SAR影像洪水淹没范围深度学习提取方法

doi:10.16265/j.cnki.issn1003-3033.2022.04.026

中国安全科学学报 ›› 2022, Vol. 32 ›› Issue (4): 177-184.doi: 10.16265/j.cnki.issn1003-3033.2022.04.026

SAR影像洪水淹没范围深度学习提取方法

郭玮¹^,²(), 袁宏永², 薛明¹, 魏平岩¹

¹ 应急管理部大数据中心,北京 100010
² 清华大学工程物理系,北京 100084

收稿日期:2021-12-15 修回日期:2022-03-17 出版日期:2022-04-28 发布日期:2022-10-28
作者简介:
郭玮 (1986—),男,河北保定人,博士,工程师,主要从事自然灾害遥感监测、人群疏散、公共安全等方面的工作。E-mail: david_science@163.com。
袁宏永, 教授。
薛明, 高级工程师。
魏平岩, 高级工程师。
基金资助:
国家重点研发计划项目(2018YFC0807000)

Flood inundation area extraction method of SAR images based on deep learning

GUO Wei¹^,²(), YUAN Hongyong², XUE Ming¹, WEI Pingyan¹

¹ Big Data Center of Ministry of Emergency Management, Beijing 100010, China
² Department of Engineering Physics, Tsinghua University, Beijing 100084, China

Received:2021-12-15 Revised:2022-03-17 Online:2022-04-28 Published:2022-10-28

摘要/Abstract

摘要：

为提高洪涝灾害应急救援辅助决策能力,快速提取洪水淹没范围,提出一种基于深度学习的合成孔径雷达(SAR)影像洪水淹没范围提取方法。首先,基于DeepLab v3+模型,建立合成孔径雷达(SAR)影像洪水淹没范围提取模型;然后,针对SAR影像标注数据获取困难的问题,提出一种基于光学影像水体指数的半自动化标注样本制作方法,该方法大幅度减少标注工作的人工量和所需时间;最后,采用Sentinel-1卫星影像验证洪水淹没范围提取模型准确度。结果表明:提出的洪水淹没范围提取模型对复杂地表适应性较强;相比于自适应阈值方法,其识别精度更高,对遥感影像中的水体边缘、小面积水体、细长线状水体识别效果更优,该模型平均交并比为0.83。

关键词: 合成孔径雷达(SAR), 洪水淹没范围, 深度学习, DeepLab v3+模型, 水体提取

Abstract:

In order to improve decision-making ability for flood disaster emergency rescue and quickly extract flood inundation areas, an extraction method of SAR images based on deep learning was proposed. Firstly, flood inundation area extraction model of SAR images was established based on DeepLab v3+ model. Then, considering difficulty in obtaining labeled samples of SAR images, a semi-automatic sample making method based on optical image water index was proposed, which greatly reduced the labor and time required for annotation. Lastly, Sentinel-1 images were used for experimental analysis to verify the model's accuracy. The results show that the proposed extraction model has strong adaptability to complex surfaces. Compared with the adaptive threshold method, it features higher recognition accuracy, and better recognition effect of water edge, small area water body and thin and long linear water body in remote sensing image, with an mean Intersection over Union of 0.83.

Key words: synthetic aperture radar (SAR), flood inundation area, deep learning, DeepLab v3+ model, water extraction

郭玮, 袁宏永, 薛明, 魏平岩. SAR影像洪水淹没范围深度学习提取方法[J]. 中国安全科学学报, 2022, 32(4): 177-184.

GUO Wei, YUAN Hongyong, XUE Ming, WEI Pingyan. Flood inundation area extraction method of SAR images based on deep learning[J]. China Safety Science Journal, 2022, 32(4): 177-184.

