水土保持研究2013，Vol.20Issue(3)：110-114,5.
DEM支持的黄土丘陵沟壑区地被遥感分类方法
Research on DEM Based on Remote Sensing Classification Method in Loess Hilly Ravine Area
摘要
Abstract
A land-cover classification method was put forward based on remote sensing image to solve the problem of low classification accuracy and lack of terrain characteristics due to the influence of complex terrain in Loess hilly Ravine area.This method uses PCA to recover the remote sensing image shadow information before land-cover classification,and to reduce the influence caused by large numbers of shadow.This method selects different samples on same cover of different terrain,and uses Bayes classification to complete land cover classification.DEM is used to classify the landform into ridges and hills,sloping fields,gullies,combines the spectrum feature analysis and terrain information processing,overlays the result with terrain classification,and restructures the figure spot based on land type Boundary.This method combines spectral features and spatial topographic information in the process of classification,therefore,the classification result precision is improved.This method makes the classification results with the terrain information,fully embodies the actual situation of the cover distribution in Loess hilly ravine area,and enhances the practical significance of classification results.关键词
黄土丘陵沟壑区/遥感影像/DEM/地被分类/地形分类Key words
Loess hilly-gully region/ remote sensing image/ DEM/ land-cover classification/ terrain classification分类
信息技术与安全科学引用本文复制引用
李小曼,王刚..DEM支持的黄土丘陵沟壑区地被遥感分类方法[J].水土保持研究,2013,20(3):110-114,5.基金项目
武警工程大学自然基金项目(WJY201113) （WJY201113）
