Fuzzy Modeling of Geometric Textures for Identifying Archipelagos in Area-Patch Generalization

Abstract

This paper describes a method to approach area-patch problems in model generalization using fuzzy set theory. The area-patch problem identifies polygons with the same semantics (feature code) but varying geometry and spatial distribution. Area-patch generalization can be understood as a pattern recognition problem with a variety of viable solutions, constrained by the purposes of the generalized output. The focus lies with identifying geometric textures associated with groups of swamps collectively identified as archipelagos in preparation for generalization. In this research, polygonal swamp/marsh features as area-patches are considered, and constraints are defined in part by USGS data production specifications. The method developed will identify archipelagos for swamps and marshes in an NHD High-Resolution subbasin dataset; and the study area for this work spans the Florida-Georgia border. Fuzzy membership functions are assigned to swamp/marsh polygons based on area, inter-polygon distance and number of neighbors within a predefined distance. These attributes are combined in a fuzzy overlay to derive degrees of membership that rank each polygon for possible aggregation into archipelagos. The final aggregation is based on 'alpha cuts' thresh-holding. A sensitivity analysis evaluates the impact of different fuzzy overlay operators and varying alpha-cuts on the resulting pattern. To validate the change in geometric properties (area, area/perimeter ratios) and overall texture from the original scale to the target scale, a comparison is made between our approach and a solution that does not account for archipelagos. Preliminary findings indicate that a fuzzy set approach allows for the identification of area-patches that would otherwise not be delineated in a generalization solution.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 7 citations exist upstream.