Evaluating label placement for augmented reality view management

Abstract
View management, a relatively new area of research inAugmented Reality (AR) applications, is about the spatiallayout of 2D virtual annotations in the view plane. Thispaper represents the first study in an actual ARapplication of a specific view management task:evaluating the placement of 2D virtual labels thatidentify information about real counterparts. Here, weobjectively evaluated four different placement algorithms,including a novel algorithm for placement based onidentifying existing clusters. The evaluation includedboth a statistical analysis of traditional metrics (e.g.counting overlaps) and an empirical user study guided byprinciples from human cognition. The numericalanalysis of the three real-time algorithms revealed thatour new cluster-based method recorded the best averageplacement accuracy while requiring only relativelymoderate computation time. Measures of objectivereadability from the user study demonstrated that inpractice, human subjects were able to read labels fastestwith the algorithms that most quickly prevented overlap,even if placement wasn't ideal.

This publication has 20 references indexed in Scilit: