Advanced polarimetric optimization for DInSAR applications with Ground-Based SAR

摘要

In this paper, the applications of advanced polarimetric optimization techniques for Differential Interferometric SAR (DInSAR) pixel selection are explored in the framework of deformation map retrieval in landslides. Unfortunately, most of high mountain landslides occur on vegetated areas with a poor density of coherent scatterers that with a fast decorrelation at X-band. The objective of the techniques proposed in this paper is to increase the number of pixels with good quality, and thus the robustness of DInSAR algorithms, by exploiting the larger density of coherent scatterers present in polarimetric SAR data. Using Ground-Based SAR zero-baseline fully-polarimetric data sets, the relationship between optimum coherences and its corresponding phase in terms of DInSAR application is carefully analyzed. In the paper, different coherence optimization methods available in the literature are studied. The main advantages and drawbacks of the different approaches are drawn.