the airport base DLR Oberpfaffenhofen and surrounding area serves as geometric calibration field for the determination of the boresight misalignment angles of the airborne scanners, which are used within different flight campaigns. The achieved accuracy for the geometric calibrated airborne systems is demonstrated.