基于深度学习的多特征融合海面目标检测方法

汪翔 汪育苗 陈星宇 臧传飞 崔国龙

汪翔, 汪育苗, 陈星宇, 等. 基于深度学习的多特征融合海面目标检测方法[J]. 雷达学报(中英文), 2024, 13(3): 554–564. doi: 10.12000/JR23105
引用本文: 汪翔, 汪育苗, 陈星宇, 等. 基于深度学习的多特征融合海面目标检测方法[J]. 雷达学报(中英文), 2024, 13(3): 554–564. doi: 10.12000/JR23105
WANG Xiang, WANG Yumiao, CHEN Xingyu, et al. Deep learning-based marine target detection method with multiple feature fusion[J]. Journal of Radars, 2024, 13(3): 554–564. doi: 10.12000/JR23105
Citation: WANG Xiang, WANG Yumiao, CHEN Xingyu, et al. Deep learning-based marine target detection method with multiple feature fusion[J]. Journal of Radars, 2024, 13(3): 554–564. doi: 10.12000/JR23105

基于深度学习的多特征融合海面目标检测方法

doi: 10.12000/JR23105
基金项目: 国家自然科学基金(62271126),衢州市财政资助科研项目(2022D008, 2022D005),广东省重点领域研发计划(2020B090905002),高等学校学科创新引智计划(B17008)
详细信息
    作者简介:

    汪 翔,博士生,主要研究方向为雷达目标检测识别和机器学习

    汪育苗,博士生,主要研究方向为雷达目标检测、杂波抑制和深度学习

    陈星宇,硕士生,主要研究方向为雷达目标检测跟踪和深度学习

    臧传飞,硕士生,主要研究方向为雷达杂波抑制、雷达目标跟踪和深度学习

    崔国龙,教授,博士生导师,主要研究方向为最优化理论和算法、雷达目标检测理论、波形多样性以及阵列信号处理等

    通讯作者:

    崔国龙 cuiguolong@uestc.edu.cn

  • 责任主编:许述文 Corresponding Editor: XU Shuwen
  • 中图分类号: TN957.51

Deep Learning-based Marine Target Detection Method with Multiple Feature Fusion

Funds: The National Natural Science Foundation of China (62271126), The Municipal Government of Quzhou (2022D008, 2022D005), The Guangdong Key Areas Research and Development Program (2020B090905002), 111 Project (B17008)
More Information
  • 摘要: 该文考虑了海杂波环境下的雷达目标检测问题,提出了一种基于深度学习的海面目标检测器。该检测器通过融合从不同数据源中提取的多种互补性特征以增加目标和杂波的差异性,从而提升对海面目标的检测性能。具体来说,该检测器首先利用两个特征提取分支分别从距离像和距离多普勒谱图中提取多层次快时间特征和距离特征;然后,设计局部-全局特征提取结构从特征的慢时间维度或多普勒维度提取序列关联性;接着,提出基于自适应卷积权重学习的特征融合模块,实现快慢时间特征和距离多普勒特征的高效融合;最后,对多层次特征进行融合、上采样和非线性映射获得检测结果。基于两个公开雷达数据集上的实验验证了所提检测器的检测性能。

     

  • 图  1  IPIX数据集#17子数据集的某段HH极化距离像

    Figure  1.  Range profiles of #17 sub-data set with HH polarization in IPIX database

    图  2  IPIX数据集#17子数据集HH极化距离多普勒谱图

    Figure  2.  Range-Doppler spectrum of #17 sub-data set with HH polarization in IPIX database

    图  3  MFF检测器架构图

    Figure  3.  Architecture of the MFF detector

    图  4  序列特征提取模块结构

    Figure  4.  The structure of sequence feature extraction block

    图  5  基于自适应卷积权重学习的多特征融合模块结构

    Figure  5.  Structure of multiple feature fusion block based on adaptive convolution weight learning

