InSAR Phase Unwrapping

Synthetic aperture radar (SAR) interferometry is a greatly important technique in the generation of practical digital elevation models (DEM's). Given a complex-valued interferogram created from two registered complex SAR images of the same scene, the terrain height is related to the absolute phase of the interferogram, whereas the measured phase is known only modulo 2Ï€ , i.e. "wrapped." Two-dimensional phase unwrapping, retrieving the absolute phase from the measured (wrapped) phase, is a key step.2-D phase unwrapping is an ill-posed inverse problem in general. We can only search for approximate solutions satisfying additional constraints rather than exact solutions.Many phase unwrapping approaches for SAR interferometry have been proposed in recent years. They can be divided mainly into three categories: path-following methods, mean-squares methods and network-flow methods. The three main approaches are reviewed in their own subsection. Their advantages and disadvantages are shown by experimental results carried out on simulated and real interferometric SAR data. Then we propose a new 2-D phase unwrapping approach using the branch-cut and finite element method. The application of our approach generally avoids possible error propagation from the unreliable regions to the reliable one. Moreover, thanks to the use of the good boundary values evaluated from reliable data, the total solution within all the isolated regions is improved. A simulated experiment demonstrates the effectiveness of the proposed approach.