高级搜索

卷积神经网络在雷达自动目标识别中的研究进展

贺丰收 何友 刘准钆 徐从安

引用本文: 贺丰收, 何友, 刘准钆, 徐从安. 卷积神经网络在雷达自动目标识别中的研究进展[J]. 电子与信息学报, doi: 10.11999/JEIT180899 shu
Citation:  Fengshou HE, You HE, Zhunga LIU, Cong’an XU. Research and Development on Applications of Convolutional Neural Networks of Radar Automatic Target Recognition[J]. Journal of Electronics and Information Technology, doi: 10.11999/JEIT180899 shu

卷积神经网络在雷达自动目标识别中的研究进展

    作者简介: 贺丰收: 男,1979年生,高级工程师,博士生,研究方向为雷达数据处理,多源信息融合,深度神经网络等;
    何友: 男,1956年生,中国工程院院士,博士生导师,研究方向为多源信息融合,信号检测,雷达数据处理等;
    刘准钆: 男,1984年生,教授,研究方向为多源信息融合,证据推理,模式识别;
    徐从安: 男,1987年生,博士,讲师,研究方向为多目标跟踪,信息融合等
    通讯作者: 贺丰收,hefengshou1979@163.com
  • 基金项目: 国家自然科学基金(61672431, 61790550, 91538201)

摘要: 自动目标识别(ATR)是雷达信息处理领域的重要研究方向。由于卷积神经网络(CNN)无需进行特征工程,图像分类性能优越,因此在雷达自动目标识别领域研究中受到越来越多的关注。该文综合论述了CNN在雷达图像处理中的应用进展。首先介绍了雷达自动目标识别相关知识,包括雷达图像的特性,并指出了传统的雷达自动目标识别方法局限性。给出了CNN卷积神经网络原理、组成和在计算机视觉领域的发展历程。然后着重介绍了CNN在雷达自动目标识别中的研究现状,其中详细介绍了合成孔径雷达(SAR)图像目标的检测与识别方法。接下来对雷达自动目标识别面临的挑战进行了深入分析。最后对CNN新理论、新模型,以及雷达新成像技术和未来复杂环境下的应用进行了展望。

English

    1. [1]

      KRIZHEVSKY A, SUTSKEVER I, and HINTON G E. Imagenet classification with deep convolutional neural networks[C]. Proceedings of the 25th International Conference on Neural Information Processing Systems, Lake Tahoe, USA, 2012: 1097–1105.

    2. [2]

      CHENG Gong, HAN Junwei, and LU Xiaoqiang. Remote sensing image scene classification: benchmark and state of the art[J]. Proceedings of the IEEE, 2017, 105(10): 1865–1883. doi: 10.1109/JPROC.2017.2675998

    3. [3]

      陈小龙, 关键, 何友, 等. 高分辨稀疏表示及其在雷达动目标检测中的应用[J]. 雷达学报, 2017, 6(3): 239–251. doi: 10.12000/JR16110
      CHEN Xiaolong, GUAN Jian, HE You, et al. High-resolution sparse representation and its applications in radar moving target detection[J]. Journal of Radars, 2017, 6(3): 239–251. doi: 10.12000/JR16110

    4. [4]

      BALL J E, ANDERSON D T, and CHAN C S. Comprehensive survey of deep learning in remote sensing: theories, tools, and challenges for the community[J]. Journal of Applied Remote Sensing, 2017, 11(4): 042609. doi: 10.1117/1.JRS.11.042609

    5. [5]

      PEI Jifang, HUANG Yulin, HUO Weibo, et al. SAR automatic target recognition based on multiview deep learning framework[J]. IEEE Transactions on Geoscience and Remote Sensing, 2018, 56(4): 2196–2210. doi: 10.1109/tgrs.2017.2776357

    6. [6]

      GOODFELLOW I, BENGIO Y, and COURVILLE A. Deep Learning[M]. Cambridge, Massachusetts, MIT Press, 2016.

