Influence of graphics effects on perceiving depth in maximized virtual environments
Abstract
Improving depth perception accuracy in 3D stereoscopic virtual environments is an important research topic to exploit 3D vision technology in practice. This research studied the impact of depth cues on human depth perception in 3D virtual space. It also explored using graphics techniques as supporting tools to enhance measurement accuracy in 3D virtual environments. For this study we developed an application that applied various graphics effects, selectively, onto 3D medical images. Two experimental studies were conducted using a one-wall 3D projection system. The preliminary experiment investigated primitive graphics effects to identify noteworthy techniques impacting users' depth perception. This study showed that linear perspectives, gridlines, and background color were notable influencing factors on human depth perception when users viewed stereoscopic 3D medical images. Based on findings from the preliminary study, we designed several graphical tools that can assist users' depth measurement in the virtual environment. A follow-up study was conducted to assess the efficiency of the developed tools and carryover effects upon depth perception accuracy. The experimental results for presenting 3D medical images in the virtual environment showed a tendency of depth underestimation with wide variations in the absence of depth cues. Mixed results of both under and over estimation were noted when different types of graphics techniques were applied in the 3D stereoscopic medical images. The study result showed that users' variance when measuring depth was notably decreased when they used proposed graphics tools based on linear perspective, grid lines, and projection techniques. The effectiveness of the proposed tools was also well accepted by participants on their post-experimental survey. Test results for the carryover effect showed strong overestimation with wide measurement variations in the maximized virtual environment. In summary, this research investigated the influence of graphics techniques objectively in a specific virtual space presenting medical images. We found that utilizing graphics techniques can be a promising approach to enhance depth perception accuracy in virtual environments. This study suggested several future works and research directions to improve the efficiency of proposed techniques and will contribute to utilizing 3D vision technology in medical applications that require accurate measurement.
Degree
M.S.
Advisors
Ng, Purdue University.
Subject Area
Computer science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.