Assessment of Tomographic SAR Processing Techniques for Forest Structure Estimation

Kurzfassung

The future SAR missions such as BIOMASS and Tandem-L will exploit the potential of Synthetic Aperture Radar Tomography to extract 3D forest structure information. Several algorithms can be applied for TomoSAR imaging. This paper analyses the performance of two non-parametric algorithms, Capon Beamforming and Compressive Sensing (CS), for forest structure applications, through a set of simulations reflecting different forest scenarios (distribution of canopy layers and temporal decorrealtion) and system parameters (baseline distribution, multilook, and noise). Results show that CS is in general more stable than Capon in front of system and scene variability, but may be more affected by artefacts.