论文部分内容阅读
选择策勒绿洲作为典型的绿洲-荒漠交错带,采用Landsat ETM+影像,分析了沙漠化土地的光谱特征及其波段间的相互运算,用分层分离的方法,提取了沙漠化土地信息.结果表明:利用修改型土壤调整植被指数(MSAVI),归一化差异水体指数(NDWI)和遥感图像缨帽变换后的亮度(Brightness)、绿度(Greenness)、湿度(Wetness)等复合识别指标,在决策树的各节点设计不同的分类器,可以划分沙漠化等级;决策树分类法可以有效地排除和避免提取地物时受多余信息的干扰及影响,其总体提取效果较好,是快速自动提取沙漠化土地信息的有效手段.
Using Cele Oasis as a typical oasis-desert ecotone, Landsat ETM + images were used to analyze the spectral characteristics of desertified land and their inter-band operation, and the desertification land information was extracted by stratified separation method. The results showed that: (MS), Modified Difference Water Body Index (NDVI) and Brightness, Greenness and Wetness of remote sensing images, , Each classifier can design different classifiers to classify the desertification level. The decision tree classification method can effectively eliminate and avoid the interference and influence of the extra information when extracting the features. The overall extraction effect is good, and it is a fast and automatic method for extracting desertification land information Effective means.