摘要
研究了基于特征提取的SAR图像理解问题 ,提出了一个基于高分辨率SAR图像的人造目标打击效果评估系统 ,该系统包括目标检测、目标提取、特征提取和分级评估四个模块 ,并重点分析了系统中特定人造目标如建筑物特征信息的选择和提取等问题 .经过适当扩展和改进 ,该系统可用于各种不同人造目标的打击效果评估和自动识别 .仿真结果表明 ,该系统可有效地对人造目标的打击效果如是否摧毁。
This paper focuses on the SAR image understanding on the basis of feature extracting, and proposes a system which can be used to evaluate some man-made targets' hitting effect. This system includes four modules: target detection; target selection; features extracting; two-level hitting effect evaluation. The selection and extraction of features of special targets, such as buildings, are analyzed. After proper improvement this system can be used to evaluate a variety of man-made targets hitting effect and work as an ATR system. The simulation results show that this system is feasible and effective.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2003年第9期1290-1294,共5页
Acta Electronica Sinica
关键词
SAR图像理解
特征选取
打击效果评估
Evaluation
Feature extraction
Image understanding
Radar target recognition
