A Robust InSAR Phase Noise Reduction Method Based on Sliding Window Tensor Model

Interferometric synthetic aperture radar (InSAR) images are severely corrupted by noise in both magnitude and phase, and as a powerful tool for extracting geophysical parameters, it is of significantly important to recover the interferometric phase during InSAR signal processing. In this paper, a novel algorithm of robust sliding window decomposition for InSAR phase denoising is proposed with the achievement of improved noise reduction. In the scheme, a sliding window tensor model of interferogram is constructed by stacking the image patches. Due to the locally scattering mechanism similarity principle, a superior low-rank property of the tensor data can be achieved, which is also confirmed by using the numerical experimental analysis. Then, a robust tensor decomposition method is used for noise reduction and an alternating direction method of multipliers (ADMM) solution is applied to robustly and accurately restore the noise-free interferometric phase. As a result, the proposed algorithm performs well in effectively restoring noise-free interferometric phase. Last but not least, the experimental analysis using measured InSAR data has been demonstrated, and excellent results are obtained.