中文摘要：

快速准确地获取城市绿地信息是城市绿地可持续经营和科学规划的基础.以南京市玄武区绿地为研究对象,以2002年SPOT5遥感数据为主要信息源,以同期南京市主城区交通旅游图、绿地统计资料为辅助信息源,通过目视解译获取各类绿地的准确资料.分别采用简单随机、系统、分层、空间平衡抽样4种计算机屏幕抽样方法,对研究对象的绿地信息进行参数估计,并比较分析各种抽样方法的抽样精度.研究结果表明：目视解译结果与统计数据相比,误差仅为5.1%;4种抽样方法的抽样精度大小排序为：空间平衡抽样（98%）（分层抽样（94%）（系统抽样（82%）（简单随机抽样（75%）;空间平衡抽样与遥感图像目视解译的误差仅有2%,可以代替目视解译,较大幅度地降低了目视解译的时间、人工成本.

英文摘要：

Timely and accurately obtaining the information of urban green space provides the basis for green space sustainable management and scientific planning.In this article,green space in Xuanwu District of Nanjing City was chosen as the research object,and remote sensing data of SPOT 5 in 2002 was collected as the main information source,while the traffic tourist map of urban Nanjing and green space statistics date in 2002 were employed as the auxiliary information source.Through visual interpretation,accurate data of various kinds of green space in case study area in 2002 was obtained.Then,four sampling methods of simple random sampling,system sampling,stratified sampling and spatial balanced sampling were applied to sampling on green space on computer screen,followed by parameter estimate of green space and accuracy comparison of different sampling methods.Research results showed that：（1） the comparative error of visual interpretation was only 5.1%;（2） as for sampling accuracy,spatial balanced sampling（98%）stratified sampling（94%） system sampling（82%） simple random sampling（75%）;（3） compared with visual interoperation,the error of spatial balanced sampling is as small as 2%,showing the possibility of replacing visual interpreting,thus greatly reducing the time and labor cost.