| 一种自适应的合成孔径雷达图像目标检测方法 |
| |
| 引用本文： | 张名成,吴秀清,王鹏伟.一种自适应的合成孔径雷达图像目标检测方法[J].计算机工程与应用,2006,42(10):200-203. |
| |
| 作者姓名： | 张名成 吴秀清 王鹏伟 |
| |
| 作者单位： | 中国科学技术大学电子工程与信息科学系,合肥,230027;中国科学技术大学电子工程与信息科学系,合肥,230027;中国科学技术大学电子工程与信息科学系,合肥,230027 |
| |
| 摘 要： | 目标检测是自动目标识别的一个重要步骤,论文提出了一种自适应的SAR图像目标检测方法,该方法采用基于Weibull分布模型的恒虚警率(CFAR)检测技术,将参考窗口分块,判断各子块类型,根据各子块类型不同,自适应选择参考样本确定阈值。在检测过程中,利用灰度和方差特征,预先排除明显不为目标的像素。对CFAR检测结果,利用目标基本形状特征排除虚警。实验证明,该方法在同质区和非同质区背景下都具有较好的检测性能。
|
| 关 键 词： | 自动目标识别 目标检测 合成孔径雷达 恒虚警率 |
| 文章编号： | 1002-8331-（2006）10-0200-04 |
| 收稿时间： | 2005-09-01 |
| 修稿时间： | 2005-09-01 |
An Adaptive Approach for Target Detection in SAR Images |
| |
| Zhang Mingcheng,Wu Xiuqing,Wang Pengwei.An Adaptive Approach for Target Detection in SAR Images[J].Computer Engineering and Applications,2006,42(10):200-203. |
| |
| Authors: | Zhang Mingcheng Wu Xiuqing Wang Pengwei |
| |
| Affiliation: | Department of Electronic Engineering and Information Science,University of Science and Technology of China,Hefei 230027 |
| |
| Abstract: | An approach based on CFAR(constant false alarm rate) processor to perform target detection in SAR image is presented.Weibull distribution model is used for ground clutter.The reference Window is splited into 4 sub-blocks.Based on statistics of each block,the approach dynamically tailors the background estimation algorithm.It employs the feature of intensity and local variance to remove unexpected pixels during CFAR process and basic geometrical feature to remove false alarm.It provides a good performance in a homogeneous environment and also performs robustly in nonhomogeneous environments including multiple targets and extended clutter edges. |
| |
| Keywords: | ATR target detection SAR CFAR |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
