海杂波中基于可控虚警K近邻的海面小目标检测

郭子薰 水鹏朗 白晓惠 许述文 李东宸

郭子薰, 水鹏朗, 白晓惠, 等. 海杂波中基于可控虚警K近邻的海面小目标检测[J]. 雷达学报, 2020, 9(4): 654–663. doi: 10.12000/JR20055
引用本文: 郭子薰, 水鹏朗, 白晓惠, 等. 海杂波中基于可控虚警K近邻的海面小目标检测[J]. 雷达学报, 2020, 9(4): 654–663. doi: 10.12000/JR20055
GUO Zixun, SHUI Penglang, BAI Xiaohui, et al. Sea-Surface small target detection based on K-NN with controlled false alarm rate in sea clutter[J]. Journal of Radars, 2020, 9(4): 654–663. doi: 10.12000/JR20055
Citation: GUO Zixun, SHUI Penglang, BAI Xiaohui, et al. Sea-Surface small target detection based on K-NN with controlled false alarm rate in sea clutter [J]. Journal of Radars, 2020, 9(4): 654–663. doi: 10.12000/JR20055

海杂波中基于可控虚警K近邻的海面小目标检测

doi: 10.12000/JR20055
基金项目: 国家自然科学基金(61871303)
详细信息
    作者简介:

    郭子薰(1994–),女,陕西西安人,西安电子科技大学博士生。主要研究方向为雷达目标检测、机器学习和海杂波信号处理。E-mail: zxguo_724@stu.xidian.edu.cn

    水鹏朗(1967–),男,陕西西安人,博士,教授。1999年在西安电子科技大学获得博士学位,现担任西安电子科技大学电子工程学院雷达信号处理国家重点实验室教授、硕导、博导。主要研究方向为海杂波建模、雷达目标检测和图像处理。E-mail: plshui@xidian.edu.cn

    白晓惠(1998–),女,陕西宝鸡人,西安电子科技大学博士生。主要研究方向为雷达目标检测、机器学习和海杂波信号处理。E-mail: xhbai@stu.xidian.edu.cn

    许述文(1985–),男,安徽黄山人,博士,副教授。2011年在西安电子科技大学获得博士学位,现担任西安电子科技大学电子工程学院雷达信号处理国家重点实验室副教授、硕导、博导。主要研究方向为雷达目标检测、机器学习、时频分析和SAR图像处理。E-mail: swxu@mail.xidian.edu.cn

    李东宸(1988–),男,陕西西安人,博士,高级工程师。2016年在西安电子科技大学获得博士学位,现就职于中国船舶工业系统工程研究院。主要研究方向为目标检测、图像处理、无人机任务规划。E-mail: charlieli.xidian@outlook.com

    通讯作者:

    郭子薰 zxguo_724@stu.xidian.edu.cn

    水鹏朗 plshui@xidian.edu.cn

  • 责任主编:刘宁波 Corresponding Editor: LIU Ningbo
  • 中图分类号: TN957.51

Sea-surface Small Target Detection Based on K-NN with Controlled False Alarm Rate in Sea Clutter

Funds: The National Natural Science Foundation of China (61871303)
More Information
  • 摘要: 由于高分辨海杂波具有复杂的特性以及海面小目标具有多样性,没有精确的简单统计模型可以较好地描述海杂波和目标回波时间序列,这导致目标检测遇到了很多阻碍。为了区分海杂波和目标回波,分别提取它们的特征将检测问题转化为特征空间中的分类问题是一种有效的方法。基于特征的检测可以归结为在特征空间中的一种2元假设检验问题,但是其有两个问题需要解决:一是目标回波数据远少于杂波数据;二是虚警概率不可控。为了解决第1个问题,一种典型小目标的仿真回波产生器被用于产生充足的典型目标回波数据,以辅佐后续检测器的设计。K近邻(K-NN)是一种简单有效的分类方法,但是因为无法精确地控制虚警率而不能直接在目标检测中使用。该文提出一种基于改进K-NN的海面小目标检测方法,可以很好地实现可控虚警。经IPIX雷达数据集验证,所提出的方法在观测时间分别为0.512 s和1.024 s时获得了85.1%和89.2%的检测概率,相比现有的检测器获得了7%和5%的提升,具有良好的检测效果和稳定性。

     

  • 图  1  仿真目标7特征与真实目标7特征对比图

    Figure  1.  The comparisons of seven features of simulated targets returns and real targets returns

