作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

›› 2005, Vol. 31 ›› Issue (11): 163-.

• 论文 • 上一篇    下一篇

基于爬峰法聚类的TM图像专题信息人机交互提取

葛宏立,方陆明,孟宪宇,唐小名   

  1. 北京林业大学资源与环境学院,浙江林学院信息工程学院,北京林业大学资源与环境学院,中国林业科学研究院资源信息研究所 北京 100083
  • 发布日期:2005-06-15

Interactive Thematic Information Extracting from TM Image Based on Peak-climbing Clustering

GE Hongli1, FANG Luming2, MENG Xianyu1, TANG Xiaoming3   

  1. (1. College of Forest Resources and Environment, Beijing Forestry University, Beijing 100083,;2. School of Information Engineering, Zhejiang Forestry College, Lin’an 311300; 3.Research Institute of Forest Resource Information Technique, CAF, Beijing 100091)
  • Published:2005-06-15

摘要: 介绍一种基于爬峰法(或称寻谷法)聚类的专题信息交互提取方法。爬峰法是一种基于高维直方图的图论方法,是一种完全的非参数聚类方法,具有适应性好、计算速度快的优点。

关键词: 图像分割, 聚类分析, 爬峰法, 非参数方法, 专题信息提取, 水体提取, 城镇用地提取

Abstract: This paper proposes an interactive thematic information extracting approach based on the peak-climbing (valley-seeking) clustering. Peak-climbing algorithm is a graph-theoretic clustering method based on a multidimensional histogram. It is a totally nonparametric method that mainly possesses advantages of having good flexibility and needs less computer time.

中图分类号: