Techniques for information extraction from radar signals have been assessed in view of remote sensing applications in forestry and land use. Several important directions were pursued including:
basic interferometric techniques;
advanced interferometric techniques for the extraction of geophysical parameters in forestry applications;
experimental characterization of synthetic aperture radar (SAR) spatial statistics as a function of several parameters, in particular study of the dependency on polarization and introduction of the concept of polarimetric texture signature;
modelling and experimental characterization of terrain relief effects on the polarimetric radar signal;
study of the dielectric properties of complex targets, from the microscopic physics point of view, and from the macroscopic scattering theory. Much effort has been devoted to the development of SAR techniques in remote sensing applications projects including:
preprocessing using a speckle filter and averager, classified and georeferencing geocoding packages;
Gamma Gamma MAP filtering, averaging, automatic edge detection, single date classification, multitemporal classification;
a study on the phenological interpretation of changes in radar backscattering.
An investigation was started to develop a technique based on multipass SAR interferometry to extract biophysical information about the temperate forest. The interferometric technique provides high resolution topographic data, which are useful for calibration and direct measurement of tree height, and yield information on the 3-dimensional structure of the canopy. These cannot be measured by any other remote sensing technique. Results achieved include:
review of literature on scattering models;
development of improved algorithms to generate height maps from earth remote sensing satellite number 1 (ERS-1) interferometric data sets;
development of algorithms and software to generate slopes maps;
validation of the algorithms using ERS-1 data sets.