东北大学学报(自然科学版) ›› 2024, Vol. 45 ›› Issue (10): 1401-1408.DOI: 10.12068/j.issn.1005-3026.2024.10.005

• 信息与控制 • 上一篇    

基于SPWVD-STFT的海面弱目标检测方法

成怡1,2, 王阳1()   

  1. 1.天津工业大学 控制科学与工程学院,天津 300387
    2.天津工业大学 天津市电气装备智能控制重点实验室,天津 300387
  • 收稿日期:2023-05-22 出版日期:2024-10-31 发布日期:2024-12-31
  • 通讯作者: 王阳
  • 作者简介:成 怡(1979-),女,黑龙江齐齐哈尔人,天津工业大学副教授.

Sea-Surface Weak Target Detection Method Based on SPWVD-STFT

Yi CHENG1,2, Yang WANG1()   

  1. 1.School of Control Science and Engineering,Tiangong University,Tianjin 300387,China
    2.Tianjin Key Laboratory of Intelligent Control of Electrical Equipment,Tiangong University,Tianjin 300387,China.
  • Received:2023-05-22 Online:2024-10-31 Published:2024-12-31
  • Contact: Yang WANG
  • About author:WANG Yang,E-mail: wangyang990327@163.com

摘要:

为了进一步提升时频域特征检测海面弱目标的能力,提出一种平滑伪魏格纳-威利分布(smoothed pseudo Wigner-Ville distribution,SPWVD)-短时傅里叶变换(short?time Fourier transform,STFT)海面弱目标检测算法.首先,采用STFT对回波信号进行时频特征分析,优化SPWVD的时频特征分析结果,并引入 K-medoids聚类算法对二者时频矩阵进行降噪处理.然后,提取时频域特征多普勒频率稳定度(Doppler frequency stability,DFS),利用快速凸包学习算法获得虚警可控的判决区域,从而判定海杂波与目标.最后,基于IPIX数据集中实测数据的实验结果表明所提出的检测算法在相同虚警率下比时频三特征检测器的平均检测概率高6.3%.

关键词: 海杂波, 弱目标检测, 时频分析, K-medoids聚类, 凸包检测器

Abstract:

To further improve the capability of time?frequency domain features to detect weak targets on the sea?surface, a smoothed pseudo Wigner-Ville distribution (SPWVD)-short?time Fourier transform (STFT) sea?surface weak target detection algorithm is proposed. Firstly, STFT is adopted to perform time?frequency features analysis on the echo signals, and to optimize the time?frequency features analysis results of SPWVD. The K-medoids clustering algorithm is introduced to denoise the time?frequency matrix. Then, the time?frequency features Doppler frequency stability (DFS) is extracted, and the fast convex hull learning algorithm is utilized to obtain the false alarm controllable judgment region, so as to determine the sea clutter and the target. Finally, results of experiments based on Ice multiparameter imaging X-Band radar (IPIX) measured data show that the detection probability of the proposed detection algorithm is 6.3% higher than that of the time?frequency tri?feature detector at the same false alarm rate.

Key words: sea clutter, weak target detection, time?frequency analysis, K-medoids clustering, convex hull detector

中图分类号: