图像数据融合的地貌类型识别分类与制图  

Automatic Classification and Mapping of Regional Landforms Based on Fusion of DEM and TM Image

作　　者：张永民[1] 周成虎[2] 张旸[2] 

机构地区：[1]河南财经学院资源与环境科学系,郑州450002 [2]中国科学院地理科学与资源研究所,北京100101

出　　处：《地球信息科学》2006年第2期131-136,共6页Geo-information Science

基　　金：河南财经学院博士科研启动基金;国家自然科学基金项目(编号:40225004)

摘　　要：计算机遥感地貌制图是利用航空像片或者卫星影像进行识别制图;另是利用DEM数据融合计算提取。对此, 本文介绍了一种对区域基本地貌形态类型进行计算机自动分类的方法。它通过识别标志在影像上对地貌分布区进行数字化,把TM影像中的地貌信息和从DEM中提取出来的地貌信息结合,以划分出详细的地貌类型:如河北省沽源县的台地、河谷平原、开阔平原、丘陵、低山和中山6大类。最后,通过一定的算法进行分类成图。With the rapid development of technologies in tem and relevant commercial softwares, great progress has computer, digital image, geographical information sysbeen made in the field of automatic classification and mapping of regional landforms since the 1980s. However, two methods have been employed till now. One is visual interpretation using digital images, and the other is the automated extraction of landform characteristics from DEM. Because of the difference in personal visual perceptions, different boundary lines of the same landform unit often appear in two workers＇ interpretation results. Thus, the results obtained from the first method are difficult for future use. As to the second method, it is difficult to get detailed classifications （for example, to distinguish a valley plain from an open plain） by using DEM alone due to the complex nature in landform characteristics. In fact, DEM and digital image contain different, yet complementary, informations related to landform features. Therefore, a new method to integrate landform information of both DEM and TM image by digitizing signing lines in TM image is presented in this paper. With this approach, six types of basic landforms were successfully classified and mapped automatically in Guyuan county of Hebei province. In addition, the spatial variability of accuracy in classification was also evaluated by an application of fuzzy set theory using the notion of entropy in Guyuan case studies.

关 键 词：地貌形态类型 DEM TM影像 沽源县 

分 类 号：P283.8[天文地球—地图制图学与地理信息工程]