Abstract
Prior work on depth perception in commercial augmented reality (AR) has indicated that egocentric distances are underestimated. One hypothesis as to why distances may be misjudged is an inability of the viewer to correctly perceive that a virtual object has surface contact with the ground rather than is floating above it. There are several factors in AR that can contribute to this effect, and they differ across types of AR technologies, making the question of how to improve depth perception in AR difficult. We carried out an experiment to examine egocentric depth judgments to virtual objects generated in state-of-the-art commercial optical and video see-through AR head-mounted displays (HMDs). In addition, we examined the influence of a rendered cast shadow on depth judgments for objects placed both above and on the ground. Although factors such as field of view, inertial characteristics of the HMDs, and rendering quality of the virtual objects differ, it is useful to make direct comparisons of technologies to understand any tradeoffs in spatial perception between them. Thus, we evaluated egocentric distance judgments in the Microsoft HoloLens 2, an optical see-through AR display, and the Varjo XR-3, a video see-through AR display. For each display, we asked participants (N = 24) to verbally report distance judgments to both grounded and floating targets that were rendered either with or without a cast shadow along the ground. We found significant underestimation of distance in both HMDs. We also found a significant difference between the two HMDs with the video see-through device (Varjo XR-3) showing more distance underestimation than the optical see-through device (HoloLens 2). Finally, we found that the vertical height of a virtual object and the presence of a cast shadow influence distance judgments