Schedule of Events | Search Abstracts | Talk Sessions | Poster Sessions
Poster Session A, Wednesday, May 20, 10:15 am – 11:00 am
Board 4
Perceived size of virtual objects at different heights in augmented reality
Maria Kon1,2, Gregory Francis3; 1U.S. Naval Research Laboratory, 2Purdue University, 3University of Southern Denmark
Outdoor visual illusions like the moon size illusion, where the moon appears especially large near the horizon, are inherently difficult to systematically study. Optical see-through augmented reality (AR) head-mounted displays offer a way of conducting controlled visual psychophysics experiments with manipulable virtual objects that appear in the real world. The few studies on size perception of these virtual objects used physical or verbal estimations, while varying target size (Chiu et al., 2025) or distance and speed (Benjamin et al., 2024), indicating near veridical size estimates. Here we investigate perceived virtual object size using the method of adjustment while varying height to see whether a version of the moon size illusion could be produced in controlled conditions with an eye level 0.3m virtual target sphere 2.5m away. On each trial, participants saw two virtual spheres and used a controller to adjust the reference sphere size to match the target. In Experiment 1, the reference could appear at 0, 0.8, or 1.6m above target height. We conducted this experiment in seven locations and found one where participants underestimated the size of the higher sphere, indicating that something like the moon size illusion holds in some environments at this scale. Using the same task and environment, Experiment 2 tested whether the effect scales with spheres at distances of 1.6, 4.3, and 7m. The effect of distance was significant, with size misestimation increasing with distance. Exploring whether the effect was due to the room’s unique ceiling, Experiment 3 was identical to Experiment 1 except for an additional block with a virtual ceiling similar to a room where we did not get the effect. Performance was nearly identical across blocks. Thus, AR enables study of the moon size illusion beyond the lab and, to some extent, investigation of what environmental features contribute to the effect.



