一种基于密集深度分离卷积的SAR图像水域分割算法

张金松 邢孟道 孙光才

张金松, 邢孟道, 孙光才. 一种基于密集深度分离卷积的SAR图像水域分割算法[J]. 雷达学报, 2019, 8(3): 400–412. doi: 10.12000/JR19008
引用本文: 张金松, 邢孟道, 孙光才. 一种基于密集深度分离卷积的SAR图像水域分割算法[J]. 雷达学报, 2019, 8(3): 400–412. doi: 10.12000/JR19008
ZHANG Jinsong, XING Mengdao, and SUN Guangcai. A water segmentation algorithm for SAR image based on dense depthwise separable convolution[J]. Journal of Radars, 2019, 8(3): 400–412. doi: 10.12000/JR19008
Citation: ZHANG Jinsong, XING Mengdao, and SUN Guangcai. A water segmentation algorithm for SAR image based on dense depthwise separable convolution[J]. Journal of Radars, 2019, 8(3): 400–412. doi: 10.12000/JR19008

一种基于密集深度分离卷积的SAR图像水域分割算法

doi: 10.12000/JR19008
基金项目: 国家重点研发计划(2017YFC1405600),国家自然科学基金创新群体基金(61621005)
详细信息
    作者简介:

    张金松(1995–),男,山东德州人,西安电子科技大学信号与信息处理专业博士研究生,研究方向为SAR图像解译,深度学习及SAR成像。E-mail: jinsongxd@163.com

    邢孟道(1975–),男,浙江嵊州人。西安电子科技大学教授,博士生导师,主要研究方向为雷达成像、目标识别和天波超视距雷达信号处理。E-mail: xmd@xidian.edu.cn

    孙光才(1984–),男,湖北孝感汉川人。西安电子科技大学副教授,博士生导师,主要研究方向为多通道波束指向 SAR 成像和 SAR 动目标成像。E-mail: rsandsgc@126.com

    通讯作者:

    张金松 jinsongxd@163.com

  • 中图分类号: TN958

A Water Segmentation Algorithm for SAR Image Based on Dense Depthwise Separable Convolution

Funds: The State Key Research Development Program (2017YFC1405600), The Foundation for Innovative Research Groups of the National Natural Science Foundation of China (61621005)
More Information
  • 摘要: SAR图像的水域分割在舰船目标检测、灾害监测等军事和民用领域具有重要意义。针对传统水域分割算法鲁棒性差、难以准确进行分割等问题,该文首先建立了基于高分三号的SAR图像水域分割数据集,并基于深度学习技术提出了基于密集深度分离卷积的分割网络架构,该网络以SAR图像作为输入,通过密集分离卷积和扩张卷积提取图像高维特征,并构造基于双线性插值的上采样解码模块用于输出分割结果。在水域分割数据集上的实验结果表明,与传统方法相比,该方法不仅在分割准确度上有大幅提高,在算法的鲁棒性和分割速度上也具有部分优势,具备较好的工程实用价值。

     

  • 图  1  水域分割对舰船检测意义

    Figure  1.  Significance of water segmentation for ship detection

    图  2  常规卷积和深度分离卷积结构对比图

    Figure  2.  Comparison of conventional convolution and depthwise separable convolution

    图  3  常规卷积和扩张卷积结构对比图

    Figure  3.  Comparison of conventional convolution and dilated convolution

    图  4  特征提取网络结构示意图

    Figure  4.  The structure of feature extraction network

    图  5  基于编码-解码结构的SAR图像水域分割网络示意图

    Figure  5.  The structure of encoder-decoder network for water segmentation

    图  6  分割网络训练结果示意图

    Figure  6.  Training results of segmentation network

    图  7  网络分割结果示意图

    Figure  7.  Segmentation results of segmentation network

    图  8  各方法分割结果对比图

    Figure  8.  Segmentation results of different methods

    图  9  不同工作模式分割结果对比图

    Figure  9.  Segmentation results of different imaging modes

    图  10  不同极化方式分割结果对比图

    Figure  10.  Segmentation results of different imaging polarizations

    表  1  高分三号成像模式

    Table  1.   The imaging modes of GF3 satellite

    工作模式分辨率(m)极化方式成像幅宽(km)
    滑块聚束(SL)1单极化10
    超精细条带(UFS)3单极化30
    精细条带1(FSI)5双极化50
    精细条带2(FSII)10双极化100
    标准条带1(QPSI)8全极化30
    下载: 导出CSV

    表  2  数据集图像组成

    Table  2.   The composition of dataset

    图像类型数量图像尺寸(像素)
    原始图像10$ \approx 10,000 \times 10,000$
    裁剪图像480$513 \times 513$
    扩充图像21180$513 \times 513$
    下载: 导出CSV

    表  3  数据扩充对分割性能的影响

    Table  3.   Segmentation effects of data augmentation

    扩充方法像素准确度交并比
    未扩充0.95690.9497
    旋转0.98060.9758
    翻转0.96200.9603
    旋转+翻转0.98870.9844
    下载: 导出CSV

