收藏 ｜ 设为首页 广告服务 ｜ 联系我们 ｜ 在线留言
高级检索

干旱区地理 >

2020 , Vol. 43 >Issue 3: 750 - 760

DOI: https://doi.org/10.12118/j.issn.1000-6060.2020.03.21

地球信息科学

基于Sentinel-1A合成孔径雷达数据和全卷积网络的城市建设用地监测方法研究

展开

  • 1 北京师范大学地表过程与资源生态国家重点实验室，人与环境系统可持续研究中心，北京 100875； 2 北京师范大学地理科学学部自然资源学院，北京 100875

卢文路（1995-），男，在读博士研究生，主要从事机器学习与景观可持续科学研究.E-mail：wenlu.lu@qq.com

收稿日期: 2019-10-14

  修回日期: 2020-01-25

  网络出版日期: 2020-05-25

基金资助

国家自然科学基金面上项目（41871185，41971271）；国家自然科学基金创新群体项目（41621061）资助

收起

A new method for detecting urban construction land based on Sentinel-1A synthetic aperture radar data and fully convolutional network

Expand

  • 1 Center for HumanEnvironment System Sustainability (CHESS),State Key Laboratory of Earth Surface Processes and 

    Resource Ecology (ESPRE),Faculty of Geographical Science,Beijing Normal University,Beijing   100875,China；  2 School of Natural Resources,Faculty of Geographical Science,Beijing Normal University,Beijing 100875,China

Received date: 2019-10-14

  Revised date: 2020-01-25

  Online published: 2020-05-25

Fold

摘要

Sentinel-1A合成孔径雷达数据不受云、雾等天气条件的影响并具有丰富的纹理信息，为提取城市建设用地信息提供了一种新的数据源。本文发展了一种基于Sentinel-1A合成孔径雷达数据和全卷积网络的城市建设用地监测方法。该方法的优势主要在于可以有效复合不同极化方式下的Sentinel-1A合成孔径雷达数据和综合集成多尺度特征。在甘肃省张掖市甘州区的应用表明，该方法的提取结果总体精度为92.50%，Kappa系数为0.85。与现有方法“KTH-Pavia城市提取器”相比，Kappa系数提高了37.10%，总体精度提高了11.50%。因此，该方法具有良好的应用潜力。

关键词： Sentinel-1A; 哨兵数据; 合成孔径雷达; 城市建设用地; 全卷积网络; 深度学习

本文引用格式

卢文路, 刘志锋, 何春阳, 夏沛 .

基于Sentinel-1A合成孔径雷达数据和全卷积网络的城市建设用地监测方法研究[J]. 干旱区地理, 2020 , 43(3) : 750 -760 . DOI: 10.12118/j.issn.1000-6060.2020.03.21

Abstract

Timely and accurate extraction of urban construction land is essential for assessing the impacts of urban expansion on the environment.Optical remote sensing is susceptible to weather conditions and contains insufficient information on the texture of urban features,and is thus not conducive to urban construction land extraction.Synthetic aperture radar (SAR) is unique in its potential to be used in all weather conditions and at any time.Furthermore,its backscattering and polarimetric information is sensitive to the dielectric and geometric properties of the urban land surface,providing a new means for rapid and accurate extraction of urban construction land.Among existing SAR data,the Sentinel-1A SAR data is widely used due to its free access.Nonetheless,existing methods are poorly adapted to fully utilize the Sentinel1A SAR data,thereby limiting its application.The fully convolutional network is a deep learning network developed on the basis of the convolutional neural network,which adopts “pixel to pixel” image recognition and has become an effective method for extracting urban construction land.This paper aims to develop a method for detecting urban construction land based on Sentinel1A SAR data,and thus a fully convolutional network.Firstly,a fully convolutional network consisting of five Inception modules was developed with reference to GoogleNet.Each Inception module includes three convolutional layers with convolutional kernel sizes of 1×1,3×3,and 5×5 respectively,three corresponding activation layers,and a concatenation layer.The Ganzhou District in Zhangye City,Gansu Province,China,where urban construction land and bare land show similar spectrum features,was used as a case study area to verify our fully convolutional network.The extraction results show that the area of construction land in the Ganzhou District was 4 006 hectares in 2018,accounting for about 20% of the total area.The patch area of urban construction land was generally found to be between 0.01 and 1 hectare.Urban construction land was mainly distributed in the west and northeast,and the circle of 2-4 km from the city center exhibited the most concentrated urban construction land.Our accuracy assessment shows an overall accuracy of 92.50% and a Kappa coefficient of 0.85.By comparing our results to extraction results using the “KTH-Pavia” method (the most widely used method for extracting urban construction land from SAR data),it was found that extraction results based on the fully convolutional network are closer to the real urban construction land in spatial patterns.Furthermore,the overall accuracy and Kappa coefficient were,respectively,11% and 37% higher than the “KTH-Pavia” method.The principal reason for this higher accuracy is that the fully convolution network can better integrate multi-polarization and multi-scale texture information from SAR data.Furthermore,the fully convolutional network contains multiple convolutional structures and supports multi-source data inputs.The method developed in this study has both greater accuracy than existing methods and is applicable to urban construction land extraction based on different SAR data in different regions.It therefore has potential for widespread application.

