摘要
湿地研究的关键在于定量化获取和分析湿地信息,特别是湿地分布信息。以三江平原为例,结合湿地的光谱特征和时相特征,利用多时相MODIS数据,采用MNF(最小噪音分离变化)技术和非监督分类获取了研究区湿地分布,其精度达79%。研究表明,利用MODIS数据可以有效提取湿地分布。
The keys of wetland research are to obtain and analyze wetland information quantitatively,especially,wetlands distribution. RS and GIS technology provide a new way to wetland research. On the basis of absorbing the results from corrective studies, this paper uses MODIS data to extract wetlands distribution,and makes validation from several aspects, which provides a reference for managing and monitoring wetland. The paper takes Sanjiang Plain as a case study,and utilizes MNF（Minimum Noise Fraction）transformation and unsupervised classification method to acquire wetland distribution in this area using multi - temporary MODIS data. Validation shows that accuracy is 79% ,and proves that MODIS can acquire wetland distribution effectively.Results indicate that it is necessary to combine spectral and temporal characteristics of different wetlands in monitoring wetland by means of RS.
出处
《地理与地理信息科学》
CSCD
北大核心
2006年第3期57-60,共4页
Geography and Geo-Information Science
基金
国家科技部项目"MODIS共享平台标准规范研究"(2003DKAlT007)
