A Novel Method for Abrupt Motions Estimation of Sparse Scatterers in Differential Tomography SAR Imaging

In urban areas with many buildings, the scatterers located within one pixel can be assumed to be discrete and sparse (m = 0, 1, or 2). When the abrupt motions of sparse scatterers only happen in a single synthetic aperture radar (SAR) acquisition, these motions cannot be calculated using 2-D or 3-D frequency spectrum estimation method based on traditional differential Tomography SAR model. In this paper, a novel approach is proposed to solve this issue. By compensating a phase factor from the acquisition which has an abrupt motion and then maximizing the 3D Elevation-Linear motion-Amplitude of periodic motion spectrum at the scatterers locations, the final abrupt motions of the scatterers could be calculated. For the scatterers simulated, the estimation accuracy of the locations along elevation direction, linear motions velocity, the amplitudes of seasonal periodic motions, the amplitudes of relative scattering values, and the abrupt motions in the last acquisition are shown. The simulation results demonstrate that the proposed method can achieve high accuracy for no-noise situation. The root-mean-square errors of the abrupt motions estimation are also acceptable (about 1–2 mm) when the signal-to-noise ratio = 10 and 8 dB. Thus, the estimation accuracy of the proposed algorithm can also fulfill the practical applications requirement.