Automatic Terrain Characterization and Feature Identification in FOPEN SAR Imagery

Interferometric SAR techniques have been successfully demonstrated to be able to locate pixel scattering centers in three dimensions: however the relative location within a scattering volume does not provide further information that can be used to makemeasurements. Similarly polarimetric techniques allow scattering behavior to be determined averaged over the entire volume. However, by the use of a fully polarimetric and interferometric SAR system, there is the potential to determine the differentscattering mechanisms within a volume as well as their locations (x, y, z). During our phase I effort we successfully applied polarimetric interferometry techniques to FOPEN SAR imagery to estimate the tree heights and bald earth topography over forestedareas. During phase II, these algorithms will be extended to allow for estimation of the canopy density as well as a more reliable estimate of the current parameters. This will be accomplished through the development of improved vegetation models, the useof multifrequency information present in the UWB UHF data, and the use of multiple repeat pass baselines. A set of software modules for performing terrain characterization using data from the DARPA FOPEN sensor will be developed for use in batch mode aswell as interactively by a user.