GIS-supported iterative filtering approach for building height estimation from InSAR data

The tandem configuration of TerraSAR-X and TanDEM-X provides high resolution and high coherence spaceborne single-pass InSAR data for the first time, which opens up the floor for new applications. These data are especially attractive for 3D information extraction - a topic also investigated here. The presented new workflow on building height extraction contains processing steps of smart phase filtering, interferometric phase simulation, and their assessment to obtain the building height. Due to the fact that interferometric phase interpretation suffers from considerable noise, an appropriate filtering is mandatory to enhance building reconstruction. In previous work, we compared conventional patch-like filters, orientation mask filters, and our new adaptive orientation mask filter. These first results showed in particular a high potential of GIS-supported filters, and emphasize some further investigations. Hence, in this paper, we present an iterative approach using GIS-supported filters and include local phase unwrapping into the filtering. Assessment results between filtered and simulated phases as well as extracted building heights are presented and discussed for single pass airborne and repeat pass spaceborne data. Eventually, the evaluation of final building height is supported by ground truth LIDAR data