GitHub Account Google Scholar Account LinkedIn Account Twitter Account flickr Account vialab Research Group Profile Page

Stereoscopic 3D Perception Study
Evaluating the Perceived Relative Depth of Stereoscopically
Rendered Two-Dimensional Shapes

October 2012 – December 2012

Technologies: C++

We performed a perception study to examine the human's ability to perceive relative depth of stereoscopically and monoscopically rendered two-dimensional shapes. Participants were required to estimate the relative distance ratios of shapes with respect to a reference object.

To display stereoscopic 3D content, a TFT-LCD monitor, model ZM-M240W, by Zalman Tech. Co., Ltd. was used. It is a passive stereoscopic 3D LCD monitor requiring polarized 3D glasses. The screen resolution was set to 1920x1080. Questions were displayed on a second monitor, and could be answered by using a mouse.

To avoid accidental motion parallax, blurring effects, as well as fluctuations in perceived depths, each participant was required to place their head on a chin rest. While seated, the distance from each participant's forehead was made to be exactly 1 metre away from the base of the 3D monitor.

Materials used:
  • Passive stereoscopic 3D LCD monitor
  • Polarized 3D glasses
  • 3D Content
  • Chin-rest
Perception Study
Our hypotheses:
  • Stereoscopic viewing condition would improve relative depth perception compared to the monoscopic viewing condition
  • Estimating relative depth distances would have an accuracy of fifty percent or less in both viewing conditions
  • Shape type would not affect relative depth estimation accuracy
  • Confidence levels would be higher for the stereoscopic viewing condition