中文摘要：

基于Landsat卫星遥感影像,将一种新的水体自动提取指数（AWEI_nsh/AWEI_sh）运用于湿地信息提取,并通过箱型图统计方法自动获取了12种相关变量的阈值范围.结果表明：利用该水体自动提取指数和归一化植被指数,结合数字高程模型建立决策树模型,可对研究区内的湿地信息进行有效提取,且2000,2014年的遥感影像湿地信息提取精度分别为83%、93%,提取方法可用于大范围湿地研究.

英文摘要：

New water indices （AWEI_nsh/AWEI_sh） were used to extract the wetland information based on the Landsat images and the threshold ranges of twelve relevant variables could be obtained automatically through an analysis of box figures. The result showed that the decision tree model combining the use of AWEI_nsh/ AWEI_sh, normalized difference vegetation index and DEM could effectively extract the wetland information in the study area. The accuracy of extraction results for 2000 and 2014 were 83% and 93% respectively. The method can be widely used in wetland research.