三江源自然保护区土地利用遥感分类方法研究  

Study on Remote Sensing Classification of Land Use in the Nature Reserve of the Three Rivers Source Region

作　　者：邹文涛[1] 张怀清[1] 鞠洪波[1] 刘华[1] 

机构地区：[1]中国林业科学研究院资源信息研究所,北京100091

出　　处：《林业资源管理》2010年第6期90-96,共7页Forest Resources Management

基　　金：中央级公益性科研院所基本科研业务费专项资金"气候变化背景下的三江源典型区湿地变化研究"(IFRIT200906);国家"十一五"科技支撑重点项目课题"综合监测技术体系集成与应用示范"(2006BAD23B06)

摘　　要：以三江源区域索加曲麻河自然保护区为例,基于TM影像进行数据变换和波段运算后获取的特征指数,采用决策树方法,探讨了高寒区域土地利用遥感分类方法。然后通过与传统的最大似然法监督分类所得到的结果进行对比,结果表明:利用基于指数的决策树分类方法对高寒区域土地利用/土地覆盖类型进行遥感分类,较传统的最大似然法监督分类总体精度提高15.48%,总体kappa系数提高0.174 1;滩地、沼泽、高覆盖度草地、低覆盖度草地、裸岩石砾地等地类的用户精度提高较大,分别提高28.13%,25.00%,17.86%,17.86%和12.50%。低、中、高3种覆盖度草地,裸岩石砾地的生产者精度也有较大幅度的提升。表明基于指数的决策树分类方法是高寒区域土地利用遥感分类的一种有效手段。This paper took the Suojiaqumahe Nature Reserve,which is located in the source regions of the three rivers（Yangtze River,Yellow River and Lancang River）,as the study site.Verification was made on the efficiency of decision tree based on indices from TM image transformation and band operation in alpine land use classifying.And then,the results were compared with the traditional maximum likelihood supervised classification.It showed that the decision tree method based on the indices can improve the overall accuracy by 15.48%,and the overall kappa coefficient by 0.1741.For bottomlands,swamp,high coverage grassland,low coverage grassland and barren land,the users accuracies were increased by 28.13%,25.00%,17.86%,17.86% and 12.50% respectively.For different coverage grassland,barren land,the producers accuracy also increased dramatically.The result indicates that the method based on indices got from image band transformation and band operation is an effective way of alpine land use/land cover remote sensing classification.

关 键 词：三江源区域 土地利用 决策树算法 特征指数 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]