摘要
湿地是自然界最富生物多样性的生态系统和人类最重要的生存环境之一,运用遥感影像提取湿地信息,进而对湿地的土地利用/覆盖进行分析,为后续湿地变化检测提供一个适宜的分析方法,对于湿地保护和可持续利用有重要意义。该研究以崇明东滩河口型湿地为例,通过对研究区ETM+影像进行K-T变换、HLS变换,主成分分析,并结合NDVI指数、均值纹理等特征,构建基于知识规则的决策树模型来提取湿地信息,结果表明,决策树模型以其自身优势,能够有针对性地实现湿地信息的快速提取,并取得较高的分类精度。
As wetland is one of the richest ecosystems in bio-diversity and the most important survival environment of human, it has important meaning for wetland protection and sustainable application to carry out information extraction and classification by using remote sensing technology. Using NDVI index, spectral characteristics, K-T transformation, HLS transformation and the principal component analysis to ETM-}-image of the study area, it presents a decision tree model based on knowledge for extracting wetland information taking Chongming Dongtan estuarine wetland as an example. The result indicates that the decision tree model can realize quickly the extraction of the targeted wetland information, which has a higher classification accuracy.
出处
《测绘工程》
CSCD
2013年第1期63-66,69,共5页
Engineering of Surveying and Mapping
基金
国家科技支撑计划资助项目(2010BAK69B13)
上海市国际科技合作基金项目(11230705700)
关键词
ETM+遥感影像
决策树
河口型湿地
K—T变换
HLS变换
主成分分析
ETM＋remote sensing image
decision tree
estuarine wetland
K-T transformation
HLStransformation
principal component analysis
