A Nonlocal Estimation Method of Optimal Coherence for PolInSAR Data

The traditional coherence optimization algorithm has been successfully applied to polarimetric interferometric synthetic aperture radar (PolInSAR) data, but often suffers from spatial discontinuity. This paper introduces an innovative approach for optimizing the scattering mechanism vector(SMV) through nonlocal estimations. The proposed method constructs a pixel-level optimization model by leveraging patch-based nonlocal similarity, effectively enhancing the accuracy of SMV estimation while preserving high resolution. Both simulated and observed ESAR data are employed to evaluate the performance of various methods. Experimental results confirm the advantages of the proposed approach in detail preservation and phase recovery.