Perceptual Display Hierarchies for Visualization

Abstract

The advent of computers with high processing power has led to the generation of large, multidimensional collections of data with increasing size and dimensionality. This has led to a critical need for ways to manage, explore and analyze large, multidimensional information spaces. Visualization lends itself well to the challenge of exploring and analyzing these datasets by managing and presenting information in a visual form to facilitate rapid, effective, and meaningful analysis of data by harnessing the strengths of the human visual system.

Most visualization techniques are based on the assumption that the display device has sufficient resolution, and that our visual acuity is adequate for completing the analysis tasks. However, this may not be true, particularly for specialized display devices (e.g., PDAs or large-format projection walls). Our goal is to: (1) determine the amount of information a particular display environment can encode; (2) design visualizations that maximize the information they represent relative to this upper-limit; and (3) dynamically update a visualization when the display environment changes to continue to maintain high levels of information content. A collection of controlled psychophysical experiments were designed, executed, and analyzed to identify thresholds for display resolution and visual acuity for four visual features: hue, luminance, size, and orientation.

In computer graphics, level-of-detail hierarchies are often used to reduce geometric complexity in situations where full resolution models are unnecessary. Using results from our experiments, we applied similar logic to a visualization environment. If certain properties of a dataset cannot be seen because their current visual representation falls below a resolution or acuity threshold, they need not be included in the visualization. We built level-of-detail perceptual display hierarchies to automatically add or remove information from a visualization as viewer's perspective on the data changes, or as the data is visualized across different display devices. Our display hierarchies are combined with existing rules on the use of perception in visualization to improve our ability to construct visualizations that are perceptually optimal for a particular dataset, analysis tasks, and viewing conditions. We conclude by visualizing multidimensional weather data using a prototype visualization system that applies our perceptual display hierarchy rules.