Distance estimation of near‐field visual objects in stereoscopic displays

Abstract

Stereoscopic displays have a promising future because of recent advancements and popularity of handheld devices and maturing head mounted displays. Gesture interaction such as pointing, selection, pinching, and manipulation are now possible in the current virtual environments, where accurate distance judgment is required. In this paper, we address the perception of exocentric distance in stereoscopic displays under two target orientations: horizontal and vertical. Three parallax conditions (on screen, 5 cm from screen, and 10 cm from screen) were considered, where the screen was fixed at a distance of 100 cm from the observer. Four levels of center‐to‐center distance between 10 and 50 cm were employed. The perceptual matching task revealed underestimation in all conditions. The overall judgment of exocentric distance was only about 80% of the actual. We also found a main effect of distance and interaction between layout and distance to be significant. The two important findings of this study are that underestimation of exocentric distance increases as the separation between virtual targets increases and that in vertical orientation, accuracy increases with closer targets. However, the main effects of layout and parallax on accuracy of judgment were not significant. Engineering implications of the results are also discussed in this paper.

Perceptions of an automotive load space in a virtual environment

“…In contrast to some of the previous research (Waller, 1999;Wartenberg and Wiborg, 2003;Lin et al 2014), our study found no evidence for systematic over-or under-estimation in either virtual or real conditions. This lack of a consistent pattern (which is seen for egocentric distance perception, Renner et al 2013) supports the notion that perception of distance and space in near-field VE applications is specific to the environment set up and task conditions.…”

“…Previous research has indicated that further studies of spatial understanding of VR are required in new contexts (Wartenberg and Wiborg, 2003) and that the results are specific to the environment and task conditions (Waller, 1999;Wartenberg and Wiborg, 2003;Lin et al, 2014;Kenyon et al, 2007). Overall, our results showed that perception of space was not significantly different between the virtual (CAVE) and real vehicle for subjective ratings of distance and volume.…”

“…Renner et al (2013) attribute this to several factors, including the method used to measure the distance, the technical attributes of the VR, the composition of the virtual environment, presence and individual differences. There is also evidence of inaccuracies in exocentric distance judgement in virtual environments (Henry and Furness, 1993;Waller, 1999;Wartenberg and Wiborg, 2003;Lin et al, 2014) although this is affected by several factors, including display type, feedback, parallax, and orientation of targets in the virtual environment. As exocentric judgement is more closely linked to estimates of load space width, height and depth, the relevant findings from previous research are outlined below.…”

“…Lin et al (2014) asked participants to judge the differences between the centers of two targets presented at different distances from a screen (parallax) on which targets were presented in either vertical or horizontal orientations. They found underestimation of distances across all conditions, which became greater with wider spaced objects.…”

Interaction and visual performance in stereoscopic displays: A review

“…Considering relative distance between objects, few studies reported the accuracy of distance judgment in exocentric space . Henry and Furness, Wartenberg and Wiborg, and Geuss et al .…”