    7. [7]

      LECUN Yann, BOTTOU Léon, BENGIO Yoshua, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278–2324. doi: 10.1109/5.726791

    8. [8]

      RUSSAKOVSKY O, DENG Jia, SU Hao, et al. Imagenet large scale visual recognition challenge[J]. International Journal of Computer Vision, 2015, 115(3): 211–252. doi: 10.1007/s11263-015-0816-y

    9. [9]

      ZEILER M D and FERGUS R. Visualizing and understanding convolutional networks[C]. Proceedings of the 13th European Conference on Computer Vision, Zurich, Switzerland, 2014: 818–833.

    10. [10]

      SZEGEDY C, LIU Wei, JIA Yangqing, et al. Going deeper with convolutions[C]. Proceedings of the 2015 IEEE Conference on Computer Vision and Pattern Recognition, Boston, USA, 2015: 1–9.

    11. [11]

      SIMONYAN K and ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[EB/OL]. http://arxiv.org/abs/1409.1556, 2014.

    12. [12]

      HE Kaiming, ZHANG Xiangyu, REN Shaoqing, et al. Deep residual learning for image recognition[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, USA, 2016: 770–778. doi: 10.1109/CVPR.2016.90.

    13. [13]

      HU Jie, SHEN Li, ALBANIE S, et al. Squeeze-and-excitation networks[EB/OL]. https://arxiv.org/abs/1709.01507, 2017.

    14. [14]

      许强, 李伟, LOUMBI P. 深度卷积神经网络在SAR自动目标识别领域的应用综述[J]. 电讯技术, 2018, 58(1): 106–112. doi: 10.3969/j.issn.1001-893x.2018.01.019
      XU Qiang, LI Wei, and LOUMBI P. Applications of Deep convolutional neural network in SAR automatic target recognition: a summarization[J]. Telecommunication Engineering, 2018, 58(1): 106–112. doi: 10.3969/j.issn.1001-893x.2018.01.019

    15. [15]

      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

    16. [16]

      杜兰, 刘彬, 王燕, 等. 基于卷积神经网络的SAR图像目标检测算法[J]. 电子与信息学报, 2016, 38(12): 3018–3025. doi: 10.11999/JEIT161032
      DU Lan, LIU Bin, WANG Yan, et al. Target detection method based on convolutional neural network for SAR image[J]. Journal of Electronics &Information Technology, 2016, 38(12): 3018–3025. doi: 10.11999/JEIT161032

    17. [17]

      GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Columbus, USA, 2014: 580–587.

    18. [18]

      HE Kaiming, ZHANG Xiangyu, REN Shaoqing, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[C]. Proceedings of the 13th European Conference on Computer Vision, Zurich, Switzerland, 2014: 346–361.

    19. [19]

      GIRSHICK R. Fast R-CNN[C]. Proceedings of the IEEE international Conference on Computer Vision, Santiago, Chile, 2015: 1440–1448.

    20. [20]

      REN Shaoqing, HE Kaiming, GIRSHICK R, et al. Faster R-CNN: Towards real-time object detection with region proposal networks[C]. Proceedings of the 28th International Conference on Neural Information Processing Systems, Montreal, Canada, 2015: 91–99.

    21. [21]

      DAI Jifeng, LI Yi, HE Kaiming, et al. R-FCN: Object detection via region-based fully convolutional networks[C]. Proceedings of the 30th International Conference on Neural Information Processing Systems, Barcelona, Spain, 2016: 379–387.

    22. [22]

      KONG Tao, YAO Anbang, CHEN Yurong, et al. Hypernet: Towards accurate region proposal generation and joint object detection[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, USA, 2016: 845–853.

    23. [23]

      LIN T Y, DOLLÁR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 936–944.

    24. [24]

      HE Kaiming, GKIOXARI G, DOLLÁR P, et al. Mask R-CNN[C]. Proceedings of the 2017 IEEE International Conference on Computer Vision, Venice, Italy, 2017: 2980–2988.