    表  4  网络结构对分割性能的影响

    Table  4.   Segmentation effects of network structure

    连接方式像素准确度交并比
    直连0.93120.9289
    仅残差0.97030.9681
    仅密集0.96790.9638
    残差+密集0.98870.9844
    下载: 导出CSV

    表  5  各水域分割算法性能对比

    Table  5.   Segmentation performance of different methods

    方法类别具体方法像素准确度交并比小图速度(s)大图速度(s)
    传统方法FCM0.67100.46448.24206.0
    MRF0.59610.54302.2957.25
    OTSU0.63030.61080.061.50
    Levelset0.71340.68683.4185.25
    深度学习Unet0.95330.94960.071.75
    DeepLabv3+0.96720.95660.102.50
    所提方法0.98870.98440.143.50
    理想值1.00001.0000
    下载: 导出CSV

    表  6  本文方法对多模式多极化下SAR图像的IoU分割结果

    Table  6.   IoU under multi-mode and multi-polarization by the proposed method

    工作模式/极化方式HHHVVHVV
    SL (1 m)0.9844
    UFS (3 m)0.9240
    FSI (5 m)0.93650.9542
    FSII (10 m)0.95490.9454
    QPSI (8 m)0.96050.96840.96860.9717
    下载: 导出CSV
  • [1] 吴一戎. 多维度合成孔径雷达成像概念[J]. 雷达学报, 2013, 2(2): 135–142. doi: 10.3724/SP.J.1300.2013.13047

    WU Yirong. Concept of multidimensional space joint-observation SAR[J]. Journal of Radars, 2013, 2(2): 135–142. doi: 10.3724/SP.J.1300.2013.13047
    [2] 艾加秋, 齐向阳, 禹卫东. 改进的SAR图像双参数CFAR舰船检测算法[J]. 电子与信息学报, 2009, 31(12): 2881–2885. doi: 10.3724/SP.J.1146.2008.01707

    AI Jiaqiu, QI Xiangyang, and YU Weidong. Improved two parameter CFAR ship detection algorithm in SAR images[J]. Journal of Electronics &Information Technology, 2009, 31(12): 2881–2885. doi: 10.3724/SP.J.1146.2008.01707
    [3] 牛世林, 郭拯危, 李宁, 等. 星载SAR水域分割研究进展与趋势分析[J]. 聊城大学学报: 自然科学版, 2018, 31(2): 72–86.

    NIU Shilin, GUO Zhengwei, LI Ning, et al. Research progress and trend analysis of water extraction by spaceborne SAR[J]. Journal of Liaocheng University:Natural Science Edition, 2018, 31(2): 72–86.
    [4] 安成锦, 牛照东, 李志军, 等. 典型Otsu算法阈值比较及其SAR图像水域分割性能分析[J]. 电子与信息学报, 2010, 32(9): 2215–2219. doi: 10.3724/SP.J.1146.2009.01426

    AN Chengjin, NIU Zhaodong, LI Zhijun, et al. Otsu threshold comparison and SAR water segmentation result analysis[J]. Journal of Electronics &Information Technology, 2010, 32(9): 2215–2219. doi: 10.3724/SP.J.1146.2009.01426
    [5] 李智, 曲长文, 周强, 等. 基于SLIC超像素分割的SAR图像海陆分割算法[J]. 雷达科学与技术, 2017, 15(4): 354–358. doi: 10.3969/j.issn.1672-2337.2017.04.003

    LI Zhi, QU Changwen, ZHOU Qiang, et al. A sea-land segmentation algorithm of SAR image based on the SLIC superpixel division[J]. Radar Science and Technology, 2017, 15(4): 354–358. doi: 10.3969/j.issn.1672-2337.2017.04.003
    [6] AMITRANO D, CIERVO F, DI MARTINO G, et al. Modeling watershed response in semiarid regions with high-resolution synthetic aperture radars[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014, 7(7): 2732–2745. doi: 10.1109/jstars.2014.2313230
    [7] OTSU N. A threshold selection method from gray-level histograms[J]. IEEE Transactions on Systems, Man, and Cybernetics, 1979, 9(1): 62–66. doi: 10.1109/TSMC.1979.4310076
    [8] SUI H G and XU C. Automatic extraction of water in high-resolution SAR images based on multi-scale level set method and Otsu algorithm[C]. Proceedings of International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Melbourne, Australia, 2012: 453–457. doi: 10.5194/isprsarchives-XXXIX-B7-453-2012.
    [9] LIU Chun, YANG Jian, YIN Junjun, et al. Coastline detection in SAR images using a hierarchical level set segmentation[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2016, 9(11): 4908–4920. doi: 10.1109/jstars.2016.2613279
    [10] 侯彪, 胡育辉, 焦李成. SAR图像水域的改进Shearlet边缘检测[J]. 中国图象图形学报, 2010, 15(10): 1549–1554. doi: 10.11834/jig.20101019

