黄土高原地形复杂区土地利用信息提取方法研究

引入影响土地利用类型分布的坡度数据和对植被覆盖反应敏感的归一化植被指数，以延河流域为研究区域，参考已有的土地利用数据，对地形复杂地区的土地利用信息提取方法进行了研究。结果表明，以监督分类和非监督分类为基础，辅以坡度信息并利用分区和多步骤信息提取方法进行土地利用信息提取，能在一定程度上提高分类精度，研究结果对建立地形复杂地区实用性强的遥感影像分类技术体系具有一定的参考价值。

RESEARCHES ON METHODS FOR LAND USE INFORMATION EXTRACTION IN COMPLICATED TERRAIN AREAS ON LOESS PLATRAU

Soil erosion is one of the important environment problems,and land use classification is an important process in this aspect.The normal automatic classification(supervised classification and unsupervised classification) based on spectral characteristics cannot meet the accuracy needed.Therefore,the slope(produced by DEM) which affects the land use type location and the NDVI(produced by TM images) which reflects the vegetation coverage sensitivity should be taken into account.With Yanhe basin as the study area,the authors carried out the research on methods for extracting the land use information in complicated areas on the loess plateau.The result indicates that the division of the image into several parts according to the TM image characteristics and the extraction of the land use type one by one assisted by the slope are suitable for the complicated terrain area.The result obtained serves as an important reference to the remote sensing classification technical system.