Using spatial updating bias to explore the robustness of a collaborative geovisualization model.

摘要

VR currently lacks a universal structure that facilitates real-time distant collaborative visualization between in situ field agents and distant experts. Additionally, procedures used by others to date in validating the robustness of such a VR model have mostly been anecdotal and can not be applied to all types of mimicked environments. This dissertation evaluated visual spatial updating during wayfinding, finding ones way, in the physical world and comparatively evaluates the robustness of spatial updating in a mimicked virtual reality. Spatial updating is described here as the ability to remember landmark location when they become occluded during wayfinding. Ecological Visual Perception Theory suggested that structures within the proximal view field bias landmark remembrance. This biasing factor was used to test robustness between the physical environment and the mimicked virtual reality, termed here a Collaborative Geovisualization Model, which is composed of a linked sequence of photorealistic panoramic images. Displayed by the computer, these panoramic images capture the scene of a pause and gaze waypoint, a specific location during wayfinding where a person would stop walking and gaze around their new position to spatial update. Subjects in each test group, the real world and virtual reality, initially viewed four landmarks and were told the location of one landmark. Being told the position of a landmark explored the concept of Spatialized Language ---obtaining audible spatial information, and whether subsequent spatial updating bias of this landmark was similar to spatial updating bias of the four initially viewed landmarks. Virtual reality subjects were not provided with movement between pause and gaze waypoints, they only hyperlinked between panorama images. By comparing both environments, this dissertation explored if proprioception, body movement, enhanced spatial updating. Results from this experiment concluded that both subject groups biased spatial updating similarly in accordance to the Ecological Visual Perception Theory, that Spatialized Language of landmarks was as robust as viewing, and that proprioception did not enhance spatial updating.