Sea-Surface Target Angular Superresolution in Forward-Looking Radar Imaging Based on Maximum A Posteriori Algorithm
Yin Zhang,Qiping Zhang,Changlin Li,Yongchao Zhang,Yulin Huang,Jianyu Yang
DOI: https://doi.org/10.1109/JSTARS.2019.2918189
IF: 4.715
2019-01-01
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Abstract:Forward-looking imaging for sea-surface target has raised many concerns in the fields of vessel monitoring and sea rescue in all-day and all-weather time. Constrained by the imaging principle, conventional high-resolution radar imaging techniques, such as synthetic aperture radar and Doppler beam sharpening, are incapable of forward-looking imaging. Currently, scanning radar is the main tool to obtain forward-looking image, but it suffers from low angular resolution. Recently, much attention has been paid to superresolution methods, which can provide improved angular resolution over that of real-beam scanning radar. However, these methods are mainly used for ground targets, which mainly consider the ground clutter. For sea surface imaging, the distribution of sea clutter is more complex than the ground clutter. Besides, the performance of sea surface imaging is limited by the moving speed of targets. If we directly use these existing methods to sea-surface target imaging, it will lead to amplification of clutter and even elimination of the real target. In this paper, we consider the sea clutter distribution and present a Rayleigh-based sparse maximum
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">a posteriori</italic>
(RSMAP) deconvolution algorithm to improve the angular resolution of the sea-surface low-speed and static targets. First, the Rayleigh distribution is considered as the likelihood function because it can describe the amplitude characteristic of the sea clutter. Second, since the target concerned is often sparse compared with the imaging region, the sparse constraint is considered as the prior information of the sea-surface targets in objective function. Finally, the targets distribution is obtained by MAP estimation under the Bayesian framework. The simulations and measured data processing results have shown that the proposed algorithm provides better angular resolution and target location accuracy than traditional methods.