摘要
以位于印度尼西亚Kalimantan热带雨林地区的1幅具有3个通道的雷达图像为研究对象,利用面向对象方法对高分辨率遥感影像的水体进行提取,在分割的基础上,经过分层分类,充分利用影像的光谱信息、拓扑信息、形状特征等构建知识库进行分类,实验结果理想,相对传统基于像素的图像分类分析方法结果图斑完整,提高了分类精度。
This paper took Kalimantan as study area.Object-oriented information extraction technology was use to extract water bodies.Based on multiresolution segmentation,we carried out a classification of twolayered.Through the whole process,the image information of spectrum,topology,shape were adequately used.The results of experiment were famous,the classification objects of water were integrate contrasting to the traditional classification based on pixel analysis.
出处
《地理空间信息》
2012年第2期64-66,181,共3页
Geospatial Information
基金
广州市城市规划勘测设计研究院2011年度院科研资助课题
