摘要
在内蒙古多伦县土地利用现状调查中,根据该区地表异质性和破碎化强烈、地物光谱特征混杂严重、土地利用/覆被遥感信息提取对目视解译依赖程度高的特点,将基于知识的遥感信息提取技术应用于该地区。通过地物光谱特征的深入分析,用线性光谱混合分解模型将主要地物覆被类型分离,并建立多个专题信息模型,依据经验知识建立了各用地类型提取规则,提取了多伦县2007年的土地利用信息。结果表明:2007年的土地利用结构仍以耕地、牧草地、未利用地为主,3种地类分别占总面积的16.4%、55.2%和20.3%,林地面积为21919hm2,占总面积的5.6%。旱地、退化草地和沙地比例偏高,多伦县的生态环境虽有所改善,但仍处于极不稳定的状态。
Knowledge-based remote sensing information extraction technology was applied to study the information extraction for land use in Duolun County,Inner Mongolia,aiming at the strong surface heterogeneity and fragmentation,the severely mixed spectral characteristics of surface features,and the high dependency of information extraction for land use/cover remote sensing on visual interpretation.By in-depth analysis of spectral characteristics,the main cover types were decomposed by linear spectral mixture model and a number of thematic information models were set up.Extraction rules of all types were set up based on empirical knowledge,and the land use information of Duolun in 2007 was extracted by this model.Results showed that farmland,grassland and unused land were still the main land types of Duolun in 2007,accounting for 16.4 percent,55.2 percent and 20.3 percent of the total area respectively.The woodland area was 21919hm^2 and occupied 5.6 percent of the total.The glebe,degenerative grassland and desert land accounted for a high proportion of area.The ecological environment has been improved in Duolun,but it is still in a very unstable state.
出处
《东北林业大学学报》
CAS
CSCD
北大核心
2010年第10期88-90,94,共4页
Journal of Northeast Forestry University
基金
"十一五"国家科技支撑计划项目(2006BAD23B05)
关键词
土地利用
遥感
决策树
多伦县
Land use
Remote sensing
Decision trees
Duolun
