An Accurate High Resolution ISAR Image Reconstruction Method Based on 2-D Nonuniform Fast Fourier Transform

摘要

We present a new method to reconstruct high resolution inverse synthetic aperture radar (ISAR) images'. When the far-field data arranged in the polar format, the two-dimensional (2-D) image can be reconstructed by a 2-D Fourier transform directly. But the fast Fourier transform (FFT) is failed for the data do not falling on a Cartesian grid. We introduce the 2-D nonuniform fast Fourier transform (NUFFT) to reconstruct the ISAR image. It is the first time to use NUFFT technology to achieve ISAR imaging. Instead of interpolating the input data, NUFFT interpolate the exponential function for each sampling point. The technology could easily extend to high dimensional by tensor product. The 2-D NUFFT is very suitable to reconstruct ISAR image. An example will be shown to demonstrate that the reconstructed images have higher resolution than the general method when using the same data.