Local configuration measures for categorical spatial data: binary regular lattices |
| |
Authors: | Barry Boots |
| |
Institution: | (1) Department of Geography and Environmental Studies, Wilfrid Laurier University, Waterloo, ON, N2L 3C5, Canada |
| |
Abstract: | This paper examines the utility of a number of pattern measures for local exploratory analysis of binary spatial data. Based
on a review of existing pattern measures in cartography, geography, image analysis, and landscape ecology, two fundamental
classes of such measures, termed compositional and configurational, are identified. The paper focuses on configurational measures
and it is suggested that as many as five such measures (join counts, patch numbers, patch sizes, patch proximity, and distribution
of the classes relative to the focal cell of the window) are required to differentiate between all possible local categorical
maps. This suggestion is explored by examining aspects of the statistical behaviour (probability distributions and correlations
between extreme values of pairs of measures) of a set of 12 configurational measures. Their use is also demonstrated by means
of an empirical example. |
| |
Keywords: | Categorical spatial data Local spatial statistics Configuration measures |
本文献已被 SpringerLink 等数据库收录! |
|