Nonlocal SAR Interferometric Phase Filtering Through Higher Order Singular Value Decomposition

Interferometric phase filtering is an indispensable step to obtain accurate measurement of digital elevation model and surface displacement. In the case of low-correlation or complicated topography, traditional phase filtering methods fail in balancing noise elimination and phase preservation, which leads to inaccurate interferometric phase. A new nonlocal interferometric phase filtering method taking advantage of higher order singular value decomposition (HOSVD) is proposed in this letter. For each pixel of the interferometric phase, a 3-D data array is established, and shrinkage is applied after HOSVD. A Wiener filter is used to improve the denoising performance in the end. Simulated and real data are employed to validate that the proposed method outperforms other traditional methods and some of the state-of-the-art nonlocal methods.