    图  2  所提检测器的流程图

    Figure  2.  The flowchart of the proposed detector

    图  3  所提检测器与其余检测器的检测概率

    Figure  3.  Detection probabilities of the proposed detector and other detectors

    图  4  所提检测器与其余检测器的检测概率

    Figure  4.  Detection probabilities of the proposed detector and other detectors

    图  5  k值不同时,所实现的虚警率变化图,其中w*=3

    Figure  5.  Realized false alarm rate when k takes different values, where the w*=3

    表  1  IPIX数据集描述[10]

    Table  1.   Description of IPIX radar database[10]

    序号数据WS(km/h)SWHs(m)Angle (degree)目标单元影响单元
    119931107_135603_starea1792.2998,10,11
    219931108_220902_starea2691.19776,8
    319931109_191449_starea30190.99876,8
    419931109_202217_starea31190.99876,8,9
    519931110_001635_starea4091.08875,6,8
    619931111_163625_starea54200.7887,9,10
    719931118_023604_stareC0000280101.613087,9,10
    819931118_162155_stareC0000310330.93076,8,9
    919931118_162658_stareC0000311330.94076,8,9
    1019931118_174259_stareC0000320280.93076,8,9
    1119980204_202225_ANTSTEP1652423,25,26
    1219980204_202525_ANTSTEP18076,8,9
    1319980204_163113_ANTSTEP1652423,25,26
    1419980205_171437_ANTSTEP18076,8,9
    1519980205_180558_ANTSTEP18076,8,9
    1619980212_195704_ANTSTEP18076,8,9
    1719980223_164055_ANTSTEP1653130,32,33
    1819980223_173317_ANTSTEP1653231,33,34
    1919980223_173950_ANTSTEP1652928,30–34
    2019980304_184537_ANTSTEP2120,22
    下载: 导出CSV

    表  2  IPIX数据集上多种检测器的平均检测概率

    Table  2.   The average detection probabilities of detectors on IPIX radar database

    检测器观测时间(s)HHHVVHVV平均
    基于分形的检测器[12]0.5120.2230.4040.4480.2410.329
    1.0240.3010.5360.5760.3280.435
    基于3特征的检测器[9]0.5120.5770.7360.7760.5690.665
    1.0240.6220.7970.8130.5980.708
    基于时频3特征的检测器[11]0.5120.7470.8260.8420.7060.780
    1.0240.8210.8820.8770.7890.842
    所提检测器0.5120.8210.8870.8950.8000.851
    1.0240.8680.9220.9210.8580.892
    下载: 导出CSV

    表  3  IPIX数据库中20组数据的平均检测结果对比

    Table  3.   The comparisons of average detection results of 20 datasets on IPIX radar database

    检测器虚警概率HHHVVHVV平均
    基于分形的检测器[12]0.0100.3220.5140.5540.3350.431
    0.0010.2230.4040.4480.2410.329
    基于3特征的检测器[9]0.0100.6730.8120.8410.6570.745
    0.0010.5770.7360.7760.5690.665
    基于时频3特征的检测器[11]0.0100.8060.8780.8820.7810.837
    0.0010.7470.8260.8420.7060.780
    所提检测器0.0100.8870.9320.9320.8680.905
    0.0010.8210.8870.8950.8000.851
    下载: 导出CSV

    表  4  基于6特征的KNN检测器在IPIX数据库上20组数据的平均检测结果对比 (%)

    Table  4.   The average detection results comparisons of KNN-based detectors using six features at 20 datasets on IPIX radar database (%)

    检测器所去掉的特征NHERAARDPHRVERIMSNR
    性能损失HH0.010.795.140.503.011.460.20
    HV0.360.022.500.014.190.700.17
    VH0.670.212.680.085.451.310.12
    VV0.010.634.620.010.891.020.02
    平均0.260.413.740.153.381.120.13
    下载: 导出CSV

    表  5  所提检测器在20组IPIX雷达数据集上实现的虚警概率

    Table  5.   The realized false alarm rate of the proposed detector of 20 datasets on the IPIX radar database