    HOU Biao, HU Yuhui, and JIAO Licheng. Improved shearlet edge detection for waters of SAR images[J]. Journal of Image and Graphics, 2010, 15(10): 1549–1554. doi: 10.11834/jig.20101019
    [11] LIU Zhongling, LI Fei, LI Ning, et al. A novel region-merging approach for coastline extraction from sentinel-1A IW mode SAR imagery[J]. IEEE Geoscience and Remote Sensing Letters, 2016, 13(3): 324–328. doi: 10.1109/LGRS.2015.2510745
    [12] SHELHAMER E, LONG J, and DARRELL T. Fully convolutional networks for semantic segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(4): 640–651. doi: 10.1109/TPAMI.2016.2572683
    [13] RONNEBERGER O, FISCHER P, and BROX T. U-Net: Convolutional networks for biomedical image segmentation[C]. Proceedings of the 18th International Conference on Medical Image Computing and Computer-assisted Intervention, Munich, Germany, 2015: 234–241. doi: 10.1007/978-3-319-24574-4_28.
    [14] LIN T Y, DOLLÁR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]. Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 936–944. doi: 10.1109/CVPR.2017.106.
    [15] ZHAO Hengshuang, SHI Jianping, QI Xiaojuan, et al. Pyramid scene parsing network[C]. Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 6230–6239. doi: 10.1109/CVPR.2017.660.
    [16] CHEN L C, PAPANDREOU G, KOKKINOS I, et al. DeepLab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(4): 834–848. doi: 10.1109/TPAMI.2017.2699184
    [17] CHEN L C, PAPANDREOU G, KOKKINOS I, et al. Semantic image segmentation with deep convolutional nets and fully connected CRFs[EB/OL]. arXiv: 1412.7062, 2014.
    [18] 张庆君. 高分三号卫星总体设计与关键技术[J]. 测绘学报, 2017, 46(3): 269–277. doi: 10.11947/j.AGCS.2017.20170049

    ZHANG Qingjun. System design and key technologies of the GF-3 satellite[J]. Acta Geodaetica et Cartographica Sinica, 2017, 46(3): 269–277. doi: 10.11947/j.AGCS.2017.20170049
    [19] 丁赤飚, 刘佳音, 雷斌, 等. 高分三号SAR卫星系统级几何定位精度初探[J]. 雷达学报, 2017, 6(1): 11–16. doi: 10.12000/JR17024

    DING Chibiao, LIU Jiayin, LEI Bin, et al. Preliminary exploration of systematic geolocation accuracy of GF-3 SAR satellite system[J]. Journal of Radars, 2017, 6(1): 11–16. doi: 10.12000/JR17024
    [20] CHOLLET F. Xception: Deep learning with depthwise separable convolutions[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 1800–1807. doi: 10.1109/CVPR.2017.195.
    [21] YU F and KOLTUN V. Multi-scale context aggregation by dilated convolutions[EB/OL]. arXiv preprint arXiv: 1511.07122, 2015.
    [22] SZEGEDY C, VANHOUCKE V, IOFFE S, et al. Rethinking the inception architecture for computer vision[C]. Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, USA, 2016: 2818–2826. doi: 10.1109/CVPR.2016.308.
    [23] HE Kaiming, ZHANG Xiangyu, REN Shaoqing, et al. Deep residual learning for image recognition[C]. Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, USA, 2016: 770–778. doi: 10.1109/CVPR.2016.90.
    [24] HUANG Gao, LIU Zhuang, VAN DER MAATEN L, et al. Densely connected convolutional networks[C]. Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 2261–2269. doi: 10.1109/CVPR.2017.243.
    [25] NAIR V and HINTON G E. Rectified linear units improve restricted boltzmann machines[C]. Proceedings of the 27th International Conference on International Conference on Machine Learning, Haifa, Israel, 2010: 807–814.
    [26] IOFFE S and SZEGEDY C. Batch normalization: Accelerating deep network training by reducing internal covariate shift[EB/OL]. arXiv preprint arXiv: 1502.03167, 2015.
    [27] MIKOLOV T, SUTSKEVER I, CHEN Kai, et al. Distributed representations of words and phrases and their compositionality[C]. Proceedings of the 26th International Conference on Neural Information Processing Systems, Lake Tahoe, USA, 2013: 3111–3119.
    [28] HANSEN L K and SALAMON P. Neural network ensembles[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1990, 12(10): 993–1001. doi: 10.1109/34.58871
    [29] DENG Jia, DONG Wei, SOCHER R, et al. ImageNet: A large-scale hierarchical image database[C]. Proceedings of 2009 IEEE Conference on Computer Vision and Pattern Recognition, Miami, USA, 2009: 248–255. doi: 10.1109/CVPR.2009.5206848.
    [30] VESE L A and CHAN T F. A multiphase level set framework for image segmentation using the Mumford and shah model[J]. International Journal of Computer Vision, 2002, 50(3): 271–293. doi: 10.1023/a:1020874308076.
  • 加载中
图(10) / 表(6)
计量
  • 文章访问数:  3519
  • HTML全文浏览量:  1362
  • PDF下载量:  258
  • 被引次数: 0
出版历程
  • 收稿日期:  2019-01-14
  • 修回日期:  2019-04-08
  • 网络出版日期:  2019-06-01

目录

    /

    返回文章
    返回