Enhanced Three-Dimensional Reconstruction Method for Polarimetric Coherent Optimal-Based Tomographic SAR

This paper introduced a three-dimensional reconstruction approach for Synthetic Aperture Radar (SAR) based on polarization-coherent optimization. It includes two-dimensional imaging of full-polarization tomographic SAR data, along with registration and amplitude-phase correction of the resultant images. Through the application of a polarization-coherent optimal strategy, a set of coherent-optimal two-dimensional complex images is derived for the full-polarization multi-channel tomographic SAR. Subsequently, a three-dimensional reconstruction is executed using a compressive sensing technique, yielding a refined three-dimensional point cloud of the target region.This innovation addresses challenges encountered in traditional single-polarization tomographic SAR systems, including issues such as low image coherence, subpar three-dimensional reconstruction quality, and inadequate detail in point cloud representation. By employing the polarization-coherent optimal method on two-dimensional complex images from multi-channel full-polarization tomographic SAR, a set of images with optimal coherence coefficients is obtained. This approach effectively enhances coherence among two-dimensional complex images of each channel in full-polarization TomoSAR, ultimately elevating the quality of three-dimensional imaging point clouds.The methodology involves initial two-dimensional imaging of full-polarization multi-channel tomographic SAR echo data, followed by registration and amplitude-phase correction of the obtained complex image set. The polarization-coherent optimal method is then applied to acquire a set of coherent-optimal two-dimensional complex images. These images are further processed using a compressive sensing-based three-dimensional reconstruction algorithm, yielding the original three-dimensional point cloud. Finally, post-processing steps such as point cloud filtering and coordinate transformation are applied to obtain a high-quality three-dimensional SAR image of the target area.The significance of this approach is demonstrated through its application in the field of Polarimetric TomoSAR (PolTomoSAR), a vital tool for acquiring three-dimensional information about urban structures. The proposed method addresses the challenges faced by PolTomoSAR, including the limited application of full-polarization data and the need to improve coherence among multi-channel complex images for effective three-dimensional reconstruction.Additionally, the effectiveness of the proposed method is validated through the acquisition and processing of real-world data from a fully polarized P-band TomoSAR radar system over the Dunhuang Quadrilateral residential area. The developed technique holds great potential in advancing the applications of polarimetric SAR data in three-dimensional imaging, urban mapping, city planning, and building monitoring.