Research Of InSAR Phase-Unwrapping

Two-dimensional phase unwrapping is considered as one of the most important procedures in the data processing and analysis of synthetic aperture radar interferometry (InSAR). Phase unwrapping is indispensable for the purpose of either creating digital elevation model (DEM) or extracting surface deformation information with the InSAR approach. The achieved accuracy level of DEM or deformation measurements is highly dependent on the accuracy in the unwrapped phases.At present, the existing phase-unwrapping algorithms can be grouped into two categories. The first category reconstructs the continuous phase surface by a path-based integrating operation, and its typical representatives are the branch-cut, mast-cut and Flynn algorithm. The second category determines the absolute phases using least squares solution, and its typical representatives are the DCT/FFT and PCG algorithm. The first category of algorithms accomplish phase unwrapping by integrating phase gradients between neighboring pixels in an interferogram, while the second category of algorithms conduct phase unwrapping by minimizing the sum of the square of gradient difference between the unwrapped and wrapped phase field.This thesis will first review the principles of several typical phase-unwrapping algorithms. These different algorithms are then evaluated by comparing their estimation accuracy and computation efficiency. By integrating the branch-cut and DCT/FFT algorithm, a new algorithm is finally extended and validated by experimental analysis.The experiments carried out in this thesis are divided into two parts.As the first part of the experiments, Chapter 5 treats comparisons and analyses for different phase-unwrapping algorithms on the computation time efficiency and the achievable accuracy by using the simulated phase data. As second part of the experiments, Chapter 6 will involve in the integration of two types of algorithms, i.e., branch-cut and least squares algorithms. Both the simulated phase data and the real phase data around Shanghai area are used for the validation of the integrated algorithm.