    25. [25]

      WANG Sifei, CUI Zongyong, and CAO Zongjie. Target recognition in large scene SAR images based on region proposal regression[C]. Proceedings of the 2017 IEEE International Geoscience and Remote Sensing Symposium, Fort Worth, USA, 2017: 3297–3300.

    26. [26]

      LI Jianwei, QU Changwen, and SHAO Jiaqi. Ship detection in SAR images based on an improved faster R-CNN[C]. Proceedings of the 2017 SAR in Big Data Era: Models, Methods and Applications, Beijing, China, 2017: 1–6.

    27. [27]

      KANG Miao, LENG Xiangguang, LIN Zhao, et al. A modified faster R-CNN based on CFAR algorithm for SAR ship detection[C]. Proceedings of the 2017 International Workshop on Remote Sensing with Intelligent Processing, Shanghai, China, 2017: 1–4.

    28. [28]

      KANG Miao, JI Kefeng, LENG Xiangguang, et al. Contextual region-based convolutional neural network with multilayer fusion for SAR ship detection[J]. Remote Sensing, 2017, 9(8): 860. doi: 10.3390/rs9080860

    29. [29]

      JIAO Jiao, ZHANG Yue, SUN Hao, et al. A densely connected end-to-end neural network for multiscale and multiscene SAR ship detection[J]. IEEE Access, 2018, 6: 20881–20896. doi: 10.1109/ACCESS.2018.2825376

    30. [30]

      ZHONG Yanfei, HAN Xiaobing, and ZHANG Liangpei. Multi-class geospatial object detection based on a position-sensitive balancing framework for high spatial resolution remote sensing imagery[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2018, 138: 281–294. doi: 10.1016/j.isprsjprs.2018.02.014

    31. [31]

      REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]. Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, USA, 2016: 779–788.

    32. [32]

      LIU Wei, ANGUELOV D, ERHAN D, et al. SSD: Single shot multibox detector[C]. Proceedings of the 14th European Conference on Computer Vision, Amsterdam, The Netherlands, 2016: 21–37.

    33. [33]

      WANG Yuanyuan, WANG Chao, ZHANG Hong, et al. Combing single shot multibox detector with transfer learning for ship detection using Chinese Gaofen-3 images[C]. Proceedings of the 2017 Progress in Electromagnetics Research Symposium - Fall, Singapore, Singapore, 2018: 712–716.

    34. [34]

      WANG Yuanyuan, WANG Chao, and ZHANG Hong. Combining a single shot multibox detector with transfer learning for ship detection using sentinel-1 SAR images[J]. Remote Sensing Letters, 2018, 9(8): 780–788. doi: 10.1080/2150704X.2018.1475770

    35. [35]

      KONG Tao, SUN Fuchun, YAO Anbang, et al. Ron: Reverse connection with objectness prior networks for object detection[C]. Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 5244–5252.

    36. [36]

      CUI Zongyong, DANG Sihang, CAO Zongjie, et al. SAR target recognition in large scene images via region-based convolutional neural networks[J]. Remote Sensing, 2018, 10(5): 776. doi: 10.3390/rs10050776

    37. [37]

      NI Jiacheng and XU Yuelei. SAR automatic target recognition based on a visual cortical system[C]. Proceedings of the 20136th International Congress on Image and Signal Processing, Hangzhou, China, 2013: 778–782.

    38. [38]

      CHEN Sizhe and WANG Haipeng. SAR target recognition based on deep learning[C]. Proceedings of the 2014 International Conference on Data Science and Advanced Analytics, Shanghai, China, 2014: 541–547.

    39. [39]

      WAGNER S. Combination of convolutional feature extraction and support vector machines for radar ATR[C]. Proceedings of the 17th International Conference on Information Fusion, Salamanca, Spain, 2014: 1–6.

    40. [40]

      WANG Haipeng, CHEN Sizhe, XU Feng, et al. Application of deep-learning algorithms to MSTAR data[C]. Proceedings of the 2015 IEEE International Geoscience and Remote Sensing Symposium, Milan, Italy, 2015: 3743–3745.