    图  6  检测层结构

    Figure  6.  The structure of detection layer

    图  7  不同检测器在IPIX数据集上检测性能

    Figure  7.  Detection performance of various detectors in IPIX database

    图  8  “2021010150614_03_staring.mat”数据集幅度图

    Figure  8.  Amplitude image of “2021010150614_03_staring.mat” data set

    图  9  不同检测器在测试集上的检测可视化

    Figure  9.  Detection visualization of various detectors on test set

    表  1  MFF检测器参数设置

    Table  1.   Parameter setting of MFF detector

    参数名称特征提取模块池化层序列特征提取模块特征融合模块上采样层检测层
    核尺寸(1,3), (1,3)
    (1,3), (1,3)
    (1,3), (1,3)
    (1,2)(3,1), (3,1)
    (1,1), (1,1)
    (1,1), (1,1)
    (1,3), (1,3)
    (1,4)
    (1,2)
    (1,3), (1,3)
    (1,1)
    步长(1,1), (1,1)
    (1,1), (1,1)
    (1,1), (1,1)
    (1,1)(2,1), (2,1)
    (1,1), (1,1)
    (1,1), (1,1)
    (1,1), (1,1)
    (1,4)
    (1,2)
    (1,1), (1,1)
    (1,1)
    填充尺寸(0,1), (0,1)
    (0,1), (0,1)
    (0,1), (0,1)
    (0,0)(1,0), (1,0)
    (0,0), (0,0)
    (0,0), (0,0)
    (0,1), (0,1)
    (0,0)
    (0,0)
    (0,1), (0,1)
    (0,0)
    通道维度 (1,16), (16,16)
    (16,32), (32,32)
    (32,64), (64,64)
    N/A(${C_1}$,$2{C_1}$),
    ($2{C_1}$,${C_1}$)
    (16,8), (8,1)
    (${C_1}$,1), (${C_1}$,1)
    (${C_1}$,${{{C_1}} \mathord{\left/ {\vphantom {{{C_1}} 2}} \right. } 2}$),
    (${{{C_1}} \mathord{\left/ {\vphantom {{{C_1}} 2}} \right. } 2}$,${C_1}$)
    (64,64)
    (32,32)
    (112,64), (64,64)
    (64,2)
    注:对于3个序列特征提取模块和特征融合模块,${C_1}$依次设置为16, 32, 64。
    下载: 导出CSV

    表  2  雷达参数

    Table  2.   Radar parameters

    数据集名称带宽(MHz)载频(GHz)脉冲重复频率(Hz)脉宽(μs)
    IPIX数据集59.3910000.2
    雷达对海探测数据集259.30~9.5016003.0
    下载: 导出CSV

    表  3  IPIX数据集子数据集信息

    Table  3.   Information of sub-dataset in IPIX database

    数据集
    索引
    子数据集名称目标主回
    波单元
    目标次回
    波单元
    #1719931107_135603_starea98, 10, 11
    #1819931107_141630_starea98, 10, 11
    #1919931107_145028_starea87, 9
    #2519931108_213827_starea76, 8
    #2619931108_220902_starea76, 8
    #3019931109_191449_starea76, 8
    #3119931109_202217_starea76, 8
    #4019931110_001635_starea75, 6, 8
    #5419931111_163625_starea87, 9, 10
    #28019931118_023604_stareC000087, 10
    #28319931118_035737_stareC0000108, 9, 11, 12
    #31019931118_162155_stareC000076, 8, 9
    #31119931118_162658_stareC000076, 8, 9
    #32019931118_174259_stareC000076, 8, 9
    下载: 导出CSV

    表  4  训练样本数和测试样本数

    Table  4.   Sample number of training set and test set

    数据集名称训练集样本数测试集样本数
    IPIX数据集每个子数据集78585237
    雷达对海探测数据集953374
    下载: 导出CSV

    表  5  MFF检测器和其变体检测器检测性能

    Table  5.   Detection performance of the MFF detector and its variant detectors

    检测器名称平均实际虚警率平均检测概率
    MFF0.00040.9870
    变体10.00050.9478
    变体20.00050.9814
    下载: 导出CSV

    表  6  不同特征融合方法检测性能

    Table  6.   Detection performance of different feature fusion approaches

    融合方法平均实际虚警率平均检测概率
    所提方法0.00040.9870
    特征串接0.00030.9477
    特征求和0.00310.9619
    下载: 导出CSV

    表  7  不同检测器在IPIX数据集不同极化数据上的平均检测概率

    Table  7.   Averagely detection probability of various detectors in IPIX database with different polarizations

    检测器名称HHHVVVVH
    MFF1.00001.00001.00001.0000
    三特征0.64650.86200.584808630
    支持向量机0.67670.76730.60530.7761
    MDCCNN0.82240.89980.82170.9100
    Bi-LSTM0.73800.88890.79700.7963
    下载: 导出CSV

    表  8  不同检测器在IPIX数据集不同极化数据上的平均实际虚警率

    Table  8.   Averagely actual false alarm rate of various detectors in IPIX database with different polarizations

    检测器名称HHHVVVVH
    MFF0.00140.00160.00160.0017
    三特征0.00370.00430.00380.0043
    支持向量机0.00330.00360.00420.0035
    MDCCNN0.00470.00200.00490.0020
    Bi-LSTM0.00420.00220.00570.0029
    下载: 导出CSV

    表  9  不同检测器在雷达对海探测数据集上的检测性能

    Table  9.   Detection performance of various detectors in SDRDSP database

    检测器名称实际虚警率检测概率
    MFF0.00040.9870
    CA-CFAR0.00830.9265
    GO-CFAR0.00520.8944
    SO-CFAR0.02250.9588
    下载: 导出CSV
  • [1] 苏宁远, 陈小龙, 关键, 等. 基于深度学习的海上目标一维序列信号目标检测方法[J]. 信号处理, 2020, 36(12): 1987–1997. doi: 10.16798/j.issn.1003-0530.2020.12.004.

