S1TBX-SE-17 Terrain correction and inverse terrain correction based on saved transformations


The goal of terrain correction is to correct the topographic distortion in the SAR image caused by the ground elevation of the targets. Due to the topographic distortion, the ground target is imaged in the SAR image to a different pixel with both range and azimuth displacements. In the terrain correction procedure, a new map domain image is generated with all the ground target images placed at the right position. Basically, for each point in the DEM grid (northing, easting), its range and azimuth indices are retrieved with the help of DEM and the orbit state vectors of the source product. Then the pixel value is extracted from the source image and output in the DEM’s map geometry. Terrain correction is a transformation from the Range-Doppler grid in the SAR image domain to the DEM grid in the map domain. However, there are cases where the inverse transformation is needed, i.e. the transformation from DEM grid to the SAR image grid. With the availability of the transformation and the inverse transformation, the images can be transformed between both geometries with the correct extent and pixel spacing. In this project, a new operator will be created for the inverse terrain correction. Similar to the terrain correction operator, the inverse terrain correction operator also starts from the DEM grid. For each point in the DEM grid, the operator also retrieves its range and azimuth indices in the SAR image. Once the range and azimuth indices are available, the operator will then distribute the pixel value of the point target in the DEM grid to the corresponding pixels in the SAR image using a distribution technique.





Jun Lu


Lisa House

Fix versions