    41. [41]

      WAGNER S. Morphological component analysis in SAR images to improve the generalization of ATR systems[C]. Proceedings of the 20153rd International Workshop on Compressed Sensing Theory and its Applications to Radar, Sonar and Remote Sensing, Pisa, Italy, 2015: 46–50.

    42. [42]

      SCHWEGMANN C P, KLEYNHANS W, SALMON B P, et al. Very deep learning for ship discrimination in Synthetic Aperture Radar imagery[C]. Proceedings of the 2016 IEEE International Geoscience and Remote Sensing Symposium, Beijing, China, 2016: 104–107.

    43. [43]

      CHO J H and PARK C G. Additional feature CNN based automatic target recognition in SAR image[C]. Proceedings of 2017 Fourth Asian Conference on Defence Technology, Tokyo, Japan, 2017: 1–4.

    44. [44]

      LIN Zhao, JI Kefeng, KANG Miao, et al. Deep convolutional highway unit network for SAR target classification with limited labeled training data[J]. IEEE Geoscience and Remote Sensing Letters, 2017, 14(7): 1091–1095. doi: 10.1109/lgrs.2017.2698213

    45. [45]

      HE Hao, WANG Shicheng, YANG Dongfang, et al. SAR target recognition and unsupervised detection based on convolutional neural network[C]. Proceedings of the 2017 Chinese Automation Congress, Jinan, China, 2017: 435–438.

    46. [46]

      田壮壮, 占荣辉, 胡杰民, 等. 基于卷积神经网络的SAR图像目标识别研究[J]. 雷达学报, 2016, 5(3): 320–325. doi: 10.12000/JR16037
      TIAN Zhuangzhuang, ZHAN Ronghui, HU Jiemin, et al. SAR ATR based on convolutional neural network[J]. Journal of Radars, 2016, 5(3): 320–325. doi: 10.12000/JR16037

    47. [47]

      CHEN Sizhe, WANG Haipeng, XU Feng, et al. Target classification using the deep convolutional networks for SAR images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 54(8): 4806–4817. doi: 10.1109/tgrs.2016.2551720

    48. [48]

      WILMANSKI M, KREUCHER C, and LAUER J. Modern approaches in deep learning for SAR ATR[C]. Proceedings of SPIE 9843 Algorithms for Synthetic Aperture Radar Imagery XXIII, Baltimore, USA, 2016: 98430N.

    49. [49]

      赵娟萍, 郭炜炜, 柳彬, 等. 基于概率转移卷积神经网络的含噪标记SAR图像分类[J]. 雷达学报, 2017, 6(5): 514–523. doi: 10.12000/jr16140
      ZHAO Juanping, GUO Weiwei, LIU Bin, et al. Convolutional neural network-based SAR image classification with noisy labels[J]. Journal of Radars, 2017, 6(5): 514–523. doi: 10.12000/jr16140

    50. [50]

      AMRANI M and JIANG Feng. Deep feature extraction and combination for synthetic aperture radar target classification[J]. Journal of Applied Remote Sensing, 2017, 11(4): 042616. doi: 10.1117/1.Jrs.11.042616

    51. [51]

      WANG Ning, WANG Yinghua, LIU Hongwei, et al. Feature-fused SAR target discrimination using multiple convolutional neural networks[J]. IEEE Geoscience and Remote Sensing Letters, 2017, 14(10): 1695–1699. doi: 10.1109/lgrs.2017.2729159

    52. [52]

      ZHENG Ce, JIANG Xue, and LIU Xingzhao. Generalized synthetic aperture radar automatic target recognition by convolutional neural network with joint use of two-dimensional principal component analysis and support vector machine[J]. Journal of Applied Remote Sensing, 2017, 11(4): 046007. doi: 10.1117/1.Jrs.11.046007

    53. [53]

      刘晨, 曲长文, 周强, 等. 基于卷积神经网络迁移学习的SAR图像目标分类[J]. 现代雷达, 2018, 40(3): 38–42. doi: 10.16592/j.cnki.1004-7859.2018.03.009
      LIU Chen, QU Changwen, ZHOU Qiang, et al. SAR image target classification based on convolutional neural network transfer learning[J]. Modern Radar, 2018, 40(3): 38–42. doi: 10.16592/j.cnki.1004-7859.2018.03.009

    54. [54]

      LI Xuan, LI Chunsheng, WANG Pengbo, et al. SAR ATR based on dividing CNN into CAE and SNN[C]. Proceedings of the 2015 IEEE 5th Asia-Pacific Conference on Synthetic Aperture Radar, Singapore, Singapore, 2015: 676–679.

