A New Ship Target Detecting Method Based on Saliency in SAR Image
Yanyan Hu, Xiaofan Wang, Wei Li, Xinhong Hei, Guo Wei Xie · 2021 7th Annual International Conference on Network and Information Systems for Computers (ICNISC) · 2021
SAR (Synthetic Aperture Radar) is a kind of active microwave sensor, and has been widely used in the field of ocean ship detection. Because SAR imaging is affected by the speckle noise, sea clutter and complex environment, it is still a challenging task about ship target detection in SAR image processing. In this paper a fast ship target detecting method based on saliency is proposed. Firstly, the SAR image is separated to land and sea in order to remove the influence of land scattering. In the next step, the sea candidate area is segmented by super pixels, the global and local contrast of super pixels are calculated by using Bhattacharyya dissimilarity measure, and then combine them to generate the saliency map. Finally, the Otsu algorithm is used to separate the target regions in the saliency maps. The experimental results show that the proposed method effectively suppressed the influence of sea and land clutter, enhanced the saliency of the target, which greatly improves the detection efficiency and accuracy.