UWBSAR叶簇遮蔽目标检测中的ROI提取方法.pdfVIP

29 11 Vol. 29 No. 11 2007 11 Systems Engineering and Electronics Nov. 2007 : 1001506X(2007) 111 3705 UWB SAR ROI 杨志国, 黄晓涛, 周智敏 ( 国防科学技术大学电子科学与工程学院, 湖南长沙410073) : 针对SAR(synthetic aperture radar) AT R( auto target recognition) 算法中的ROI(region of interest) 提取通常由一个CFAR( constant false alarm rat e) 检测器和聚类算法来完成, 方法在高波段SAR 目标检测中具 有优良的性能, 而在UWB SAR( ultraw ide band synthetic aperture radar) 叶簇遮蔽目标检测中效果不佳提出了 一种适于叶簇遮蔽目标检测的ROI 提取方法, 方法由小滑窗中值滤波低门限CFAR 检测形态学操作和聚类 算法四部分组成, 能够在叶簇遮蔽目标检测中很好地完成ROI 提取, 基于实际UWB SAR 图像的ROI 提取结果 验证了 方法的有效性 : 超宽带合成孔径雷达; 目标识别; 叶簇遮蔽; ROI 提取; 中值滤波; CFAR 检测; 形态学; 聚类 : T N 95 : A ROI extract method f or UWB SAR foliageconcealed target detection YANG Zhiguo, HU ANG Xiaotao, ZHOU Zhimin ( School of Electr onic S cience and Eng ineering , N ational Univ . of Def ense Technolog y , Changsha 410073 , China Abstract: ROI ( region of interest) extract in SA R ( synthetic aperture radar) AT R ( auto target recogni tion) algorithm is usually accomplished by a CFAR ( Constant False Alarm rate) detector followed by a cluste ring algorithm. It has an excellent perf ormance on target detection in high frequency SAR images. But it isnt effective to extract ROI when it is used to detect f oliageconcealed target s in UWB SAR images. So a ROI ex tract method is proposed for the foliageconcealed target detection. The proposed method is composed of a medi an filter of small window , a CFAR detector of low threshold, a morphologic operation and a clustering algo rithm. It is eff ective to complete ROI extract in the foliageconcealed target detection. T he ROI extract result s of different extract methods based on a factual UWB SA R