    55. [55]

      李松, 魏中浩, 张冰尘, 等. 深度卷积神经网络在迁移学习模式下的SAR目标识别[J]. 中国科学院大学学报, 2018, 35(1): 75–83. doi: 10.7523/j.issn.2095-6134.2018.01.010
      LI Song, WEI Zhonghao, ZHANG Bingchen, et al. Target recognition using the transfer learning-based deep convolutional neural networks for SAR images[J]. Journal of University of Chinese Academy of Sciences, 2018, 35(1): 75–83. doi: 10.7523/j.issn.2095-6134.2018.01.010

    56. [56]

      ØDEGAARD N, KNAPSKOG A O, COCHIN C, et al. Classification of ships using real and simulated data in a convolutional neural network[C]. Proceedings of the 2016 IEEE Radar Conference, Philadelphia, PA, USA, 2016: 1–6.

    57. [57]

      KARABAYIR O, YUCEDAG O M, KARTAL M Z, et al. Convolutional neural networks-based ship target recognition using high resolution range profiles[C]. Proceedings of the 201718th International Radar Symposium, Prague, Czech Republic, 2017.

    58. [58]

      BENTES C, VELOTTO D, and LEHNER S. Target classification in oceanographic SAR images with deep neural networks: Architecture and initial results[C]. Proceedings of the 2015 IEEE International Geoscience and Remote Sensing Symposium, Milan, Italy, 2015: 3703–3706.

    59. [59]

      WANG Zhaocheng, DU Lan, WANG Fei, et al. Multi-scale target detection in SAR image based on visual attention model[C]. Proceedings of the 2015 IEEE 5th Asia-Pacific Conference on Synthetic Aperture Radar, Singapore, Singapore, 2015: 704–709.

    60. [60]

      YUAN Lele. A time-frequency feature fusion algorithm based on neural network for HRRP[J]. Progress in Electromagnetics Research, 2017, 55: 63–71. doi: 10.2528/PIERM16123002

    61. [61]

      BENGIO Y, MESNARD T, FISCHER A, et al. STDP-compatible approximation of backpropagation in an energy-based model[J]. Neural Computation, 2017, 29(3): 555–577. doi: 10.1162/NECO_a_00934

    62. [62]

      LECUN Y, BENGIO Y, and HINTON G. Deep learning[J]. Nature, 2015, 521(7553): 436–444. doi: 10.1038/nature14539

    63. [63]

      HOWARD A G, ZHU Menglong, CHEN Bo, et al. Mobilenets: Efficient convolutional neural networks for mobile vision applications[EB/OL]. http://arxiv.org/abs/1704.04861, 2017.

    64. [64]

      HUANG Gao, LIU Zhuang, VAN DER MAATEN L, et al. Densely connected convolutional networks[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 2261–2269.

    65. [65]

      GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets[C]. Proceedings of the 27th International Conference on Neural Information Processing Systems, Montreal, Canada, 2014: 2672–2680.

