基于面向对象的重大工程土地利用变化信息提取——以国家体育场(鸟巢)建设工程为例  

THE EXTRACTION OF LAND USE CHANGE CAUSED BY THE CONSTRUCTION OF KEY PROJECTS BASED ON OBJECT-ORIENTED METHOD:A CASE STUDY OF THE NATIONAL STADIUM

作　　者：刘浩[1,2,3] 胡卓玮[1,2,3] 赵文慧[1,2,3] 

机构地区：[1]首都师范大学资源环境与旅游学院,北京100048 [2]三维信息获取与地学应用教育部重点实验室,北京100048 [3]资源环境与地理信息系统北京市重点实验室,北京100048

出　　处：《国土资源遥感》2009年第4期86-89,95,共5页Remote Sensing for Land & Resources

基　　金：国家科技支撑计划项目(编号:2007BAH15B02;2006BAC08B02);北京市教育委员会科技计划重点项目(编号:09224010028)共同资助

摘　　要：以国家体育场(鸟巢)工程建设为例,研究基于面向对象的重大工程土地利用变化信息提取技术,以及建设工程对周边环境土地利用类型变化的影响。选取鸟巢2001年(动工之前)和2005年(建设中期)两期高分辨率遥感图像,运用多尺度影像分割和面向对象分类方法,进行土地利用信息提取及其变化对比分析。结果表明:利用面向对象方法取得了理想的土地利用分类结果,两期影像的分类精度分别达到93.23%和94.55%,Kappa系数分别为0.9158和0.9332;鸟巢的建设对周边土地利用类型影响较大,尤其是植被、道路和耕地的变化显著,草地和道路明显增加,耕地明显减少。Exemplified by the National Stadium, this paper probes into the influence of the construction of key projects on the periphery areas and the technology of land - cover change based on the object - oriented method. Selecting two images of the National Stadium in 2001 （before the construction） and in 2005 （the intermediate stage of the construction） and using muhi - scale segmentation and object - oriented classification method, the authors carried out the land use information extraction and a comparative analysis of the land use change information resulting form the National Stadium construction. The results show that： （ 1 ） the ideal land cover classification results were obtained by using the object -oriented method, with the overall accuracy being 93.23% and 94.55% and the coefficient of Kappa being 0. 915 8 and 0. 933 2, which make up the foundation for rapid and accurate acquisition of land cover change information; （2） the construetion of the National Stadium had a significant influenee on periphery areas, as shown by the obvious change of vegetation, roads and cultivated land. Grasses and roads increased obviously while the cultivated land was reduced remarkably.

关 键 词：多尺度分割 面向对象分类 变化检测 重大工程 国家体育场(鸟巢) 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]