Abstract
Recent advancements in augmented reality (AR) have led to the development of several applications in domains such as architecture, engineering, and medical training. Typically, these applications present users with 3D virtual images in real environments that would not easily be portrayed otherwise (e.g., floor plans, arteries, etc.). The way users perceive the scale (i.e., size, distance, etc.) of such displays is important for decision making and learning outcomes. The current study aimed to asses users’ perception of distance to AR images, which has previously been shown to be underestimated in other virtual technologies. We focused our investigation on the influence of ground contact, which is an important cue for distance perception that many AR images lack because they are presented above the ground surface. Furthermore, binocular cues should be particularly important for users to overcome the lack of ground contact in many AR images. To test both the influence of ground contact and the importance of binocular cues, we conducted a study where participants were asked to blind walk to AR cubes presented at 3m, 4.5m, and 6m. Participants completed this task with cubes rendered on the ground surface or 0.2m above the ground surface. Additionally, we had each participant perform this task under monocular and binocular viewing conditions. We found that participants blind walked farther to AR cubes presented above the ground surface and that this effect was exaggerated under monocular viewing conditions. However, we found that participants blind walked shorter to AR cubes presented on the ground which was not expected. We also found underestimation of cube distance, regardless of where the cubes were presented or the viewing condition. Our results suggest that distance in AR environments is generally underestimated and that a lack of ground contact influences users’ perception of distance to AR images.
Acknowledgement: Office of Naval Research grant N00014 - 18 - 1 - 2964