Perceived Depth in Virtual and Physical Environments

Theoretically, stereopsis provides accurate depth information if information regarding absolute distance is accurate and reliable. However, assessments of stereopsis often report depth distortions, particularly for virtual stimuli. These distortions are often attributed to misestimates of viewing distance caused by limited distance cues and/or the presence of conflicts between ocular distance cues in virtual displays. To understand how these factors contribute to depth distortions, I conducted a series of experiments in which depth was estimated under a range of viewing conditions and cue combinations. In the first series (Chapter 2), I evaluated if conflicts between oculomotor distance cues drive depth underconstancy observed in virtual environments by comparing judgments of virtual and physical objects. The results showed that depth judgments of physical stimuli were accurate and exhibited depth constancy, but judgments of virtual stimuli failed to achieve depth constancy. This failure was due in part to the presence of the vergence-accommodation conflict. Further, prior experience with each environment had a profound effect on depth judgments, e.g., performance in virtual environments was enhanced by limited exposure to a similar task using physical objects. In Chapter 3, I assessed if limitations of virtual environments contributed to previous failures of linear combination models to account for the integration of stereopsis and motion cues. I measured the perceived depth of virtual and physical objects defined by motion parallax, binocular disparity, or their combination. Accuracy was remarkedly similar for both environments, but estimates were more precise when depth was defined by binocular disparity than motion parallax. A linear combination model did not adequately describe performance in either physical or virtual conditions. In Chapter 4, I evaluated if reaching to virtual objects provides distance information that can be used to scale stereopsis using an interactive ring game. Brief experience reaching to virtual objects improved the accuracy and scaling of subsequent depth judgements. Overall, experience with physical objects or reaching-in-depth enhanced performance on tasks dependent on distance perception. To fully understand how binocular depth perception is used to interact with objects in the real world, it is important to assess these cues in a rich, full-cue natural scenes.