Low Computational Complexity SAR Imaging Algorithm for Ship Monitoring via 2-D Band-Limited Sparse Fourier Transform

For synthetic aperture radar (SAR) imaging with big data, traditional fast Fourier transform (FFT)-based imaging algorithms are no longer fast enough for some real-time applications, such as ship monitoring and early warning. The imaging scene of ships in a calm sea background presents block sparse feature. In this article, considering the block sparse feature, we propose a low computational complexity imaging algorithm. First, we introduce a sublinear complexity method called 2-D band-limited sparse Fourier transform (2D-BLSFT) to accelerate 2D-FFT. Then, <tex-math notation="LaTeX">ω K </tex-math> -BLSFT algorithm for SAR imaging is designed based on <tex-math notation="LaTeX">ω K </tex-math> algorithm and 2D-BLSFT. Due to the sublinear computational and sample complexity of 2D-BLSFT, the computational complexity in 2-D inverse FFT (2D-IFFT) step and Stolt mapping step is reduced. Furthermore, the robust processing is studied considering the nonideal block sparse feature in practical scenes. Numerical simulations confirm that 2D-BLSFT outperforms the state-of-the-art sparse FFT (SFFT) algorithms. Moreover, experimental results based on simulation data and GF-3 spaceborne SAR data verify that the imaging performance of <tex-math notation="LaTeX">ω K </tex-math> -BLSFT algorithm surpasses that of existing SFFT-based algorithms.