摘要
GIS栅格数据能表述连续变化的空间现象,它每个像元的属性值有连续数据和离散数据2种。针对栅格连续数值空间进行分析,突破传统的基于栅格本身邻域位置的研究方法,从属性值本身来定义邻域,基于属性距离度量公式来定义数据的领域。不同的邻域大小,一个栅格像元可以得到不同的邻域;并且从粗糙逼近的角度来探讨邻域信息粒子样本的上、下近似,此研究将为栅格数据分类等数据分析方法提供一种新的视角。
GIS grid data can express continuous spatial phenomenon,each pixel's value of grid data has two types,that is continuous data and discrete data.This research aimed at continuous data space to study.It broke the former study that was based on the neighborhood of its own position,and it defined neighborhood from attribute value and distance formula.Different neighbor size could obtain different neighborhood for every grid pixel.We discussed upper and lower approximation of neighborhood particle from rough set.This research could provide a new method for grid data reclass and data analysis.
出处
《地理空间信息》
2011年第4期31-33,189-190,共3页
Geospatial Information
关键词
栅格数据
邻域粒子
粗糙集
距离度量
grid data,neighborhood particle,rough set,distance measure
