中文摘要：

城市裸地通常面积较小且分散分布,其光谱特征与建筑用地相似,传统监督分类方法很难准确提取。采用分层分类思想,利用改进后的归一化水体指数和归一化植被指数分别提取水体、植被后并去除,再利用光谱特征分析实现建筑用地和裸地分离,最后提取出城市裸地。以武汉市主城区为研究区,选取Landsat7/ETM＋和Landsat8/OLI影像,分层分类法裸地提取精度分别达到94.5%和96%,而作为对比实验的监督分类法裸地提取精度分别为89%和90%,可见,分层分类法可以较高的精度提取城市裸地信息。

英文摘要：

It is difficult to extract city bare land using supervised classification because of its small area and dispersive distribution and also because it shares the similar spectral signature with city construction land.In this study,guided by the idea of hierarchical classification,Modified Normalized Difference Water Index（MNDWI）and Normalized Difference Vegetation Index（NDVI）are adopted for water and vegetation extraction and then they are removed;separation of construction land and bare land is based on spectral signature analysis in the following in order to extract city bare land.Taking main urban area of Wuhan city as a case study,accuracy of bare land extraction for Landsat7/ETM＋and Landsat8/OLI is 94.5%and 96% with hierarchical classification;correspondingly,accuracy is 89% and 90% with supervised classification.The results indicate that hierarchical classification can extract city bare land with high accuracy.