以陕西省榆林市神木县北部神府东胜矿区为研究区,利用LandsatETM影像,在ENVI软件的支持下,分析了影像的光谱特征及NDVI,NDBI,NDWI特征值,并对影像进行缨帽变换,确定各地类的综合阈值,建立决策树模型,得到分类结果,并且评价了决策树分类在遥感数据分类中的作用。
Taking Shenfu Dongsheng mine in north of Yulin Shenmu County,Shanxi Province as a study case,and with the support of ENVI software, Landsat ETM image was used to analyze the spectral characteristics of the image and its value of NDVI,NDBI,NDWI. And this image was transformed with tasseled cap to determine the threshold value for different lands and build a decision tree. Then, the classification results were obtained and the role of decision tree in remote sensing data classification was evaluated.