Sar Tomographic Imaging Demonstration Using Gf-3 Data

Synthetic aperture radar (SAR) tomography (TomoSAR) is well established for three-dimensional (3D) reconstruction in urban buildings. There are many methods usually require a large data stacks for a reliable reconstruction and very few methods to handle small data stacks. This paper proposes a new processing method for 3D reconstruction of tall buildings using extremely small data stacks with large baseline interval. In this method, it is necessary to select persistent scatterers (PSs) with high signal-to-noise ratio (SNR) as reference points and estimate the elevation ambiguity issue by using the geometric relationship of SAR imaging. This paper presents the experimental results and shows a well performance of the proposed method in 3D reconstruction of tall buildings with simulated data and SAR data acquired by Chinese spaceborne SAR GF-3 satellite.