Cooperative Enhancement of Ship Targets in SAR Images Based on Density Features
Xueqian Wang, Gang Li, You He · 2021 CIE International Conference on Radar (Radar) · 2021
In this paper, we propose a new method to improve the target-to-clutter ratios (TCRs) of SAR image chips cropped from a large SAR image from the perspective of cooperative enhancement (CE) with the density features. First, the superpixel-based saliency map of each SAR image chip is independently obtained. Then, we innovatively design the superpixel-histogram density (SHD) feature of salience maps to accumulate target cues from all the SAR image chips containing ship targets, where the weight of cues from each SAR image chip is calculated in accordance with its self-quality. The SHD feature exploits the dependence of all the SAR image chips, i.e., sparse ship targets and non-sparse sea clutter. Finally, we fuse the salience maps and SHD features of SAR image chips. In this way, the ship target therein is enhanced and the sea clutter is relatively suppressed from the CE viewpoint. Experimental results on Gaofen-3 SAR images are reported to demonstrate that the newly proposed method provides higher TCRs than the state-of-the-art with better target enhancement performance.