Research On Interferometric Phase Processing Of InSAR

This thesis first demonstrates the basic principle of the interferometric synthetic aperture radar signal processing, which is followed by a more detailed discussion on the key steps of the interferometric phase processing through simulation and analysis, including the complex image coregistration, flat effect removing, noise filtering of interferometric phase, phase unwrapping, and so on. New ideas and jobs we have done in filtering and unwrapping are also explained in this thesis.The main jobs of this thesis are:Chapter one first presents the principle of InSAR and the basic flow of InSAR signal processing.Chapter two introduces image coregistration and flat effect removing. Several coregistration methods are dissected and compared, which include correlation-based algorithm and maximum spectra algorithm. Following, the flat effect analysis and removal in InSAR Imaging are dissertated.Chapter three studies the problem of interferometric phase images filtering and estimation, and details spin filtering with curve windows along the fringe tangent orientation and an estimation method based on the technique of joint space-image subspace projection. This thesis proposes a method based on Moving Least-Squares to generate the fringe tangent orientation and gives an improving joint space-image subspace projection method with curve windows along the fringe tangent orientation through the combination of the fringe tangent orientation and the estimation method. Then we prove their advantages through comparing with the traditional filtering methods.Chapter four illustrates the problems of two dimensional phase unwrapping of single and multi-baseline, and again proves the importance of filtering in improving the quality of interferometric phase images and the unwrapping result through comparing the corresponding noisy and filtering phase unwrapping results.Finally, we suggest the future experiments in filtering and phase unwrapping according to the simulation results.