NACIS 2017

Visual Contrast and Hierarchy Relations of Cartographic Features Across Multi-Scale Map Displays: Map Evaluation Results | Le contraste visuel et les relations hiérarchiques des éléments cartographiques dans les affichages de cartes multi-échelles : résultats d’évaluations de cartes
Presenter: P. William Limpisathian, University of Oregon
Copresenter: Cynthia A. Brewer, The Pennsylvania State University
Visual contrast is a fundamental principle employed in cartographic design to highlight foreground elements on maps. However, limited software resources and research exist on how to maintain consistent visual contrast and hierarchy relations across map scales. This task is especially challenging in map series that are constrained by a color and layout template that affects fore-to-background feature recognition. Four combinations of lightness contrast, ten hue contrast combinations, and two line-width contrast levels are tested in combination with six basemap locations, each with two layout variants, generating 960 individual test maps or 480 map pairs. The results of aesthetic and hierarchy evaluations of five map pairs by each Amazon Mechanical Turk worker are analyzed with a Friedman non-parametric test and linear regression models. The project lays the empirical foundation for an application that provides users with map design schemes that preserve the cartographer's intended visual contrast and hierarchy relations.

Map-based Visual Storytelling: An Assessment of Emerging Design Considerations | La narration visuelle basée sur les cartes : évaluation des considérations émergentes en matière de conception
Presenter: Zihan Song, Esri
Copresenter: Robert Roth, University of Wisconsin–Madison
The use of maps in media increasingly is described as "data journalism," "spatial narratives," or "visual storytelling." Can cartographers actually adapt their visual designs to "tell" a story and, if so, what makes such map-based visual stories more or less understandable and meaningful? To this end, we conducted an online survey with 125 Amazon Mechanical Turk participants to evaluate emerging design considerations for map-based visual storytelling. Participants reviewed stories on two controversial topics—sea-level rise and presidential election campaign funding—and responded to questions on story retention, comprehension, and preference. Designs varied by their genre, or stylistically coherent visual story structures, and tropes, or additional visual design devices used to unfold a narrative. Results showed successful application of a three-act narrative to the visual stories, revealed differences in retention and comprehension by the presented genre or trope design, and illuminated the impact of individual differences on visual story reading.

Visual Hierarchy and Layout in the Practice of Map Design | La hiérarchie visuelle et la disposition dans la pratique de la conception cartographique
Presenter: Alex Tait, National Geographic Society
Writers of map design texts invariably include a chapter or two about the structure and composition of the map image and the map page. The main organizing terms that describe what a map designer should keep in mind when constructing a map are visual hierarchy and layout. Many other terms come into the conversation like balance, contrast, focus, gestalt principles and of course the biggie: figure-ground. My presentation will review many map design texts for descriptions of visual hierarchy and layout and then look carefully at how they work in the practice of mapmaking. Which principles are most important, surfacing in every project? Do all principles apply to both the map itself and the page or screen on which it resides with other components? Is the current body of knowledge for this topic complete and how can it be improved?

Texture Shading 2.0: Directional Texture Shading and More | Ombrage de texture 2.0 : l’ombrage de texture directionnel et autres évolutions
Presenter: Leland Brown
Texture shading is an alternative method for displaying terrain relief that relies on a mathematical operator called a fractional Laplacian. Fractional derivatives are well-suited to processing terrain data due to their scale invariance property, which respects the (approximately) fractal nature of real terrain, and their ability to emphasize ridge and canyon structures in the data. Related techniques are used in medical image analysis and computer vision applications.

New research shows that expanding the repertoire of fractional differential operators used in the algorithm allows a variety of enhancements, while preserving the inherent scale invariance. Incorporating fractional gradient information produces "directional texture shading," which is visually more analogous to traditional hillshading. Another new operator, the fractional Hessian matrix, shows promise in further expanding the versatility of texture shading. Meanwhile, interesting parallels with diffuse sky models have been found, which provide a link between these new methods and physical models of illumination.