Research On Data Simulation And Phase Unwrapping For Multi-baseline Insar

Multi-baseline Interferometric Synthetic Aperture Radar (InSAR) has been proposed since 1990s to solve the problems in deriving accurate DEM for rugged areas with traditional single baseline InSAR. Nowadays, multi-baseline InSAR stands for a new direction in InSAR development. The research of multi-baseline InSAR in China is still at the beginning, and so the key techniques of multi-baseline InSAR were studied in the thesis. The simulation methods of multi-baseline InSAR data were designed and validated with different DEMs. The means of multi-baseline InSAR phase unwrapping were studied based on the simulated multi-baseline InSAR data. The main work and innovations were listed as follows:1. The development and applications of InSAR were introduced. And then the disadvantages of traditional single baseline InSAR were analyzed. The contents and their meaning of the dissertation were introduced based on the presentation of the background, the status and the key techniques of multi-baseline InSAR.2. Pixel resolutions and pixel spacing of SAR were compared. The reference frames, the characters of SAR images and the imaging models were recommended. The fundamental of deriving DEM from InSAR was analyzed, and the modes, the advantages and the processing flowchart of multi-baseline InSAR were discussed.3. The simulation of multi-baseline InSAR data was researched. Both the direct simulation method and the indirect simulation method were designed based on the F.Leberl imaging model. The interferogram simulation experiments were done with different DEMs. According to the simulation results, the different simulation methods and therir efficiency were compared. And the simulation of SAR intensity images was finished. All the simulation experiments could provid data for multi-baseline InSAR phase unwrapping study.4. Chinese Remainder Theory (CRT) was introduced into multi-baseline InSAR phase unwrapping. Phase unwrapping method based on Chinese Remainder Theory for multi-baseline InSAR was first devised. The approach of inverting modulars non-prime to each other to modulars prime to each other was proposed, which broke through the restrict of special vertical baselines'rates. The proposed method could make full use of the relationship of vertical baselines, extend the un-fuzzy interval of wrapped phases and solve the phase unwrapping problems with sub-sampling and aliasing phases.5. The multi-baseline InSAR phase unwrapping method based on difference filtering was proposed, which used the idea of Zero-Intermediate Frequency Vector Filtering for reference. In this method, the filtering and phase unwrapping of interferogram with shortest baseline were done firstly, and then the reference interferogram for longer baseline was calculated using the unwrapped interferogram with shorter baseline and their relationship. The differential interferogram was derived from interferogram with longer baseline and its reference interferogram. The differential interferogram was filtered and unwrapped one by one, until the interferogram with longest baseline was unwrapped. The method could use the unwrapped interferogram with shorter baseline to unwrap the interferogram with longer baseline, and solve the difficulty in phase unwrapping for sub-sampling areas with longer baseline. It was robust and less affected by noise.6. Based on the simulated interferograms, experiments of phase unwrapping with the proposed two methods for multi-baseline InSAR were finished. The phase unwrapping results were addressed, and these two methods were compared and analyzed for their applying cases and characters.A software for multi-baseline data simulation and phase unwrapping was coded with the study. The research could promote the development of multi-baseline InSAR.