Key words： Sentinel-1A; sentinel ; data; synthetic aperture radar; urban construction land; fully convolutional network; deep learning

参考文献

［1］国家统计局.中国城市统计年鉴2017［M］.北京：中国统计出版社,2017.［National Bureau of Statistics of the People’s Republic of China.China city statistical yearbook 2017［M］.Beijing:China Statistics Press,2017.］ ［2］刘纪远,匡文慧,张增祥,等.20世纪80年代末以来中国土地利用变化的基本特征与空间格局［J］.地理学报,2014,69(1)：3-14.［LIU Jiyuan,KUANG Wenhui,ZHANG Zengxiang,et al.Spatiotemporal characteristics,patterns and causes of land use changes in China since the late 1980s［J］.Acta Geographica Sinica,2014,69(1)：3-14.］ ［3］匡文慧,刘纪远,陆灯盛.京津唐城市群不透水地表增长格局以及水环境效应［J］.地理学报,2011,66(11)：1486-1496.［KUANG Wenhui,LIU Jiyuan,LU Dengsheng.Pattern of impervious surface change and its effect on water environment in the Beijing-Tianjin-Tangshan metropolitan area［J］.Acta Geographica Sinica,2011,66(11)：1486-1496.］ ［4］王静,周伟奇,许开鹏,等.京津冀地区城市化对植被覆盖度及景观格局的影响［J］.生态学报,2017,37(21)：7019-7029.［WANG Jing,ZHOU Weiqi,XU Kaipeng,et al.Spatiotemporal pattern of vegetation cover and its relationship with urbanization in BeijingTianjin-Hebei megaregion from 2000 to 2010［J］.Acta Ecologica Sinica,2017,37(21):7019-7029.］ ［5］HE C Y,LIU Z F,TIAN J,et al.Urban expansion dynamics and natural habitat loss in China:A multiscale landscape perspective［J］.Global Change Biology,2014,20(9)：2886-2902. ［6］陆大道,陈明星.关于“国家新型城镇化规划(2014—2020)”编制大背景的几点认识［J］.地理学报,2015,70(2)：179-185.［LU Dadao,CHEN Mingxing.Several viewpoints on the background of compiling the “National new urbanization planning (2014-2020)”［J］.Acta Geographica Sinica,2015,70(2)：179-185.］ ［7］LIU Z F,DING M H,HE C Y,et al.The impairment of environmental sustainability due to rapid urbanization in the dryland region of northern China［J］.Landscape and Urban Planning,2019,(187),165-180. ［8］LIU Z F,HE C Y,ZHOU Y Y,et al.How much of the worl[JP8]d’[JP]s land has been urbanized,really a hierarchical framework for avoiding confusion［J］.Landscape Ecology,2014,29(5)：763-771. ［9］杜培军.城市遥感的研究动态与发展趋势——“城市遥感”专栏导读［J］.地理与地理信息科学,2018,34(3)：1-4.［DU Peijun.Progress and trends of urban remote sensing:Reading guidance for the special column on urban remote sensing［J］.Geography and Geo-Information Science,2018,34(3)：1-4.］ ［10］WU J G.Urban ecology and sustainability:The state of the science and future directions［J］.Landscape & Urban Planning,2014,125(2)：209-221. ［11］何春阳,史培军,陈晋,等.北京地区土地利用/覆盖变化研究［J］.地理研究,2001,(6)：679-687,772.［HE Chunyang,SHI Peijun,CHEN Jin,et al.A study on land use/cover change in Beijing area［J］.Geographical Research,2001,(6)：679-687,772.］[JP] ［12］宋金超,李新虎,吝涛,等.基于夜晚灯光数据和Google Earth的城市建成区提取分析［J］.地球信息科学学报,2015,17(6)：750-756.［SONG Jinchao,LI Xinhu,LIN Tao,et al.A method of extracting urban built-up area based on DMSP/OLS nighttime data and Google Earth［J］.Journal of Geo-information Science,2015,17(6)：750-756.