    SU Ningyuan, CHEN Xiaolong, GUAN Jian, et al. One-dimensional sequence signal detection method for marine target based on deep learning[J]. Journal of Signal Processing, 2020, 36(12): 1987–1997. doi: 10.16798/j.issn.1003-0530.2020.12.004.
    [2] 赵文静, 刘畅, 刘文龙, 等. K分布海杂波背景下基于最大特征值的雷达信号检测算法[J]. 电子与信息学报, 2018, 40(9): 2235–2241. doi: 10.11999/JEIT171092.

    ZHAO Wenjing, LIU Chang, LIU Wenlong, et al. Maximum eigenvalue based radar signal detection method for K distribution sea clutter environment[J]. Journal of Electronics &Information Technology, 2018, 40(9): 2235–2241. doi: 10.11999/JEIT171092.
    [3] 陈小龙, 关键, 何友. 微多普勒理论在海面目标检测中的应用及展望[J]. 雷达学报, 2013, 2(1): 123–134. doi: 10.3724/SP.J.1300.2012.20102.

    CHEN Xiaolong, GUAN Jian, and HE You. Applications and prospect of micro-motion theory in the detection of sea surface target[J]. Journal of Radars, 2013, 2(1): 123–134. doi: 10.3724/SP.J.1300.2012.20102.
    [4] LO T, LEUNG H, LITVA J, et al. Fractal characterisation of sea-scattered signals and detection of sea-surface targets[J]. IEE Proceedings F (Radar and Signal Processing), 1993, 140(4): 243–250. doi: 10.1049/ip-f-2.1993.0034.
    [5] XU Xiaoke. Low observable targets detection by joint fractal properties of sea clutter: An experimental study of IPIX OHGR datasets[J]. IEEE Transactions on Antennas and Propagation, 2010, 58(4): 1425–1429. doi: 10.1109/TAP.2010.2041144.
    [6] 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.
    [7] 邵夫驰, 行鸿彦. 基于FRFT的多重分形海面小目标检测[J]. 探测与控制学报, 2020, 42(1): 69–74, 80.

    SHAO Fuchi and XING Hongyan. Small target detection based on multi-fractal characteristics of sea clutter FRFT spectrum[J]. Journal of Detection &Control, 2020, 42(1): 69–74, 80.
    [8] 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.
    [9] LI Yuzhou, XIE Pengcheng, TANG Zeshen, et al. SVM-based sea-surface small target detection: A false-alarm-rate-controllable approach[J]. IEEE Geoscience and Remote Sensing Letters, 2019, 16(8): 1225–1229. doi: 10.1109/LGRS.2019.2894385.
    [10] 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.
    [11] YAN Kun, BAI Yu, WU H C, et al. Robust target detection within sea clutter based on graphs[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(9): 7093–7103. doi: 10.1109/TGRS.2019.2911451.
    [12] 时艳玲, 姚婷婷, 郭亚星. 基于图连通密度的海面漂浮小目标检测[J]. 电子与信息学报, 2021, 43(11): 3185–3192. doi: 10.11999/JEIT201028.

    SHI Yanling, YAO Tingting, and GUO Yaxing. Floating small target detection based on graph connected density in sea surface[J]. Journal of Electronics &Information Technology, 2021, 43(11): 3185–3192. doi: 10.11999/JEIT201028.
    [13] 许述文, 焦银萍, 白晓惠, 等. 基于频域多通道图特征感知的海面小目标检测[J]. 电子与信息学报, 2023, 45(5): 1567–1574. doi: 10.11999/JEIT220188.

    XU Shuwen, JIAO Yinping, BAI Xiaohui, et al. Small target detection based on frequency domain multichannel graph feature perception on sea surface[J]. Journal of Electronics &Information Technology, 2023, 45(5): 1567–1574. doi: 10.11999/JEIT220188.
    [14] YAN Yujia, WU Guangxin, DONG Yang, et al. Floating small target detection in sea clutter using mean spectral radius[J]. IEEE Geoscience and Remote Sensing Letters, 2022, 19: 4023405. doi: 10.1109/LGRS.2022.3165163.
    [15] 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.
    [16] 陈世超, 高鹤婷, 罗丰. 基于极化联合特征的海面目标检测方法[J]. 雷达学报, 2020, 9(4): 664–673. doi: 10.12000/JR20072.