    所需的虚警率实现的虚警率
    HHHVVHVV平均
    0.0100.0087740.0091530.0087760.0085630.008816
    0.0010.0010110.0010280.0010100.0010150.001016
    下载: 导出CSV
  • [1] WANG Shibin, CHEN Xufeng, WANG Yan, et al. Nonlinear squeezing time–frequency transform for weak signal detection[J]. Signal Processing, 2015, 113: 195–210. doi: 10.1016/j.sigpro.2015.01.022
    [2] WARD K, TOUGH R, and WATTS S. Sea Clutter: Scattering, the K Distribution and Radar Performance[M]. 2nd ed. London: The Institution of Engineering and Technology, 2013. doi: 10.1049/PBRA025E.
    [3] RABIDEAU D J and PARKER P. Ubiquitous MIMO multifunction digital array radar[C]. The 37th Asilomar Conference on Signals, Systems & Computers, Pacific Grove, USA, 2003: 1057–1064. doi: 10.1109/ACSSC.2003.1292087.
    [4] CARTA P, GALATI G, PIRACCI E G, et al. Implementation of the ubiquitous radar concept with a conformal array[C]. 2015 European Radar Conference (EuRAD), Paris, France, 2015: 1463–1466. doi: 10.1109/EuRAD.2015.7346332.
    [5] SHI Sainan and SHUI Penglang. Detection of low-velocity and floating small targets in sea clutter via income-reference particle filters[J]. Signal Processing, 2018, 148: 78–90. doi: 10.1016/j.sigpro.2018.02.005
    [6] 陈世超, 罗丰, 胡冲, 等. 基于多普勒谱非广延熵的海面目标检测方法[J]. 雷达学报, 2019, 8(3): 344–354. doi: 10.12000/JR19012

    CHEN Shichao, LUO Feng, HU Chong, et al. Small target detection in sea clutter background based on Tsallis entropy of Doppler spectrum[J]. Journal of Radars, 2019, 8(3): 344–354. doi: 10.12000/JR19012
    [7] 左磊, 产秀秀, 禄晓飞, 等. 基于空域联合时频分解的海面微弱目标检测方法[J]. 雷达学报, 2019, 8(3): 335–343. doi: 10.12000/JR19035

    ZUO Lei, CHAN Xiuxiu, LU Xiaofei, et al. A weak target detection method in sea clutter based on joint space-time-frequency decomposition[J]. Journal of Radars, 2019, 8(3): 335–343. doi: 10.12000/JR19035
    [8] 苏宁远, 陈小龙, 关键, 等. 基于卷积神经网络的海上微动目标检测与分类方法[J]. 雷达学报, 2018, 7(5): 565–574. doi: 10.12000/JR18077

    SU Ningyuan, CHEN Xiaolong, GUAN Jian, et al. Detection and classification of maritime target with micro-motion based on CNNs[J]. Journal of Radars, 2018, 7(5): 565–574. doi: 10.12000/JR18077
    [9] SHUI Penglang, LI Dongchen, and XU Shuwen. Tri-feature-based detection of floating small targets in sea clutter[J]. IEEE Transactions on Aerospace and Electronic Systems, 2014, 50(2): 1416–1430. doi: 10.1109/taes.2014.120657
    [10] Cognitive Systems Laboratory. IPIX radar database[EB/OL]. http://soma.mcmaster.ca//ipix.php, 2012.
    [11] SHI Sainan and SHUI Penglang. Sea-surface floating small target detection by one-class classifier in time-frequency feature space[J]. IEEE Transactions on Geoscience and Remote Sensing, 2018, 56(11): 6395–6411. doi: 10.1109/TGRS.2018.2838260
    [12] HU Jing, TUNG W W, and GAO Jianbo. Detection of low observable targets within sea clutter by structure function based multifractal analysis[J]. IEEE Transactions on Antennas and Propagation, 2006, 54(1): 136–143. doi: 10.1109/TAP.2005.861541
    [13] LI Dongchen and SHUI Penglang. Floating small target detection in sea clutter via normalised Hurst exponent[J]. Electronics Letters, 2014, 50(17): 1240–1242. doi: 10.1049/el.2014.1569
    [14] GUO Zixun and SHUI Penglang. Sea-surface floating small target detection based on feature compression[J]. The Journal of Engineering, 2019, 2019(21): 8160–8164. doi: 10.1049/joe.2019.0694
    [15] SHUI Penglang, GUO Zixun, and SHI Sainan. Feature-compression-based detection of sea-surface small targets[J]. IEEE Access, 2019, 8: 8371–8385. doi: 10.1109/ACCESS.2019.2962793
    [16] ZHANG Shichao, LI Xuelong, and ZONG Ming. Efficient kNN classification with different numbers of nearest neighbors[J]. IEEE Transactions on Neural Networks and Learning Systems, 2018, 29(5): 1774–1785. doi: 10.1109/TNNLS.2017.2673241
    [17] FAYED H A and ATIYA A F. A novel template reduction approach for the K-nearest neighbor method[J]. IEEE Transactions on Neural Networks, 2009, 20(5): 890–896. doi: 10.1109/TNN.2009.2018547
  • 加载中
图(5) / 表(5)
计量
  • 文章访问数:  928
  • HTML全文浏览量:  429
  • PDF下载量:  191
  • 被引次数: 0
出版历程
  • 收稿日期:  2020-05-08
  • 修回日期:  2020-07-06
  • 网络出版日期:  2020-07-29

目录

    /

    返回文章
    返回