张文博,孔金玲,杨园园,李彤

• 1:长安大学地球科学与资源学院
• 2:长安大学地质工程与测绘学院

摘要(Abstract)：

针对旱区植被分类尺度过大、种群无法准确提取的问题,该文提出了面向对象的CFS-RF分类模型,即利用CFS算法对先验样本数据集进行特征优选,结合随机森林构建分类规则,完成分类过程。以新疆阿勒泰为研究区,利用GF-2数据,通过CFS、ReliefF两种不同特征选择方法和J48、SVM、RF 3种分类算法构造出6种面向对象分类方案来实现小尺度植被种群提取。结果表明,经过特征选择,上述分类方案的精度和效率均得到了提升。其中,CFS-RF算法最优,总体精度达到92.41%,Kappa系数为0.90,更适用于旱区植被遥感精细分类。

关键词(KeyWords)： 高分影像;旱区植被;特征优选;面向对象分类

Abstract:

Keywords:

基金项目(Foundation): 新疆乌伦古河流域水文地质环境地质调查项目（S17-2-XJ07）

作者(Author): 张文博,孔金玲,杨园园,李彤

DOI: 10.16251/j.cnki.1009-2307.2021.01.018

参考文献(References)：

• [1]SCHEIEWE J,TUFTE L,EHLERS M.Potential and problems of multi-scale segmentation methods in remote sensing[J].Geo-BIT/GIS,2001,6:34-39．
• [2]王永吉，孟庆岩，杨健，等．一种基于特征选择的面向对象遥感影像分类方法[J]．科学技术与工程，2016,16(32):107-113.(WANG Yongji,MENG Qingyan,YANG Jian,et al.Object based remote sensing image classification based on feature selection method[J].Science Technology and Engineering,2016,16(32):107-113．）
• [3]张秀英，冯学智，江洪．面向对象分类的特征空间优化[J]．遥感学报，2009,13(4):664-669.(ZHANG Xiuying,FENG Xuezhi,JIANG Hong.Feature set optimization in object-oriented methodology[J].Journal of Remote Sensing,2009,13(4):664-669．）
• [4]李冰，卢小平，李新社，等．特征优选的GF-2影像湿地地表覆盖要素提取[J]．测绘与空间地理信息，2018,41(9):49-52.(LI Bing,LU Xiaoping,LI Xinshe,et al.Extraction of gf-2image surface cover elements with feature optimization[J].Mapping and Spatial Geographic Information,2008,41(9):49-52．）
• [5]AVRIM L.BLUM,LANGLEY P.Selection of relevant features and examples in machine learning[J].Artificial Intelligence,1997,97(1):245-271．
• [6]肖艳，姜琦刚，王斌，等．基于ReliefF和PSO混合特征选择的面向对象土地利用分类[J]．农业工程学报，2016,32(4):211-216.(XIAO Yan,JIANG Qigang,WANGBin,et al.Object-oriented land use classification based on mixed feature selection of ReliefF and PSO[J].Transactions of the Chinese Society of Agricultural Engineering,2016,32(4):211-216．
• [7]HALL M A.Correlation-based feature subset selection for machine learning[EB/OL].[2019-10-09]https：∥www.docin.com/p-896976522.html．
• [8]温小乐，钟奥，胡秀娟．基于随机森林特征选择的城市绿化乔木树种分类[J]．地球信息科学学报，2018,20(12):1777-1786.(WEN Xiaole,ZHONG Ao,HUXiujuan.Classification of urban greening trees based on random forest feature selection[J].Journal of GeoInformation Science,2008,20(12):1777-1786．）
• [9]刘建平，赵英时．高光谱遥感数据解译的最佳波段选择方法研究[J]．中国科学院研究生院学报，1999,16(2):153-161.(LIU Jianping,ZHAO Yingshi.Research on best band selection method for hyperspectral remote sensing data interpretation[J].Journal of Graduate School,Academia Sinica,1999,16(2):153-161．）
• [10]黄辉．基于ReliefF的多标签特征选择算法研究[D]．广州：广东工业大学，2018.(HUANG Hui.Research on multi-label feature selection algorithm based on ReliefF[D].Guangzhou:Guangdong University of Technology,2018．）
• [11]KONONENKO I.Estimating attributes:analysis and extensions of RELIEF[J].Proceedings of the European Conference on Machine Learning on Machine Learning.[S.l.]:IEEE,1994:171-182．
• [12]刘晓宇．C4.5算法的一种改进及其应用[D]．青岛：中国海洋大学，2013.(LIU Xiaoyu.An improvement of C4.5algorithm and its application[D].Qingdao:Ocean University of China,2013．）
• [13]单治彬，孔金玲，张永庭，等．面向对象的特色农作物种植遥感调查方法研究[J]．地球信息科学学报，2018,20(10):1509-1519.(SHAN Zhibin,KONG Jinling,ZHANGYongting,et al.Remote sensing investigation method of object-oriented crops with special characteristics[J].Journal of Geo-Information Science,2008,20(10):1509-1519．）
• [14]BREIMAN L.Random forests[J].Machine Learning,2001,45(1):1．
• [15]戴莉莉，李海涛，顾海燕，等．特征优选下的遥感影像面向对象分类规则构建[J]．测绘科学，2019,44(2):26-32.(DAI Lili,LI Haitao,GU Haiyan,et al.Construction of object oriented classification rules for remote sensing images under feature optimization[J].Science of Surveying and Mapping,2019,44(2):26-32.)