    1. [1]

      郭晨, 简涛, 徐从安, 何友, 孙顺. 基于深度多尺度一维卷积神经网络的雷达舰船目标识别. 电子与信息学报,

    2. [2]

      叶磊, 王勇, 杨强, 邓维波. 基于对数行列式散度与对称对数行列式散度的高频地波雷达目标检测器. 电子与信息学报,

    3. [3]

      杨宏宇, 王峰岩. 基于深度卷积神经网络的气象雷达噪声图像语义分割方法. 电子与信息学报,

    4. [4]

      左国玉, 杜婷婷, 马蕾, 卢佳豪, 龚道雄. 基于动作注意策略的树形DDQN目标候选区域提取方法. 电子与信息学报,

    5. [5]

      孙彦景, 石韫开, 云霄, 朱绪冉, 王赛楠. 基于多层卷积特征的自适应决策融合目标跟踪算法. 电子与信息学报,

    6. [6]

      李煜, 陈杰, 张渊智. 合成孔径雷达海面溢油探测研究进展. 电子与信息学报,

    7. [7]

      王玉莹, 张志敏, 李宁, 范怀涛, 赵庆超. 高分宽幅SAR系统下的方位多通道运动目标成像算法研究. 电子与信息学报,

    8. [8]

      王超, 王岩飞, 王琦, 詹学丽. 基于回波序列最小二乘拟合的高分辨率SAR运动目标速度估计. 电子与信息学报,

    9. [9]

      吴元. 一种基于参数更新的机载SAR图像目标定位方法. 电子与信息学报,

    10. [10]

      秦华标, 曹钦平. 基于FPGA的卷积神经网络硬件加速器设计. 电子与信息学报,

    11. [11]

      王鑫, 李可, 宁晨, 黄凤辰. 基于深度卷积神经网络和多核学习的遥感图像分类方法. 电子与信息学报,

    12. [12]

      王巍, 周凯利, 王伊昌, 王广, 袁军. 基于快速滤波算法的卷积神经网络加速器设计. 电子与信息学报,

    13. [13]

      袁野, 贾克斌, 刘鹏宇. 基于深度卷积神经网络的多元医学信号多级上下文自编码器. 电子与信息学报,

    14. [14]

      刘静, 刘涵, 黄开宇, 苏立玉. 基于自动秩估计的黎曼优化矩阵补全算法及其在图像补全中的应用. 电子与信息学报,

    15. [15]

      王斐, 吴仕超, 刘少林, 张亚徽, 魏颖. 基于脑电信号深度迁移学习的驾驶疲劳检测. 电子与信息学报,

    16. [16]

      刘政怡, 段群涛, 石松, 赵鹏. 基于多模态特征融合监督的RGB-D图像显著性检测. 电子与信息学报,

    17. [17]

      吴培良, 杨霄, 毛秉毅, 孔令富, 侯增广. 一种视角无关的时空关联深度视频行为识别方法. 电子与信息学报,

    18. [18]

      刘亚波, 刘霖, 童智勇, 喻忠军. S波段高分辨宽幅SAR辐射定标及误差分析方法. 电子与信息学报,

    19. [19]

      潘洁, 王帅, 李道京, 卢晓春. 基于方向图和多普勒相关系数的天基阵列SAR通道相位误差补偿方法. 电子与信息学报,

    20. [20]

      杨磊, 李埔丞, 李慧娟, 方澄. 稳健高效通用SAR稀疏特征增强算法. 电子与信息学报,

  • 图 1  LeNet-5网络的结构示意图

    图 2  ILSVRC历年的冠军成绩

    图 3  深度网络和深度卷积网络在雷达图像领域发表的文章数示意图

    表 1  光学图像和雷达图像的差异

    特性光学图像雷达图像
    波段可见光,红外微波段
    信号形式多波段灰度信息单波段复信号
    成像原理能量聚焦积累相位相干积累
    尺度特性和成像距离有关目标尺寸不随成像距离变化
    成像方向俯仰角-方位角距离向-方位角
    下载: 导出CSV

    表 2  部分典型网络的参数总结

    LeNet5AlexNetOverfeatfastVGG16GoogleNetV1ResNet50
    输入图像尺寸28×28227×227231×231224×224224×224224×224
    卷积层数量255135753
    全连接层数量233311
    卷积核大小53,5,113,5,1131,3,5,71,3,7
    步长11,41,411,21,2
    权值参数数量60 k61 M146 M138 M7 M25.5 M
    乘积运算数量341 k724 M2.8 G15.5 G1.43 G3.9 G
    Top-5误差N/A16.414.27.46.75.25
    下载: 导出CSV

