Joint phase unwrapping for multi-frequency SAR Interferometry

Abstract

An important step in the interferometric processing chain is the phase unwrapping, which transforms the ambiguous phase (wrapped in the interval -/+ 180°) in an unambiguous phase. This unambiguous phase can then be used for the computation of the scene height model. Standard phase unwrapping methods work with data from a single frequency and a single baseline at a time. The unwrapping becomes extremely challenging for data with small height of ambiguity (HoA), which is usually the case when working with high frequencies and large baselines. The use of multichannel information can help to solve the ambiguities, improving the unwrapping results. However, for airborne repeat-pass InSAR data, the presence of uncompensated residual motion errors usually affects the efficiency of multibaseline approaches. For systems capable of acquiring data simultaneously in different frequency bands, as is the case of the F-SAR of DLR, multifrequency approaches can be explored to cope with the residual errors issue. In this thesis, different unwrapping algorithms are evaluated for the unwrapping of data with small height of ambiguity. Furthermore, methods to jointly unwrap interferograms from multifrequency acquisitions in a robust manner are analyzed.