Research On Image Processing Of Spaceborned Interferometric Synthetic Aperture Radar

Synthetic aperture radar (SAR) is a coherent active microwave imaging method. It can observe the earth in all days and all weathers. It has been widely used in many fields, such as agriculture, forestry, geology, ocean, military, etc. Interferometric SAR (InSAR) is developed on the basis of SAR. It exploits the phase difference between two complex-valued SAR images acquired in the same area to generate its three dimensional elevation. InSAR has been used in mapping, geological detection and other fields. The research on a spaceborned SAR system and its application is one of the hotspots in the remote sensing all over the world.The study on image processing techniques for interferometric synthetic aperture radar will be the essential preparation and technique support for the wide applications of spaceborned InSAR. The basic principle of InSAR is given and the working mode, history and applications are systematically addressed at first.The data used in an interferometric SAR system are single looked complex-valued (SLC) images. The displaying effect of an SLC image is important to choose InSAR data and to match image initially. On the basis of comparing and analyzing the characteristics of the algorithms raised at present, the modified displaying algorithm based on modified nonlinear mapping is proposed and the displaying effect is verified by the simulation experiments with the spaceborned SAR data.The researching content covered by interferometric SAR is wide and deep. From the viewpoints of basic conceptions and common problems, the space geometry, height ambiguity, flatten phase and decorrelation sources are systemtically discussed or summarized. The common flow chart of signal processing in an InSAR system is given subsequently.The flatten effect is the specific phenomenon in an InSAR system. It increases the difficulty of the behaviour of signal processing. The algorithm based on spectrum estimation of an interferogram is presented to eliminate flatten phase. The errors of frequency estimation and their influence on phase compensation are systematically analyzed. The feasibility and validity of the proposed algorithm are verified by a series of simulation experiments.According to the flow chart mentioned above and beginning with the choice of interferomatric images, several key parts of signal processing are discussed systematically, such as image matching, interferogram generation, flatten effect elimination, interferogram denoising, phase unwrapping, baseline estimation and generating digital elevation model, etc. Several more robust algorithms are used to process the data come from RadarSat-1 with repeat orbits. Nearly all phases in the flow chart are realized in the dissertation.