Radar-Coding and Geocoding Lookup Tables for the Fusion of GIS and SAR Data in Mountain Areas

Synthetic aperture radar (SAR) image orthorectification induces an important alteration of information due to the side-looking geometry of SAR acquisition. In high-relief areas, the difficulty is increased by the foldover effect: The images acquired with low incidence angles cannot be registered by a bijective transformation like polynomial transformations, as usually proposed by conventional software. In this letter, a simple and efficient method, fitted to geocoded data and SAR images, is introduced to propose a generic coregistration tool that takes SAR geometry into account without requiring the exact sensor model, specific parameters, and precise navigation data. This method is based on a simulated SAR image and on the computation of lookup tables (LUTs) that represent the coordinate transform from one geometry to the other. Results are presented on a high-relief area in the Alps, where satellite and airborne SAR images are used for glacier evolution monitoring. A comparison to other sensor-independent approaches has been performed, showing that the proposed approach performs better in mountain areas. The resulting LUTs allow merging SAR data with the georeferenced data, either in ground geometry by orthorectifying the SAR information or in radar geometry by the inverse transformation, namely, radar-coding data from a geographic information system, to improve the analysis of SAR images and the result interpretation.