Perception of spatial layout in real and virtual worlds as a function of geometric field of view

Abstract

As human-machine interfaces grow more immersive and graphically-oriented, virtual environment systems become more prominent as the medium for human machine communication. Often virtual environment (VE) are built to provide exact metrical representations of existing or proposed physical spaces. To evaluate the process of model development, the present experiment examined participant``s ability to reproduce a complex spatial layout of objects having experienced them previously under different viewing conditions. The layout consisted of seven boxes arranged on a flat plane. These objects could be viewed in a free binocular real-world condition and in non-immersive virtual conditions varying with different geometric fields of view (GFOV). Viewing conditions were a between-subject variable with 24 participants randomly assigned to each condition. Performance was assessed using mapping accuracy by comparisons of relative inter-object distances. Mapping results showed a significant effect of viewing condition where the real world condition and VE with 60° GFOV condition were superior to both VE conditions with 30° and 90° GFOV. Also results showed that there is no significant difference between real world condition and VE with 60° GFOV. These results suggest that the situation model resulting from interaction with a virtual environment was indistinguishable from interaction with real objects at least within the constraints of the present procedure. Also they suggest that of the three geometric fields of view tested (30°, 60°, 90°), a 60° degree geometric field of view was found to produce least overall relative distance judgment error within the constrains of the present experiment.