The “Sense of Depth” of a Stereoscopic 3D Capture and Display System

Stereoscopic 3D imaging systems try to mimic the binocular vision of our eyes. However, there are many examples of 3D photography and video where the stereo camera does not have the same separation as our eyes and the display does not show the same field of view as our normal vision. Examples include stereo microscopy and endoscopy and 3D telephoto sports photography. Unaided vision never sees scenes such as these, yet very comfortable and effective 3D images can be displayed in an electronic system. How do we compare the 3D “experience” of one camera/display configuration to another? — We all know that camera separation is an important factor for providing a stereoscopic depth effect, but so is the convergence distance of the cameras and their magnification. Other important factors are the size of the screen, the viewer's distance from the screen, and his or her eye separation. How do all these parameters work together? And, in turn, how can we use them to provide the viewer a good viewing experience? — This paper will parameterize the overall stereoscopic sense of depth as the product of a camera contribution (CDE = MC*CRC) and a display contribution (DDE = MD/CRV). (MC and MD are camera and display magnification factors and CRC and CRV are camera and viewer Convergence Ratios. “Convergence Ratio” is stereo baseline divided by convergence distance.) We will use these parameters to compare the depth effect produced by various stereo camera systems, including our unaided and aided eyesight, and we will see how the depth effect of normal vision can be used to evaluate the sense of depth produced by a pair of telescopic cameras zoomed into the quarterback in a football game!

Published

2011-10

Content type

Original Research

Keywords

Stereoscopic, 3D, Stereo 3D, Depth Effect, Parallax, Stereo Base, IPD, Stereo Cameras, Stereo Displays, Stereo Imaging, 3D Imaging, Stereoscopic Computer Vision

DOI

10.5594/M001095

ISBN

978-1-61482-940-9