High-rise building feature extraction using high resolution spotlight TanDEM-X data

The inherent spatial scales of buildings are determined by the typical floor height of three meter and the distance of adjacent windows. With meter-resolution images delivered by modern SAR satellites like TerraSAR-X and TanDEM-X, it is now possible to map urban areas from space in very high level of detail using many advanced interferometric techniques such as PSI and TomoSAR processing, whereas these multi-pass interferometric techniques are based on a great number of images. In the paper we propose a workflow, taking advantages of the single-pass high resolution InSAR data free of motion and temporal decorrelation, to extract features of high rise building. These features include building masks, orientations and especially, a new feature - the iso-height lines. Incorporating the extracted building features as prior knowledge into the TomoSAR inversion may reduce the required number of images. The proposed approach is validated using a high resolution spotlight TanDEM-X pair over a high rise area in Las Vegas.