］ ［13］卓莉,李强,史培军,等.基于夜间灯光数据的中国城市用地扩展类型［J］.地理学报,2006,(2)：169-178.［ZHUO Li,LI Qiang,SHI Peijun,et al.Identification and characteristics analysis of urban land expansion types in China in the 1990s using DMSP/OLS data［J］.Acta Geographica Sinica,2006,(2)：169-178.］ ［14］黄世奇.合成孔径雷达成像及其图像处理［M］.北京：科学出版社,2015.［HUANG Shiqi.Synthetic aperture radar imaging and image processing［M］.Beijing:Science Press,2015.］ ［15］张风丽,邵芸.城市目标高分辨率SAR遥感监测技术研究进展［J］.遥感技术与应用,2010,25(3)：415-422.［ZHANG Fengli,SHAO Yun.Urban target monitoring using high resolution SAR data［J］.Remote Sensing Technology and Application,2010,25(3)：415-422.］ ［16］张庆君,韩晓磊,刘杰.星载合成孔径雷达遥感技术进展及发展趋势［J］.航天器工程,2017,26(6)：1-8.［ZHANG Qingjun,HAN Xiaolei,LIU Jie.Technology progress and development trend of spaceborne synthetic aperture radar remote sensing［J］.Spacecraft Engineering,2017,26(6)：1-8.］ ［17］ZHANG H S,LIN H,WANG Y P.A new scheme for urban impervious surface classification from SAR images［J］.ISPRS Journal of Photogrammetry and Remote Sensing,2018,(139)：103-118. ［18］车美琴,阿里木•赛买提,杜培军,等.利用旋转不变特征提取全极化SAR影像人工地物［J］.遥感学报,2016,20(2)：303-314.［CHE Meiqin,SAMAT A,DU Peijun,et al.Urban man-made target extraction from quad-polsar imagery with roll-invariant parameters［J］.Journal of Remote Sensing,2016,20(2):303-314.］ ［19］NIU X,BAN Y.Multitemporal RADARSAT-2 polarimetric sar data for urban land cover classification using object-based support vector machine and rule-based approach［J］.International Journal of Remote Sensing,2012,34(1)：1-26. ［20］张鸿生,林殷怡,王挺,等.融合光学与雷达遥感数据的城市不透水面提取方法［J］.地理与地理信息科学,2018,34(3)：39-46,129.［ZHANG Hongsheng,LIN Yinyi,WANG Ting,et al.Fusing optical and SAR remote sensing data for urban impervious surface estimation［J］.Geography and Geo-Information Science,2018,34(3)：39-46,129.］ ［21］SANTORO M ,ASKNE J I H ,WEGMIILLER U,et al.Observations,modeling,and applications of ERS-ENVISAT coherence over land surfaces［J］.IEEE Transactions on Geoscience and Remote Sensing,2007,45(8)：2600-2611. ［22］DEKKER R J.Texture analysis and classification of ERS SAR images for map updating of urban areas in the Netherlands［J］.Geoscience & Remote Sensing IEEE Transactions,2003,41(9)：1950-1958. ［23］DEL FRATE F,PACIFICI F,SOLIMINI D.Monitoring urban land cover in Rome,Italy,and its changes by single-polarization multitemporal SAR images［J］.IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2008,1(2)：87-97. ［24］王纬华,江涛,张永红,等.基于高分SAR影像的城市用地变化检测［J］.测绘与空间地理信息,2016,39(7)：141-143,147.［WANG Weihua,JIANG Tao,ZHANG Yonghong,et al.Detection of urban land use changes based on high resolution SAR images［J］.Geomatics & Spatial Information Technology,2016,39(7)：141-143,147.］ ［25］GENG J,WANG H Y,FAN J C,et al.Deep supervised and contractive neural network for SAR Image classification［J］.IEEE Transactions on Geoscience and Remote Sensing,2017,55(4)：2442-2459. ［26］陈筠力,李威.