Stereoscopic viewing

Authors: Jim Blanchard and Reiko Tsuneto

One of the most important features of any display of environments that are three-dimensional is the perception of the third dimension, depth. This brief article introduces the geometry of stereoscopic vision, and then walks you through a couple of examples of the use of stereoscopic vision in displays.

Geometry

The human visual system (described in the articles about perception and depth cues) has a physical configuration that supports two separate images to be gathered (each eye). Because the brain combines these into one, a small, but important mathematical difference exist between these images. This minuscule difference is represented by the figure below.

In this figure, A equals the angle of the eye to see images and motion (motion only at high acute and obtuse angles), B equals the point at which the nose interferes with the eyes ability to see images and motion, and C equals the distance between the centerline of the eyes' lens. This tiny offset, C, causes the location of the image in the field of view for that eye, A in the left eye, for example, to be different than the same image in the right eye Look at the figure below to see how the eyes each view the same image (object) differently.

This effect adds to the depth perception information the brain needs to derive an image with three dimensions. Combining these images makes a three-dimensional image.

From this small offset in focal centerlines, there are problems as well. Some of these will become apparent when designing virtual space with objects that are certain varying distances. The human vision system only supports this stereoscopic depth perception for objects that are between approximately 8 inches and 18 feet. Beyond that, the vision system uses other cues to detect depth and distance. To explore one of the phenomenon of stereoscopic vision systems, use the following figure to find your own blind spot. Hold the figure at arm's length. Close the left eye and focus of the circle/cross. Move the figure slowly towards you. Notice the spot disappear as the figure gets closer. This is your blind spot.

At greater distances, the eyes can detect very subtle movements and changes in spatial relationship between objects. These changes are the primary source of information in the distance and depth perception cues for viewing in excess of 18 feet.

Advantages of stereopsis

With complex images it has been found that subjects could easily use stereopsis to enhance 3-D spatial judgments when monocular depth cues were ambiguious. Depth cues additively interact with each other to get more vivid 3-D representation. Stereopsis appears to be a compelling depth cue except when in conflict with motion or occlusion. It has certain advantages:

Surface properties such as luster, scintillation, and sheen are difference in luminance and color between the left and right retinal images, and cannot be seen in single image. Surface inspection and analysis, or similar tasks may require a stereoscopic image.

With stereoscopic vision, the brain can more easily seperate a correlated image from static and dynamic visual interference(noise). When there is poor image quality due to interference such as low resolution, limited grey scale, motion blurr, and raster noise, binocular vision diminishes the influence of the noise.

The brain is not being stressed to extract a three dimensional comprehension from two dimensional input. Offering stereo displays reduces a perceptual workload.

Note that these advantages are additive; stereoacuity is the cumulative result of all dimensional depth cues.

References

[CARTER92] Carter, William, The advantage of single lens stereopsis, SPIE Vol.1669 Stereoscopic Displays and Applications III. 204-210.

Human Interface Technology Laboratory