Interferometric Sar Imaging Algorithm Research

Interferometric Synthetic Aperture Radar (In-SAR) is nowadays a focus ofresearch in the remote sensing area and has attracted many researchers' interests.This is due to the fact that it has many advantages over other imaging methods, suchas large coverage, high spatial resolution, high elevation accuracy, and the capabilityof working at all time and under all-weather conditions. It is the most effectivemethod to acquire the 3D Digital Elevation Models (DEMs).Based on this background and the practical requirement of our institute, weinvestigated the algorithms of In-SAR imaging both theoretically and practically,brough forward some new method and got promising results using simulated andreal data. Beginning with a brief ittroduction to the rationale and the basicarithmetic of In-SAR, this dissertation emphasizes on the two most importantaspects in improving the quality of reconstructed image: phase noise filtering andphase unwrapping.A good interferometric image is a prerequisite of getting good DEM map.Hence proper phase noise filtering need to be done before phase unwrapping. A newmethod is put forward and testified with real data. This method is proved to befeasible in that it can greatly improve the quality of the interferometric image.Phase unwrapping is a critical step of In-SAR imaging. Based on previous workdone by an old studen, we did some further research on phase unwrapping problemby multigrid method. The work is mainly focused on weighted multigrid method.Furthermore, not only simulated data but also authentic data are used in ourexperiments, and many meaningful results are obtained.