地学粗空间的理论与应用
Theory and Application of Geo-rough Space
-
摘要: 提出了由粗实体、粗关系和粗算子组成的地学粗空间。在分析通过粗集研究地球空间信息学不足的基础上,提出了简化的实用粗符号系统;定义了地学粗空间的内涵和外延,把空间论域全域划分为下近似集、边界集和非集,将空间实体以上、下近似集表示为粗实体,并用粗度衡量其逼近程度;给出了兼顾位置和属性的不确定性的粗关系及其矩阵表示;研究了地学粗空间的粗算子,重点讨论了粗交集近似通用矩阵、基于影像灰度值的粗隶属函数和冗余属性剔除等粗算子;以基于地学粗空间的河流专题遥感影像分类为例,检验了地学粗空间的实用性。结果表明,地学粗空间可以较大可能地逼近地学实体的真实存在形式,决策信息较为丰富、可靠。Abstract: In this paper,a geo-rough space was proposed as a new approach to handle spatial entities in data analysis.And its basis was rough set.In the geo-rough space,the set of symbols from rough set are simplified and standardized,which includes both rough interpretation and specialized indication.The boundary set was used to take the place of the upper approximation in a practical computation and decision-making.The geo-rough space was characterized by rough entities,rough relationships and rough algorithms in spatial discourse universe,whose element was rough element with a rough degree to measure the roughness.This paper discussed it in Geomatics,paying more attention to the computerized GIS.The result of the case study on thematic image extraction indicated that the geo-rough space was confidential and practical.