Shape Computation Lab

Spatial Similarity Metrics

 


Title:

Spatial Similarity Metrics: Graph Theoretic Distance Measurement and Floor Plan Abstraction

Authors:

Thomas Grasl and Athanassios Economou

Editors:

Andy Dong, Andrew Moere Vande, and Asko Riitahuhta

Conference:

Proceedings of the Twelfth International Conference of Computer-Aided Architectural Design Futures (CAADFutures)

Series:

CAADFutures 07

Pages:

251-263

Publisher:

Springer, Dordrecht: Sydney, Australia

Publication date:

June 2007

DOI:

10.1007/978-1-4020-6528-6_19

Keywords:

Graph representation, Graph grammars, Abstraction

Abstract:

Comparing space allocation programs computationally is a resource intensive task. This paper introduces a method which reduces the complexity of floor plan graphs to facilitate the problem. In a first step the nodes are labeled according to a classification scheme. This in its own right reduces the complexity of the comparison by introducing more diversity among the nodes. Subsequently the graph can be coarsened in several steps based on the category information. This does not only further reduce the necessary computation, but can also offer a visualization of and additional insight into the essential structure of the space allocation program at hand.

links