图/表 6

图1

图2

图3

图4

图5

图6

参考文献 22

[1]	陈超, 闫艳. 应用灾害数字孪生体的应急预案演练系统[J]. 中国安全科学学报, 2021, 31(7):90-96. doi: 10.16265/j.cnki.issn 1003-3033.2021.07.013
	CHEN Chao, YAN Yan. Emergency plan drilling system based on disaster digital twin[J]. China Safety Science Journal, 2021, 31(7):90-96. doi: 10.16265/j.cnki.issn 1003-3033.2021.07.013
[2]	王喆, 孔维磊, 方丹辉, 等. 基于贝叶斯网络的城镇洪涝应急情景推演研究[J]. 中国安全科学学报, 2021, 31(6):182-188. doi: 10.16265/j.cnki.issn 1003-3033.2021.06.024
	WANG Zhe, KONG Weilei, FANG Danhui, et al. Research on urban flood and waterlog emergency scenario deduction based on Bayesian network[J]. China Safety Science Journal, 2021, 31(6):182-188. doi: 10.16265/j.cnki.issn 1003-3033.2021.06.024
[3]	苏龙飞, 李振轩, 高飞, 等. 遥感影像水体提取研究综述[J]. 国土资源遥感, 2021, 33(1):9-19.
	SU Longfei, LI Zhenxuan, GAO Fei, et al. A review of remote sensing image water extraction[J]. Remote Sensing for Land and Resources, 2021, 33(1):9-19.
[4]	徐鹏杰, 邓磊. 遥感技术在减灾救灾中的应用[J]. 遥感技术与应用, 2011, 26(4):512-519.
	XU Pengjie, DENG Lei. Application of remote sensing in the field of disaster reduction and response[J]. Remote Sensing Technology and Application, 2011, 26(4):512-519.
[5]	于胜利, 李茂, 王保江, 等. 卫星遥感技术在铁路沿线环境防护中的应用[J]. 中国安全科学学报, 2018, 28(增2):88-92.
	YU Shengli, LI Mao, WANG Baojiang, et al. Application of satellite remote sensing technology in environmental protection along railway lines[J]. China Safety Science Journal, 2018, 28(S2):88-92. doi: 10.16265/j.cnki.issn1003-3033.2018.S2.016
[6]	MCFEETERS S K. The use of the normalized difference water index (NDWI) in the delineation of open water features[J]. International Journal of Remote Sensing, 1996, 17(7):1425-1432. doi: 10.1080/01431169608948714
[7]	闫霈, 张友静, 张元. 利用增强型水体指数(EWI)和GIS去噪音技术提取半干旱地区水系信息的研究[J]. 遥感信息, 2007(6):62-67.
	YAN Pei, ZHANG Youjing, ZHANG Yuan. A study on information extraction of water system in semi-arid regions with the enhanced water index (EWI) and GIS based noise remove techniques[J]. Remote Sensing, 2007(6):62-67.
[8]	FEYISA G L, MEILBY H, FENSHOLT R, et al. Automated water extraction index: a new technique for surface water mapping using Landsat imagery[J]. Remote Sensing of Environment, 2014, 140: 23-35. doi: 10.1016/j.rse.2013.08.029
[9]	李艳华, 丁建丽, 闫人华. 基于国产GF-1遥感影像的山区细小水体提取方法研究[J]. 资源科学, 2015, 37(2):408-416.
	LI Yanhua, DING Jianli, YAN Renhua. Extraction of small river information based on China-made GF-1 remote sense images[J]. Resources Science, 2015, 37(2):408-416.
[10]	PAUL A, TRIPATHI D, DUTTA D. Application and comparison of advanced supervised classifiers in extraction of water bodies from remote sensing images[J]. Sustainable Water Resources Management, 2018, 4(4):905-919. doi: 10.1007/s40899-017-0184-6
[11]	杨文亮, 杨敏华, 祁洪霞. 利用BP神经网络提取TM影像水体[J]. 测绘科学, 2012, 37(1):148-150.
	YANG Wenliang, YANG Minhua, QI Hongxia. Water body extracting from TM image based on BPNN[J]. Science of Surveying and Mapping, 2012, 37(1):148-150.
[12]	李士进, 王声特. 基于混合特征空间MRF(Markov Random Filed)模型的高分辨率遥感影像水体提取[J]. 南京师大学报:自然科学版, 2017, 40(1):13-19.
	LI Shijin, WANG Shengte. A new algorithm based on hybrid feature space MRF (Markov Random Filed) model for water information extraction from high resolution remote sensing imagery[J]. Journal of Nanjing Normal University :Natural Science Edition, 2017, 40(1):13-19.
[13]	REZAEE M, MAHDIANPARI M, ZHANG Yun, et al. Deep convolutional neural network for complex wetland classification using optical remote sensing imagery[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018, 11(9):1-10.
[14]	贾诗超, 薛东剑, 李成绕, 等. 基于Sentinel-1数据的水体信息提取方法研究[J]. 人民长江, 2019, 50(2):213-217.
	JIA Shichao, XUE Dongjian, LI Chengrao, et al. Study on new method for water area information extraction based on Sentinel-1 data[J]. Yangtze River, 2019, 50(2):213-217.
[15]	吴效勇, 王晓青, 丁玲, 等. 基于光学与SAR影像的老挝溃坝洪涝灾害监测与评估[J]. 灾害学, 2020, 35(1):211-215.
	WU Xiaoyong, WANG Xiaoqing, DING Ling, et al. Monitoring and assessment of dam break flood disaster in Laos based on SAR images[J]. Journal of Catastrophology, 2020, 35(1):211-215.
[16]	湛南渝, 李小涛, 路京选, 等. 基于SLIC的哨兵1号雷达数据水体信息提取[J]. 人民长江, 2020, 51(4):213-217,225.
	ZHAN Nanyu, LI Xiaotao, LU Jingxuan, et al. Water body information extraction by Sentinel-1 radar data based on SLIC[J]. Yangtze River, 2020, 51(4):213-217,225.
[17]	陈媛媛, 郑加柱, 魏浩翰, 等. 基于Sentinel-1A数据的南京市水体信息提取[J]. 地理空间信息, 2020, 18(9):62-65,67.
	CHEN Yuanyuan, ZHENG Jiazhu, WEI Haohan, et al. Water information extraction of Nanjing city based on Sentinel-1A data[J]. Geospatial Information, 2020, 18(9):62-65,67.
[18]	李玲玉, 张毅. 一种基于新差异算子和纹理的SAR图像水体变化检测算法[J]. 计算机与现代化, 2020(4):85-94.
	LI Lingyu, ZHANG Yi. A water change detection method of SAR images based on new difference operator and texture[J]. Computer and Modernization, 2020(4):85-94.
[19]	PAI M M M, MEHROTRA V, VERMA U, et al. Improved semantic segmentation of water bodies and land in SAR images using generative adversarial networks[J]. International Journal of Semantic Computing, 2020, 14(1):55-69. doi: 10.1142/S1793351X20400036
[20]	FILIPPONI F. Sentinel-1 GRD Preprocessing workflow[C]. 3rd International Electronic Conference on Remote Sensing, 2019: DOI: 10.3390/ECRS-3-06201. doi: 10.3390/ECRS-3-06201
[21]	李成绕, 薛东剑, 张露, 等. 基于Sentinel-1A卫星SAR数据的水体提取方法研究[J]. 地理空间信息, 2018, 16(1):38-40.
	LI Chengrao, XUE Dongjian, ZHANG Lu, et al. Research on water extraction method based on Sentinel-1A satellite SAR data[J]. Geospatial Information, 2018, 16(1):38-40.
[22]	CHEN Liangchieh, ZHU Yukun, PAPANDREOU G, et al. Encoder-decoder with atrous separable convolution for semantic image segmentation[C]. 15th European Conference on Computer Vision, 2018:833-851.

