系统工程与电子技术 ›› 2021, Vol. 43 ›› Issue (2): 300-310.doi: 10.12305/j.issn.1001-506X.2021.02.03

• 电子技术 • 上一篇    下一篇

基于EMD能量占比的海面漂浮小目标特征检测

时艳玲(), 刘子鹏(), 张学良(), 顾为亮()   

  1. 南京邮电大学通信与信息工程学院, 江苏 南京 210003
  • 收稿日期:2020-06-03 出版日期:2021-02-01 发布日期:2021-03-16
  • 作者简介:时艳玲 (1983-),女,副教授,博士,主要研究方向为海杂波的散射特性分析、复杂电磁环境下的目标检测和雷达信号处理。E-mail:ylshi@njupt.edu.cn|刘子鹏 (1995-),男,硕士研究生,主要研究方向为海杂波背景下基于特征的海面漂浮小目标的检测。E-mail:1218012603@njupt.edu.cn|张学良 (1995-),男,助理工程师,硕士,主要研究方向为基于特征的海杂波背景下的目标检测和雷达信号处理。E-mail:1217012431@njupt.edu.cn|顾为亮 (1996-),男,硕士研究生,主要研究方向为复杂海面环境下对海面弱小目标的检测。E-mail:1318014609@njupt.edu.cn
  • 基金资助:
    国家自然科学基金(61201325);南京邮电大学国自孵化基金(NY218045)

Feature detection of floating small target on the sea surface based on EMD energy proportion

Yanling SHI(), Zipeng LIU(), Xueliang ZHANG(), Weiliang GU()   

  1. School of Communication and Information Engineering, Nanjing University of Posts and Telecommunications, Nanjing 210003, China
  • Received:2020-06-03 Online:2021-02-01 Published:2021-03-16

摘要:

针对传统的时频分析方法对海杂波分析有限的问题,提出一种基于经验模态分解(empirical mode decomposition, EMD)能量占比的海面漂浮小目标特征检测方法。首先,采用EMD将接收回波分为独立不同尺度的若干个固有模态(intrinsic mode function, IMF)分量,实现对接收回波的频率从高频到低频的分解。然后,分别建立IMF分量与接收回波数据的相关系数,并利用平均均值-标准差之比作为筛选IMF分量的准则,自动筛选出能量较大且波动平稳的低阶IMF分量。最后,提取IMF分量在原始信号中的平均能量占比作为特征,利用蒙特卡罗方法设置门限,进行海面目标异常检测。实测数据的结果显示,所提算法的性能优于对比算法。

关键词: 经验模态分解, 相关系数, 能量占比, 漂浮小目标检测

Abstract:

Aiming at the problem that the traditional time-frequency analysis method is limited to the analysis of sea clutter, a feature detection method of floating small target on the sea surface based on empirical mode decomposition (EMD) energy proportion is proposed. Firstly, the received echo is divided into several independent intrinsic mode function (IMF) components with different scales by EMD, which can realize the decomposition of the frequency for the received echo from high frequency to low frequency. Secondly, the correlation coefficients between the IMF components and the received echo data are established respectively. And the ratio of mean to standard deviation is used as the criterion to filter the IMF components, which can automatically select the low-order IMF components with large energy and stable fluctuation. Finally, the average energy proportion of IMF component in the original signal is extracted as the feature, and the Monte Carlo method is used to set the threshold for sea surface target anomaly detection. The results of the measured data show that the performance of the proposed algorithm is better than that of the comparison algorithm.

Key words: empirical mode decomposition (EMD), correlation coefficient, energy ratio, floating small target detection

中图分类号: