中文摘要：

以南京南部高淳县为研究区，采用2010年ETM＋多光谱遥感影像作为遥感信息源，选择影像的地形因素、植被指数（NDVI）作为辅助分类特征，基于改进CRUISE算法构建决策树，实现了研究区的地物分类，并与其他分类方法的结果相比较。实验结果表明，与普通的决策树分类相比，基于改进CRUISE算法的分类可以有效地提高土地分类结果精度，具有良好的适用性。

英文摘要：

Traditional decision tree algorithm has been lagged behind because of its spectra characteristics .CRUISE algorithm becomes better in the way of space expression ,arithmetic speed and classification accuracy .The case study in Gaochun County ,south of Nanjing ,adapting ETM and multispectral remote sensing image in 2010 ,with topographic factor and normalized difference vegetation index （ND-VI） into consideration as aided classification feature ,aims to establish decision tree based on improved CRUISE algorithm ,to realize the differentiation of the surface features ,and to compare with the other classification methods .The result shows ,different from the traditional decision tree classification ,the classification based on improved CRUISE algorithm can effectively improve the precision of land classification and has a good applicability .