    表 3  MSTAR数据集的目标类型和样本数量

    数据集2S1BMP2BRD M2BTR 60BTR 70D7T62T72ZIL 131ZSU 234
    训练集299233298256233299299298299299
    测试集274587274195196274196274274274
    下载: 导出CSV

    表 4  常见数据增强技术

    名称主要方法
    旋转变换将图像旋转一定角度
    翻转变换沿水平或垂直方向翻转图像
    缩放变换放大或缩小图像
    平移变换在图像平面上对图像进行平移
    尺度变换对图像按照置顶的尺度因子进行缩放,改变图像内容的大小或模糊程度
    反射变换对称变换,包括轴反射变换和镜面反射变换
    噪声扰动在图像内增加噪声,如指数噪声,高斯噪声,瑞利噪声,椒盐噪声等
    下载: 导出CSV

    表 5  基于CNN的目标检测方法对比

    方法提出场合核心思想MAP(%)主要特点
    候选窗方法RCNNECCV 2014选择搜索方法生成候选窗66.0训练分多个阶段,每个候选窗都需要用CNN处理,占用磁盘空间大,处理效率低
    Fast RCNNICCV2015加入了SPPnet70.0选择搜索方法生成候选窗,耗时长,无法满足实时应用
    Faster RCNNNIPS2015提出了RPN网络,融合区域生成与CNN73.2性能与速度较好的折中,但区域生成方式计算量依然很大,不能实时处理
    R-FCNNIPS2016RPN+位置敏感的预测层+ROI polling+投票决策层76.6速度比faster RCNN快,且精度相当
    回归方法YOLOCVPR2016将检测问题变为回归问题57.9没有区域生成步骤,网格回归的定位性能较弱,检测精度不高。
    SSDECCV2016YOLO+Proposal+多尺度73.9速度非常快,性能也不错
    下载: 导出CSV

    表 6  CNN在雷达图像识别应用进展的思想与方法概要

    提升类型主要思想引用文献和方法概要说明
    快速算法快速寻优预训练文献[47]:带动量小批量随机梯度下降,快速寻找全局最优点
    文献[45]:预训练较浅卷积网络,实现无监督快速检测。
    文献[53]:用大样本数据对卷积网络进行预训练
    用其他结构取代全连接层文献[40,47]:低自由度稀疏连通卷积结构
    文献[39]:SVM代替FC
    文献[53]:用超限学习机替换FC
    抽取特征再训练文献[54]:先抽取特征再训练的两步快速训练方法
    提升算法提高网络的泛化能力文献[47]:Dropout和早期停止
    文献[52]:将卷积层与二维PCA方法结合
    代价函数改进文献[46]:代价函数中引入类别可分性度量提高类别区分能力
    含噪样本训练文献[49]:基于概率转移模型增强含噪标记下分类模型鲁棒性。
    扩展算法迁移学习文献[26,53,55]:大样本预训练,迁移学习加快训练速度
    CAD模型仿真文献[56]: 采用CAD模型目标仿真解决SAR真实数据有限问题
    文献[57]: CAD模型生成不同方位和俯仰角度的HRRP图像
    预处理提升信息的利用率文献[41]:形态学成分分析预处理提升性能
    文献[58]:采用去噪自编码器预训练
    小样本深度训练网络文献[42,44]:卷积高速公路单元在小样本条件下训练深度网络
    文献[58]:无监督和有监督训练结合,应对标签数据有限情况
    下载: 导出CSV
  • 加载中
图(3)表(6)
计量
  • PDF下载量:  97
  • 文章访问数:  618
  • HTML全文浏览量:  348
文章相关
  • 通讯作者:  贺丰收, hefengshou1979@163.com
  • 收稿日期:  2018-09-18
  • 录用日期:  2019-02-18
  • 网络出版日期:  2019-03-21
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

/

返回文章