国外SAR卫星最新进展与趋势展望［J］.上海航天,2016,33(6)：1-19.［CHEN Junli,LI Wei.Recent advances and trends of SAR satellites in foreign countries［J］.Aerospace Shanghai,2016,33(6)：1-19.］ ［27］欧阳伦曦,李新情,惠凤鸣,等.哨兵卫星Sentinel-1A数据特性及应用潜力分析［J］.极地研究,2017,29(2)：286-295.［ OUYANG Lunxi,LI Xinqing,HUI Fengming,et al.Sentinel-1A data products’ characteristics and the potential applications［J］.Chinese Journal of Polar Research,2017,29(2)：286-295.］ ［28］杨魁,杨建兵,江冰茹.Sentinel-1卫星综述［J］.城市勘测,2015,(2)：24-27.［YANG Kui,YANG Jianbing,JIANG Bingru.Sentinel-1 satellite overview［J］.Urban Geotechnical Investigation & Surveying,2015,(2)：24-27.］ ［29］TORRES R,SNOEIJ P,GEUDTNER D,et al.GMES Sentinel-1 mission［J］.Remote Sensing of Environment,2012,120(6)：9-24. ［30］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)：38-40,7.［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)：38-40,7.］ ［31］汤玲英,刘雯,杨东,等.基于面向对象方法的Sentinel-1A SAR在洪水监测中的应用［J］.地球信息科学学报,2018,20(3)：377-384.［TANG Lingying,LIU Wen,YANG Dong,et al.Flooding monitoring application based on the object-oriented method and Sentinel1A SAR data［J］.Journal of Geo-information Science,2018,20(3)：377-384.］ ［32］向海燕,罗红霞,刘光鹏,等.基于Sentinel-1A极化SAR数据与面向对象方法的山区地表覆被分类［J］.自然资源学报,2017,32(12)：2136-2148.［XIANG Haiyan,LUO Hongxia,LIU Guangpeng,et al.Land cover classification in mountain areas based on Sentinel1A polarimetric SAR data and object oriented method［J］.Journal of Natural Resources,2017,32(12)：2136-2148.］ ［33］BAN Y F,WEBBER L,GAMBA P,et al.EO4Urban:Sentinel-1A SAR and Sentinel-2A MSI data for global urban services［C］// Urban Remote Sensing Event.IEEE,2017：1-4. ［34］HEIKO B ,BETH C,CHRISTIAN T,et al.Mapping CORINE land cover from Sentinel-1A SAR and SRTM digital elevation model data using random forests［J］.Remote Sensing,2015,7(11)：14876-14898. ［35］CAO H,ZHANG H,WANG C,et al.Operational builtup areas extraction for cities in China using Sentinel-1 SAR data［J］.Remote Sensing,2018,(10)：874. ［36］GAMBA P,ALDRIGHI M,STASOLLA M.Robust extraction of urban area extents in HR and VHR SAR images［J］.IEEE Journal of Selected Topics in Applied Earth Observations & Remote Sensing,2011,4(1)：27-34. ［37］BAN Y F,JACOB A,GAMBA P.Spaceborne SAR data for global urban mapping at 30 m resolution using a robust urban extractor［J］.ISPRS Journal of Photogrammetry and Remote Sensing,2015,103(5)：28-37. ［38］LONG J,SHELHAMER E,DARRELL T.Fully convolutional networks for semantic segmentation［J］.IEEE Transactions on Pattern Analysis & Machine Intelligence,2014,39(4)：640-651. ［39］LECUN Y,BENGIO Y,HINTON G.Deep learning［J］.Nature,2015,521(7553)：436. ［40］LECUN Y L,BOTTOU L,BENGIO Y,et al.Gradient-based learning applied to document recognition［J］.Proceedings of the IEEE,1998,86(11)：2278-2324. ［41］刘文涛,李世华,覃驭楚.