Matching qualitative spatial scene descriptions á la Tabu

Abstract

Matching spatial scene descriptions requires appropriate representations and algorithms based on the application at hand. This work outlines a simple model for matching qualitatively described spatial scenes extracted from sketch maps. Standard qualitative constraint networks are combined to provide a suitable qualitative representation for a sketched spatial scene. Two scenes are then matched using an implementation of the Tabu search metaheuristic, employing standard and specialised data structures. We give a detailed description of the representation and algorithm, and examine the performance of the model using an example dataset. © Springer-Verlag 2013.