中文摘要：

以西安市长安区TM影像为例，研究关于城市建筑用地信息快速、准确提取的方法。通过对归一化差异型指数构成原理的分析，选取土壤调节植被指数（ SAVI ）、归一化水体指数（ NDWI ）和归一化差异型建筑指数（ NDBI ）来提取植被、水体和城市建筑用地专题影像，并将其构建为一幅新影像，分析新影像谱间特征，运用逻辑运算将城市建筑用地信息提取出来。本文方法总体提取效果十分有效，尤其是对于面积较大的城市建筑用地，总精度高达85．3％。综合指数法弥补了单靠某一指数提取城市建筑用地信息的不足，提取结果客观可信，是一种不经人为干预、快速有效的提取城市建筑用地的方法。

英文摘要：

The paper studies the method of fast and accurate information extraction for urban built -up land, taking TM image of Chang ’ an district of Xi ’ an as an example .With the analysis of normalized difference indices composition principle , the paper select-ed three indices , Soil Adjusted Vegetation Index （SAVI） , Normalized Difference Water Index （NDWI） and Normalized Difference Built-up Index（NDBI） to make urban vegetation thematic image , water body thematic image and urban build -up land thematic im-age, respectively.The three thematic images are used as three bands to compose a new image .After the analysis of spectral signature , the built-up land is finally extracted through a simple logic calculation .The results of this research show the method adopted in this paper is proved to be effective , especially large area urban built -up land, the result achieves a high accuracy of 85.3%.The com-prehensive index method makes up a defect of single method which only uses normalized difference index to extract urban build -up land information, this result is impersonal and credible , it is a fast and effective method to extract urban built -up land information without human interference .