1 May 2023 in IEEE Transactions on Geoscience and Remote Sensing
Due to the advantage of flexible and rapid deployment, unmanned aerial vehicle (UAV) based synthetic aperture radar (SAR) tomography (TomoSAR) is a promising technology in three-dimensional (3D) urban mapping.The long baseline is indispensable for P-band SAR systems to achieve high elevation resolution. It will introduce two problems.On the one hand, the unavoidable spatial decorrelation brings serious phase noise and sidelobes in 3D imaging.On the other hand, the noticeable image distortion fails the image registration and the TomoSAR data stack (TDS) construction. Aiming at the above problems, this paper proposes a multi-master (MM) TomoSAR approach via three main contributions. First, the traditional TomoSAR signal model is extended to the MM case to improve the number of baselines and the average image coherence of the TDS, and suppress the sidelobes. Second, a short-baseline-recursion image registration method is proposed to achieve high-precision image registration. Third, a TDS optimization processing consisting of interferometric SAR (InSAR) phase screening and baseline sign reassignment is introduced. Moreover, a clustering-based outliers elimination method is also adopted to ensure the 3D imaging quality. Computer simulation and long-baseline P-band UAV-SAR experiment validate the proposed approach.
13 March 2025 in ISPRS Journal of Photogrammetry and Remote Sensing