摘要
以北碚主城区为研究区域,利用面向对象方法对高分辨率遥感影像的水体进行提取,首先对QuickBird卫星影像进行分辨率融合,然后充分利用影像的光谱信息、拓扑关系、形状特征、大小信息等构建知识库进行分类。实验表明,该分类方法消除单纯利用光谱信息的缺陷,提高了分类精度。
The paper took Beibei urban as study area.Object-oriented information extraction technology was used to extract water bodies.Firstly,the resolution fusion method was used for QuickBird PAN and Multi-spectral images.Finally repositories were established by spectral feature,topological rules,shape feature,size information and so on and applied for to automatically extract water bodies types.The results indicates this method eliminates the deficiency of using spectral information only,and reaches high classification accuracy.
出处
《地理空间信息》
2010年第3期64-66,共3页
Geospatial Information
基金
国家十五科技攻关资助项目(2008BAB8B01)
重庆市自然科学基金资助项目(2008BB7010)
关键词
面向对象
遥感影像
数据处理
object-oriented
remote sensing image
data processing