    CHEN Shichao, GAO Heting, and LUO Feng. Target detection in sea clutter based on combined characteristics of polarization[J]. Journal of Radars, 2020, 9(4): 664–673. doi: 10.12000/JR20072.
    [17] XU Shuwen, ZHENG Jibin, PU Jia, et al. Sea-surface floating small target detection based on polarization features[J]. IEEE Geoscience and Remote Sensing Letters, 2018, 15(10): 1505–1509. doi: 10.1109/LGRS.2018.2852560.
    [18] 施赛楠, 杨静, 董泽远. 基于高维特征域随机森林的海面小目标检测[J]. 现代雷达, 2022, 44(3): 63–69. doi: 10.16592/j.cnki.1004-7859.2022.03.011.

    SHI Sainan, YANG Jing, and DONG Zeyuan. Detection of small sea-surface target based on random forest in high-dimensional feature domain[J]. Modern Radar, 2022, 44(3): 63–69. doi: 10.16592/j.cnki.1004-7859.2022.03.011.
    [19] CHEN Simin, FENG Chen, HUANG Yong, et al. Small target detection in X-band sea clutter using the visibility graph[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 5115011. doi: 10.1109/TGRS.2022.3186283.
    [20] WANG Jin’gang and LI Songbin. Maritime radar target detection in sea clutter based on CNN with dual-perspective attention[J]. IEEE Geoscience and Remote Sensing Letters, 2023, 20: 3500405. doi: 10.1109/LGRS.2022.3230443.
    [21] WAN Hao, TIAN Xiaoqing, LIANG Jing, et al. Sequence-feature detection of small targets in sea clutter based on Bi-LSTM[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 4208811. doi: 10.1109/TGRS.2022.3198124.
    [22] WANG Yumiao, ZHAO Wenjing, WANG Xiang, et al. Nonhomogeneous sea clutter suppression using complex-valued U-Net model[J]. IEEE Geoscience and Remote Sensing Letters, 2022, 19: 4027705. doi: 10.1109/LGRS.2022.3214633.
    [23] QU Qizhe, WANG Yongliang, LIU Weijian, et al. A false alarm controllable detection method based on CNN for sea-surface small targets[J]. IEEE Geoscience and Remote Sensing Letters, 2022, 19: 4025705. doi: 10.1109/LGRS.2022.3190865.
    [24] 李骁, 施赛楠, 董泽远, 等. 基于时频域深度网络的海面小目标特征检测[J]. 雷达科学与技术, 2022, 20(2): 209–216, 230. doi: 10.3969/j.issn.1672-2337.2022.02.013.

    LI Xiao, SHI Sainan, DONG Zeyuan, et al. Feature detection of small sea-surface target via deep network in time-frequency domain[J]. Radar Science and Technology, 2022, 20(2): 209–216, 230. doi: 10.3969/j.issn.1672-2337.2022.02.013.
    [25] CHEN Xiaolong, SU Ningyuan, HUANG Yong, et al. False-alarm-controllable radar detection for marine target based on multi features fusion via CNNs[J]. IEEE Sensors Journal, 2021, 21(7): 9099–9111. doi: 10.1109/JSEN.2021.3054744.
    [26] IPIX Radar. The McMaster IPIX radar sea clutter database[EB/OL]. http://soma.ece.mcmaster.ca/ipix/, 2021.
    [27] TRABELSI C, BILANIUK O, ZHANG Ying, et al. Deep complex networks[C]. 6th International Conference on Learning Representations, Vancouver, Canada, 2018: 1–19.
    [28] 刘宁波, 丁昊, 黄勇, 等. X波段雷达对海探测试验与数据获取年度进展[J]. 雷达学报, 2021, 10(1): 173–182. doi: 10.12000/JR21011.

    LIU Ningbo, DING Hao, HUANG Yong, et al. Annual progress of the sea-detecting X-band radar and data acquisition program[J]. Journal of Radars, 2021, 10(1): 173–182. doi: 10.12000/JR21011.
  • 加载中
图(9) / 表(9)
计量
  • 文章访问数:  2002
  • HTML全文浏览量:  671
  • PDF下载量:  835
  • 被引次数: 0
出版历程
  • 收稿日期:  2023-06-14
  • 修回日期:  2023-07-15
  • 网络出版日期:  2023-08-15
  • 刊出日期:  2024-06-28

目录

    /

    返回文章
    返回