Multi-Polarization SAR Joint 3-D Reconstruction of Forested Area Based on L2,1/2-Norm Regularization

Synthetic aperture radar tomography (TomoSAR) is a common measurement technique for three-dimensional (3-D) imaging of forested areas. It extends the synthetic aperture principle into the elevation direction, obtains the elevation scattering information of the observed target, and then realizes the 3-D reconstruction. In order to achieve high-resolution imaging, compressed sensing (CS) technology is widely used in TomoSAR. Given that many existing forested area datasets comprise multi-polarization modes, leveraging the complete potential of polarimetric data to achieve high-resolution 3-D recovery of the observed scene is essential. Therefore, this paper proposes a novel L2,1/2 -norm regularization based TomoSAR imaging method of the forested area, which jointly reconstructs the multi-polarization data and utilizes the correlation between each polarization to obtain more accurate 3-D reconstruction results. Experimental results based on real BioSAR 2008 L-band dataset are used to verify the proposed method.