Abstract:

Synthetic aperture radar tomography (TomoSAR) applies the synthetic aperture principle into the three-dimensional imaging of the elevation direction. Compared with the traditional two-dimensional imaging, it increases the elevation information. Traditional spectrum estimation methods can be used for SAR tomography, but the elevation resolution is lower. For the sparse scenes in the elevation direction, compressive sensing (CS) methods can be used to reconstruct the elevation information, and have the super-resolution capability. The iterative shrinkage-thresholding(IST) algorithm can be used for TomoSAR, but its convergence rate is slower. A fast iterative shrinkage-thresholding (FIST) algorithm is introduced for TomoSAR, which will not only maintain the accuracy of the IST algorithm, but also has a faster convergence rate. The characteristics of the FIST algorithm are shown in multiple scatterers distinguish, single-scatter position estimation and other aspects through simulation experiments. Then, the TerraSAR-X actual data of Beijing region is used to analyze the imaging result of TomoSAR. The research result shows that the FIST algorithm has obvious advantages in the aspects of multiple scatterers distinguish and single-scatter position estimation. It can apply to TomoSAR with better imaging results.