A Comprehensive Multi-criteria Model for High Cartographic Quality Point-Feature Label Placement

Abstract

The lettering process, including assigning names to point features, is an essential part of map production. While there have been numerous and varied research efforts to automate point-feature label placement (PFLP), none of them seems to have taken into account the many well-established cartographic precepts for point-feature annotation used by human cartographers. As a result, current fully automated solutions are limited in their expressive power. The PFLP problem is still vital, therefore, and solving it is a compelling challenge. This article presents a comprehensive multi-criterion model that complies with almost all well-defined cartographic placement principles and requirements for PFLP, allowing for a significant increase in toponym density without affecting legibility. The proposed model, expressed as a qualityevaluation function, can be used by any mathematical optimization algorithm to resolve the automated label-placement problem. Through an application of the proposed model tested on volunteered geographic (VGI) data and the creation of sample parameter settings, the article illustrates that a high level of cartographic quality for PFLP can be achieved through the integrated approach, comparable to the lettering produced by an expert cartographer. RÉSUMÉLe processus de lettrage, y compris l'attribution de noms à des caractéristiques ponctuelles, fait partie essentielle de la production de cartes. De nombreux efforts de recherche variés ont visé à automatiser la mise en place d'étiquettes de caractéristiques ponctuelles (MPECP), mais aucun de ces efforts ne semble avoir tenu compte des nombreux préceptes cartographiques bien établis portant sur l'annotation des caractéristiques ponctuelles utilisées par les cartographes humains. Il s'ensuit que les solutions entièrement automatisées courantes ont un pouvoir expressif limité. Le problème de la MPECP demeure donc vital et sa solution constitue un défi difficile à relever. Cet article présente un modèle multicritériel détaillé qui se conforme à presque tous les principes bien définis de mise en place cartographique et aux exigences de la MPECP, ce qui permet d'augmenter considérablement la densité toponymique sans que cela ait d'effet sur la lisibilité. Le modèle proposé, exprimé sous forme d'une fonction qualité-évaluation, peut être utilisé par n'importe quel algorithme d'optimisation mathématique pour régler le problème de la mise en place automatisée des étiquettes. En appliquant le modèle proposé mis à l'essai sur des données géographiques fournies volontairement et en créant des contextes paramétriques témoins, l'article démontre qu'il est possible d'obtenir une grande qualité cartographique pour la MPECP grâce à l'approche intégrée, comparable au lettrage produit par un cartographe expert.Mots clés : cartographie automatisée, mise en place automatisée des étiquettes, MPECP, optimisation combinatoire, cartographie SIG

Beyond Maximum Independent Set: An Extended Integer Programming Formulation for Point Labeling

“…With respect to cartographic quality, it would be interesting to see how our model extension performs with a more sophisticated weight function such as the one by Rylov and Reimer [9]. For example, it would be reasonable to reduce the weight of a label candidate if it occludes much information of the map background.…”

“…An extension of the ILP for the basic map-labeling problem is necessary, however, to address the following two criteria of the model of [9]:…”

“…More precisely, the following criteria of the model of [9] can be modeled with label weights that depend only on the input:…”

“…Kern and Brewer [8] and, more recently, Rylov and Reimer [9] have provided overviews of criteria that have been considered in the cartographic literature on map labeling. Thereby, Rylov and Reimer have developed what they call a "comprehensive multi-criteria model" for point labeling, which also deals with graphical clutter and ambiguous label-point associations.…”

“…Before we summarize our contribution, we note that Rylov and Reimer [9] expressed their model in a form that is similar to an ILP but does not satisfy the definition of an ILP. In fact, they do not use the term integer (linear) program but the much more general term combinatorial optimization problem, which correctly describes their model.…”