Large-Baseline TomoSAR Imaging Using Range–Elevation Spectrum Scaling

High-resolution tomographic synthetic aperture radar (TomoSAR) 3-D imaging requires a large baseline in elevation to achieve a sufficiently good resolution along the elevation axis. However, the large elevation baseline invalidates the traditional TomoSAR processing framework, which follows a sequence of 2-D imaging, coregistration, and finally elevation imaging. The core difficulty lies in the space-variant slant-range history from the baselines to the targets at different elevations, resulting in those targets being unable to be aligned after coregistration. The error introduced by the coregistration causes a range–elevation migration across multiple 2-D images, yielding defocusing in the 3-D image. In order to address this problem, a range–elevation spectrum scaling (RESS) method is proposed, correcting this migration through decoupling the range–elevation frequency components in the spectrum. First, the range–elevation migration is analyzed across the 2-D images according to the mapping from 3-D position onto 2-D images of the targets in the scenario. Second, the phase history and range–elevation spectrum are derived, and then the scaling method for migration correction is applied. After elevation focusing on the migration-corrected data, the 3-D image is distorted due to the incident angle variance of the scene, and finally, the space-variant distortion of the proposed RESS method is analyzed and corrected. The simulations and real unmanned aerial vehicle (UAV) experiments verify the effectiveness of the RESS.