When presenting virtual images to a user performing a simple task requiring depth perception, the use of stereoscopic projections results in a ten-fold reduction in mean alignment error as compared to the use of monocular projections. Although average physiological interocular distance is 6.3 cm, it is found that any interocular distance of greater than 3 cm used in the stereo projection model is adequate to provide a user with maximal performance in the depth perception task. No statistically significant increase in performance can be correlated to increasing interocular distances greater than 3 cm. Since it is often beneficial to reduce retinal disparity between the left and right images to increase the presentable depth range, reduce image fusion problems, and reduce operator fatigue, these results suggests that smaller than physiological interocular distances should be considered when implementing a stereoscopic vision system for virtual environments and telepresence systems
Published in:
Virtual Reality Annual International Symposium, 1993., 1993 IEEE
Date of Conference: 18-22 Sep 1993