Polarimetric Sar Tomography Using `2,1 Mixed Norm Sparse Reconstruction Method

The growing interest of Radar community in retrieving the 3D reflectivity map makes both polarimetric SAR interferometry and SAR tomography hot topics in recent years. It is expected that combining these two techniques would provide much better discriminating ability for scatterers lying in the same pixel. Generally, this is about reconstruction of scattering profiles from limited and irregular polarimetric measurements. As an emerging technique, Compressive Sensing (CS) provides a powerful tool to achieve the purpose. In this paper, we propose a `2,1 mixed norm sparse reconstruction method for jointly processing multibaseline PolInSAR data based on multiple measurement vector compressive sensing (MMV-CS) model, and also address the signal leakage problem with MMV-CS inversion by presenting a window based iterative algorithm. The results obtained by processing simulated data show that the proposed method possesses superior performance advantage over existing methods.