GEE平台和CART方法的北京市土地解译Land-cover and land-use classification in Beijing based on CART and GEE
胡云锋,商令杰,王召海,张千力
摘要(Abstract)：
针对传统遥感影像解译效率较低、人力物力需求量大等问题,该文以谷歌地球引擎为依托平台,利用Landsat5TM影像,采用分类回归树算法对2010年北京市土地覆被/土地利用类型开展了解译研究,并从类型构成、类型混淆和空间一致性3个方面将解译所得LUC-2010产品与Globeland30-2010产品进行空间一致性分析。研究表明,谷歌地球引擎(GEE)平台通过编程运算,数据处理速度极快,大幅提高工作效率。解译产品与训练样本交叉验证的学习精度为94.2%。两套产品总体对比发现,林地、水体和耕地的空间一致性比率分别为84.28%、74.75%和73.56%;林地、水体和人工地表的地类纯净度分别为87.23%、77.04%和72.97%;总体分布空间一致性为74.0%。两套产品局部对比发现,LUC-2010产品分类结果更准确和精细,精度更高。
关键词(KeyWords)： 谷歌地球引擎;分类回归树;遥感土地解译;土地利用与土地覆被
基金项目(Foundation): 国家重点研发项目(2016YFB0501502,2016YFC0503701);; 高分专项(00-Y30B14-9001-14/16)
作者(Author): 胡云锋,商令杰,王召海,张千力
DOI: 10.16251/j.cnki.1009-2307.2018.04.014
参考文献(References)：
- [1]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.)
- [2]陈军,陈晋,廖安平,等.全球30m地表覆盖遥感制图的总体技术[J].测绘学报,2014,43(6):551-557.(CHEN Jun,CHEN Jin,LIAO Anping,et al.Concepts and key techniques for 30mglobal land cover mapping[J].Acta Geodaetica et Cartographica Sinica,2014,43(6):551-557.)
- [3]MICHALSKI R S,BRATKO I,KUBAT M,et al.机器学习与数据挖掘:方法和应用[M].朱明,译.北京:电子工业出版社,2004.(MICHALSKI R S,BRATKO I,KUBAT M,et al.Machine learning and data mining:methods and applications[M].ZHU Ming,translation.Beijing:Publishing House of Electronics Industry,2004.)
- [4]BREIMAN L.Classification and regression trees[M].Belmont:Wadsworth International Group,1984.
- [5]TORBICK N,LEDOUX L,SALAS W,et al.Regional mapping of plantation extent using multisensor imagery[J].Remote Sensing,2016,8(3):236.
- [6]赵萍,傅云飞,郑刘根,等.基于分类回归树分析的遥感影像土地利用/覆被分类研究[J].遥感学报,2005,9(6):708-716.(ZHAO Ping,FU Yunfei,ZHENG Liugen,et al.Cart-based land use/cover classification of remote sensing images[J].Journal of Remote Sensing,2005,9(6):708-716.)
- [7]陈云,戴锦芳,李俊杰.基于影像多种特征的CART决策树分类方法及其应用[J].地理与地理信息科学,2008,24(2):33-36.(CHEN Yun,DAI Jinfang,LI Junjie.CART-based decision tree classifier using multifeature of image and its application[J].Geography and Geo-Information Science 2008,24(2):33-36.)
- [8]JOHANSENK,PHINN S,TAYLOR M.Mapping woody vegetation clearing in Queensland,Australia from landsat imagery using the google earth engine[J].Remote Sensing Applications:Society and Environment,2015(1):36-49.
- [9]YOHANNES Y,HODDINOTT J,YOHANNES Y.Classification and regression trees:an introduction technicalguide#3[M/OL].[2017-07-04].https://www.researchgate.net/publication/242370834_Classification_and_Regression_Trees_An_Introduction.
- [10]李金莲,刘晓玫,李恒鹏.SPOT5影像纹理特征提取与土地利用信息识别方法[J].遥感学报,2006,10(6):926-931.(LI Jinlian,LIU Xiaomei,LI Hengpeng.Extraction of texture feature and identification method of landuse information from SPOT5image[J].Journal of Remote Sensing,2006,10(6):926-931.)
- [11]PARMENTERA W,HANSEN A,KENNEDY R E,et al.Land use and land cover change in the Greater Yellowstone ecosystem:1975-1995[J].Ecological Applications,2003,13(3):687-703.
- [12]ROMO-LEON J R,LEEUWEN W J D,CASTELLANOS-VILLEGAS A.Using remote sensing tools to assess land use transitions in unsustainable arid agroecosystems[J].Journal of Arid Environments,2014,106(7):27-35.
- [13]YANG S,LUNETTA R S.Comparison of support vector machine,neural network,and CART algorithms for the land-cover classification using limited training data points[J].ISPRS Journal of Photogrammetry and Remote Sensing,2012(70):78-87.
- [14]WAHEED T,BONNELL R B,PRASHER S O,et al.Measuring performance in precision agriculture:CART-a decision tree approach[J].Agricultural Water Management,2006,84(1/2):173-185.
- [15]胡云锋,张千力,戴昭鑫,等.多源遥感土地覆被产品在欧洲地区的一致性分析[J].地理研究,2015,34(10):1839-1852.(HU Yunfeng,ZHANG Qianli,DAI Zhaoxin,et al.Agreement analysis of multi-sensor satellite remote sensing derived land cover products in the Europe continent[J].Geographical Research,2015,34(10):1839-1852.)