一种SAR图像舰船目标快速检测方法

基于SAR图像的舰船目标自动检测是海洋监视应用的重要方面,但随着SAR成像能力和图像分辨率的提高,传统的CFAR检测方法已不能满足舰船目标自动检测的要求。针对中高分辨率SAR图像中舰船目标自动检测问题,提出一种基于像素筛选G0分布的SAR图像舰船目标快速检测方法,该方法首先根据像素灰度值出现频率选取阈值对杂波像素进行筛选,然后通过抽样定理对图像进行降分辨率处理,最后再在经过像素筛选的降分辨率图像中实现基于G0分布的自适应CFAR检测。NASA/JPL AIR-SAR实测数据的实验结果表明,该方法不仅能有效减少中高分辨率SAR图像舰船目标自动检测的虚警,而且能显著提高检测效率。

A Fast Method for Ship Detection in SAR Imagery

Automatic ship detection plays an important role in the ocean surveillance application of SAR imagery. As the increasing of SAR imaging ability and image resolution, the conventional CFAR can＇t meet the requirement of automatic ship detection. For the ship detection in medium and high resolution SAR im- agery, a fast method for ship detection from SAR imagery based on pixel excision in C;~＇ distributed sea clutter is proposed in this paper. The method excises clutter pixets with a threshold determined by the appearance frequency first, then reduces resolution by the sampling procedure and carries through automatic ship detec- tion finally. Detection results on JPL/NASA AIRSAR data illustrate the method can reduce the number of false alarm and improve the calculation efficiency remarkably.