秦淮河丘陵地区土地利用遥感信息提取及制图

The Techniques of Deriving Landuse Information from SPOT Images and Study on Thematic Mapping in the field of Qinhuai River

着重论述了使用SPOT卫星遥感图像提取土地利用专题信息的技术与方法。利用数字地形模型(DEM)派生的坡度、坡向等辅助信息 ,对遥感影像的光谱特征空间进行扩展 ,建立基于知识的统计分析扩展模型 ,对不同参数的分类结果进行评价。研究结果表明 ,该方法能有效地提取SPOT卫星遥感土地利用专题信息 ,特别适合我国南方丘陵地区土地利用遥感信息提取。同时 ,通过地物分层分类信息提取方法和遥感影像数据融合技术 ,编制了研究区土地利用现状图。

As the composition of China-Germany cooperation project“Land use and Sustainable Development (SILUP)”,Jiangning county of Nanjing,in the South of China,was selected as a test area. The SPOT-2 and SPOT-4 remotely sensed images were used for deriving such thematic information as landuse,water bodies,roads and settlements and their thematic mapping. In this paper,some comprehensive methods and techniques of extracting the information from SPOT data were described,specially a improved method of land use/land cover classification by using SPOT images and ancillary data. In the land use/land cover classification,the feature sets containing the SPOT original band,and a digital elevation model were tested using MLC clustering. Incorporation of slop data generated from DEM was found to be able to improve the classification accuracy of land use. Further land cover discrimination was improved when using fusion images of PAN and XI. The study also put forward techniques of screen display editing to get the thematic maps. Some thematic maps were generated and their accuracy was assessed and discussed. The results showed that much good thematic information can be derived from SPOT images by using the methods and techniques introduced in this paper,and some extracted information can be used for drawing thematic maps in a scale of 1∶50000.