基于全卷积神经网络的建筑物屋顶自动提取［J］.地球信息科学学报,2018,20(11)：1562-1570.［LIU Wentao,LI Shihua,QIN Yuchu.Automatic building roof extraction with fully convolutional neural network［J］.Journal of Geo-information Science,2018,20(11)：1562-1570.］ ［42］潘旭冉,杨帆,潘国峰.采用改进全卷积网络的“高分一号”影像居民地提取［J］.电讯技术,2018,58(2)：119-125.［PAN Xuran,YANG Fan,PAN Guofeng.Extraction of residential areas in GF-1 remote sensing images based on improved fully convolutional network［J］.Telecommunication Engineering,2018,58(2)：119-125.］ ［43］HE C Y,LIU Z F,GOU S Y,et al.Detecting global urban expansion over the last three decades using a fully convolutional network［J］.Environmental Research Letters,2019,14(3)：doi：10.1088/1748-9326/aaf936. ［44］FU G,LIU C J,ZHOU R,et al.Classification for high resolution remote sensing imagery using a fully convolutional network［J］.Remote Sensing,2017,9(5)：498. ［45］ZHANG P B,KE Y H,ZHANG Z X,et al.Urban land use and land cover classification using novel deep learning models based on high spatial resolution satellite imagery［J］.Sensors (Basel),2018,18(11)：3717. ［46］王一航,夏沛,刘志锋,等.中国绿洲城市土地利用/覆盖变化研究进展［J］.干旱区地理,2019,42(2)：341-353.［WANG Yihang,XIA Pei,LIU Zhifeng,et al.Research progress of urban land use/cover change in the oasis cities of China［J］.Arid Land Geography,2019,42(2)：341-353.］ ［47］李骞国,石培基,魏伟.干旱区绿洲城市扩展及驱动机制——以张掖市为例［J］.干旱区研究,2015,32(3)：598-605.［LI Qianguo,SHI Peiji,WEI Wei.Research on urban expansion and drive mechanism in an oasis town of arid region: A case of Zhangye City［J］.Arid Zone Research,2015,32(3)：598-605.］ ［48］梁峰,李骞国,石培基,等.张掖城市主要功能用地演变特征及其动因分析［J］.干旱区地理,2019,42(2)：414-422.［LIANG Feng,LI Qianguo,SHI Peiji,et al.Evolution characteristics and driving factors of urban functional land in Zhangye City,the Hexi Corridors,northwest China［J］.Arid Land Geography,2019,42(2)：414-422.］ ［49］冯斌,陈晓键.西北中小城市建设用地增长的特征、动因及其分类引导［J］.干旱区地理,2019,42(2)：376-384.［FENG Bin,CHEN Xiaojian.Characteristics,motivation and classification guidance of urban construction land growth of the small and mediumsized cities in northwest China［J］.Arid Land Geography,2019,42(2)：376-384.］ ［50］SZEGEDY C,LIU W,JIA Y,et al.Going deeper with convolutions［C］// IEEE Conference on Computer Vision and Pattern Recognition (CVPR).IEEE,2015. ［51］JIA Y,SHELHAMER E,DONAHUE J,et al.Caffe:Convolutional architecture for fast feature embedding［C］//Proceeding of the 22nd ACM International Conference on Multimedia.Orlando,USA:ACM,2014:675-678. ［52］JANSSEN L L F,WEL F J M V D.Accuracy assessment of satellite derived land cover data:A review［J］.Photogrammetric Engineering & Remote Sensing,1994,60(4)：419-426. ［53］斋藤康毅.深度学习入门：基于Python的理论与实现［M］.北京:人民邮电出版社,2018.［KOKI Saitoh.Deep learning from scratch［M］.Beijing:Posts and Telecommunications Press,2018.］
Options
文章导航

/

〈 〉