Study On Extracting And Phase Reconstructi On Of Low-coherence Region With Airborne InterferometricSAR Images

Synthetic Aperture Radar Interferometry (InSAR) is a milestone in the development of Synthetic Aperture Radar (SAR); it is the extension and the further development of SAR function. With high resolution spaceborne SAR system, such as RadarSat2, Cosmo-SkyMed, TerraSAR-X launched successively, producing DEM with high resolution InSAR is getting a hot and emphasis research topic in InSAR. The research on reducing the impact of low-coherence region in high-resolution InSAR plays a pivotal role in improving the effect of interferogram unwrapping results and obtaining high-precision DEM data. In addition, it makes great sense in mapping the complex terrain, special landscape area.In the special case of topography, the large area of low coherence area result in unwrapping failure with the process of Interfere and not be able to obtain high accuracy DEM and restricting the use of InSAR. To solve this problem, Paper presented a approach to extracting the mask map of low coherence area based on the coherence and Intensity information of SAR interferometry datas, and a method which using linear interpolation to reconstruct the phase of low-coherence region for getting the complete interferogram. The method successfully resolved the problem of unwrapping failure caused by the large area of low coherence area.Begin with SAR imaging technique, the SAR imaging principle, SAR image feature, SAR image geometric characters are introduced firstly in this thesis, as well as the basic principle of InSAR. Then the method of producing the intensity image, interferogram and coherence map was explained. Particularly discusses the methods and procedures to solve the two major problems which were extracting low-coherence areas and reconstruction phase with the three above-mentioned images. And two pair airborne interference SLC images which were typical characteristics of low-coherence area terrain were employed in the experiment. The results were unwrapped and the unwrapping results show that the method discussed in this thesis is able to remove the unwrapping interference of the whole images caused by large area of low coherence area and satisfy the accuracy needed in the following InSAR processing steps.