SAR影像洪水淹没范围深度学习提取方法

Flood inundation area extraction method of SAR images based on deep learning

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 6

参考文献 22

相关文章 8

编辑推荐

Metrics

本文评价

[1]	黄珍珍, 肖硕, 王钰, 陈伟, 王升志, 江海峰. 铁路工人人体行为识别模型[J]. 中国安全科学学报, 2022, 32(6): 17-22.
[2]	张萌, 韩豫, 刘泽锋. 深度学习下建筑工人高空安全防护装备检测方法[J]. 中国安全科学学报, 2022, 32(5): 140-146.
[3]	段照斌, 杜海龙, 张鹏. 基于QAR2Vec模型的QAR数据特征提取[J]. 中国安全科学学报, 2021, 31(1): 145-152.
[4]	赵欣然, 张琪, 王卫东, 徐志强. 可燃性粉尘云的图像检测方法[J]. 中国安全科学学报, 2020, 30(4): 8-13.
[5]	徐丹, 代勇, 纪军红. 基于卷积神经网络的驾驶人行为识别方法研究[J]. 中国安全科学学报, 2019, 29(10): 12-17.
[6]	汪洋, 王俊刚. 基于深度学习算法的铁路列车运行安全检测^*[J]. 中国安全科学学报, 2018, 28(S2): 41-45.
[7]	佟瑞鹏, 陈策, 崔鹏程, 傅贵, 安宇. 基于深度学习的施工安全泛场景数据获取方法[J]. 中国安全科学学报, 2017, 27(5): 1-6.
[8]	佟瑞鹏, 崔鹏程. 基于深度学习的不安全因素识别和交互分析[J]. 中国安全科学学报, 2017, 27(4): 49-54.