【摘要】 以南京市为研究对象,获取研究区域Landsat8 OLI遥感影像,利用QUEST决策树的分类方法对影像进行分类。将植被覆盖指数(NDVI)、迭代自组织数据分析技术(ISODATA)非监督分类作为地学辅助数据因子添加到分类波段中,构建多源数据集进行不同特征的融合,处理目标类别之间的非线性关系。该方法灵活性大,总精度达91.045%,Kappa系数为0.851,取得了比普通方法更好的精度。精度的提高有助于解决南京市复杂的规划、决策和管理等问题。更多还原

【Abstract】 Taking Nanjing city as the research object, Landsat 8 OLI remote sensing image of the study area was acquired and classified using the QUEST decision tree classification method. The NDVI(Normalized difference vegetation index), ISODATA(Iterative self-organizing date analysis technique) unsupervised classification of vegetation were added to the classification band as the geoscience auxiliary data factors. Multi-source data was constructed to fuse different feature, and nonlinear relationship among the target categories was analyzed. This method had great flexibility and better accuracy than conventional methods, with accuracy was 91.045% of the total and Kappa coefficient was 0.851. The improvement of the accuracy could help to solve complex planning, decision-making and management issues of Nanjing city.更多还原