SAR 3D Scene Reconstruction Using Fourier Imaging Techniques

Synthetic aperture radar (SAR) images are conventionally collected from platforms following a straight and level flight path. The resulting Fourier domain data exist on a 2D plane which is projected onto the ground plane for imaging, producing a 2D image of the target scene. To obtain a 3D image of the target scene interferometric SAR is typically employed, this involves determining the phase difference between two independent SAR images. However if the SAR data is collected from a nonlinear 3D flight path then the Fourier domain data exist on a 2D surface with extent in all three dimensions. Scene reconstruction from this type of flight path has been covered in the literature previously; however the algorithms used for scene reconstruction are computationally intensive. This paper covers 3D scene reconstruction using Fourier imaging techniques. A 3D equivalent to the polar format algorithm (PFA) is used for scene reconstruction, where the 2D data surface is resampled onto a regular rectangular 3D grid of points to allow use of the fast Fourier transform (FFT). The irregularly sampled 2D surface is mapped to another domain where resampling to the regular rectangular 3D grid points near the surface is a simpler operation. The FFT is then performed in each direction to reconstruct the target scene. This method of scene reconstruction has been simulated in MATLAB and the results obtained have been compared to results obtainable from the same data using a matched filter (MF) based approach. The results from two different nonlinear 3D flight paths are discussed to demonstrate the algorithm, an orthogonal aperture and a circular curved arc. The results of the simulations show that Fourier imaging methods can provide a less computationally intensive alternative to the methods discussed in the literature