SAR图像自动目标识别技术分析

A survey on SAR ATR algorithms

合成孔径雷达(SAR)图像自动目标识别(ATR)是SAR图像解译的关键技术之一,其旨在获取感兴趣区域可能包含的目标类别,为战场情报分析提供有力支撑。文中在广泛阅读相关文献的基础上,从基于模板和基于模型两个方面对现有的SAR图像目标识别方法进行了较为全面的综述,对该方向的主要研究结论进行了归纳总结。同时,针对SAR图像目标识别的潜在方向进行了展望。

Automatic target recognition （ATR） of synthetic aperture radar （SAR） images is one of thekey techniques in SAR image interpretation. SAR ATR aims to determine the target type in a region ofinterest thus providing supports for the battlefield information analysis. Based on the extensiveinvestigation of the published works, this paper conducts a comprehensive survey on the SAR ATRalgorithms from two aspects： template - based methods and model - based methods. Some conclusions aredrawn based on the previous works. Meanwhile, some potential directions of this filed are also stated.