摘要
提出了一种用于SAR图像目标检测的基于区域分类的智能恒虚警算法.该算法组合常规的单元平均恒虚警、选小恒虚警和选大恒虚警的有利特性,在多窗口划分和区域类型分类的基础上,实现目标的智能检测.实验表明,该方法在同质性区域保持较好检测性能的同时,在多目标和杂波边缘的异质性区域也有较强的鲁棒性.这必将极大的降低后续辨别和识别阶段的复杂度,从而提高了整个ATR系统的性能.
Based on multiple child-windows and region classification, this paper proposes a new intelligent CFAR method that makes a full use of CA-CFAR, SO-CFAR and GO-CFAR. The experimental results on SAR image show that the new method has a good detection performance not only in homogenous clutter background but also in non-homogenous clutter background including multi-target environment and clutter edge environment. This will greatly reduce the complexity of discrimination and classification in the next stage, and improve the performance of the whole ATR system.
出处
《武汉大学学报(理学版)》
CAS
CSCD
北大核心
2004年第1期104-108,共5页
Journal of Wuhan University:Natural Science Edition
基金
国家自然科学基金资助项目(60072041
60372057)
关键词
区域分类
智能恒虚警
SAR图像
目标检测
合成孔径雷达
target detection
region classification
constant false alarm rate
synthetic aperture radar