摘要
在回顾前人运用数字图像处理技术从遥感图像中提取空间变化特征经典算法的基础上,总结出基于特征的、基于图层的和基于像素的三种方式提取影像特征的方法。实验结果表明该方法适用于轮廓较明显的目标,同时具有速度快、精度高、稳定性强的优点。
With many high quality images and new models coming forth, remote sensing technology has made quite great progress in the past 40 years, having abroad applications in the field of geography. There have been many case studies in extracting cartographical features from images with digital image technology. This paper summarizes three approaches, feature-based, region-based and pixel-based, after reviewing classical algorithms. The experimental results indicate that the proposed algorithm is efficient, precise, and robust.
出处
《地理空间信息》
2008年第2期43-45,共3页
Geospatial Information
基金
国家自然科学基金资助项目(40501058)
关键词
遥感影像
图像处理
数据库
数据更新
remote sensing image
image processing
database
data revision