中文摘要：

本文利用CBERS-02B影像，通过分析各个地类的光谱特征，发现了长沙市城市建设用地和其他背景地物的区别，并在此基础上选取了土壤调节植被指数（SAVI）、归一化水体指数（NDWI）和比值居民地指数（RRI）怍为三个指数波段，重新进行波段组合，从而减少了波段数据的冗余，最后采用最大似然法分类，提取出城市建设用地信息，其正确率达到85．6％。

英文摘要：

Based on the CBERS-O2B image data, this paper analyzed the Spectral Characteristics of land use types, and found the difference between the urban built-up land and others. The study selected three indices, i.e. Soil Adjusted Vegetation Index （ SAV I）, normalized difference water index （NDWI） and Ratio Resident-area Index（RRI） , to represent three major index bands generated from the original muhispectral bands and dramatically decreased band correlation. Using Maximum Likelihood Classification （MLC） to retrieve urban built-up areas in the region of Changsha, the overall accuracy of them were 85.6%.