基于峰值点形态信息的SAR图像舰船尾迹检测算法  

Detection Algorithm of the Ship Wakes from SAR Imagery Based on the Peak Morphological Information

作　　者：邹焕新[1] 郁文贤[1] 匡纲要[1] 郑键[1] 

机构地区：[1]国防科技大学电子科学与工程学院,湖南长沙410073

出　　处：《国防科技大学学报》2005年第2期87-91,共5页Journal of National University of Defense Technology

摘　　要：在分析合成孔径雷达(SAR)图像中舰船尾迹线性特性的基础上,针对SAR图像线性特征产生因素的多样性,及由此引起的Radon变换域中尾迹峰值点检测的不确定性,提出了一种基于峰值点形态信息的尾迹检测算法。算法门限化图像的Radon变换系数,提取出所有可能的峰值点;对这些相应的局部峰值点的一维截面进行连续小波变换峰值点匹配,根据提取到的参数形成决策矢量在特征空间中进行决策。仿真和实际数据处理的结果表明,该方法能有效、准确地检测到SAR图像中的舰船尾迹并判决其真假类别。The linear characteristics of the ship wakes in SAR image is analyzed. Due to the fact that there are many factors which can generate the linear features, so aiming at the problem of the uncertainty in detecting the ship wake peaks in the Radon transform domain, an algorithm based on the peak morphological information to detect the ship wakes in SAR ocean imagery is proposed. The algorithm extracts all possible peaks by thresholding the Radon coefficients, and perform the continuous wavelet transform to the peak match for all these 1-dimensional local peak sections, and makes decision in the feature space finally using the decision vectors formed by the extracted parameters. The simulation and real data processing results show that the algorithm is reliable and can efficiently improve the accuracy of detection.

关 键 词：SAR图像 尾迹检测 RADON变换 连续小波变换 特征空间 决策 

分 类 号：TN957[电子电信—信号与信息处理]