Perspective Projection - Transformations and Projections in Computer Graphics - page 130

4. At a certain point, the four images will merge into three. Concentrate on the
central image and you will suddenly see it in three dimensions. The effect is more

noticeable if the original image pair is in vivid colors.
You can try this technique on the image pair of Figure 3.49, but avoid prolonged
viewing and concentration, which may lead to eye fatigue.
The Anaglyph Method
This approach to stereoscopic viewing combines the right-eye and left-eye images (partly
overlapping) in one image but in different colors, normally red and blue (or cyan)
but sometimes red and green. The method requires the use of special glasses with
different color filters for the two eyes, as illustrated here. (See also [kspark 05] for several
well-known Escher drawings that have been converted to three dimensions, mostly as
anaglyphs.)

Red

Blue

The red filter on the left eye looks red because it reflects red light. Any other colors
are partly absorbed and partly transmitted through the filter. Thus, this filter lets the
blue parts of the image through to the left eye. Similarly, the blue filter on the right
lets only the red parts of the image to the right eye.
Warning . Some people may be sensitive to these glasses. If you feel discomfort or
if you get tired very quickly, take off the glasses and take a break. In any case, try to
use these glasses for short periods and only to view an anaglyph image. They are not
intended for normal use!

From the Dictionary
Anaglyph. From Late Latin anaglyphus, carved in
low relief. Also from Greek anagluphos (to carve).

An anaglyph image is encoded in one of three ways as follows:
Color . The left-eye image is left mostly in its original colors, but certain crucial
parts, such as edges, curves, lines, and points, are painted blue (or cyan or green). The
right-eye image is treated similarly with red. Thus, a color anaglyph (Figure 3.50a)
preserves much of the original colors of the image, but its red/blue (or red/green)
stereo information is diluted throughout the image. The result is that many images
lose their depth information in this format and don't look three-dimensional. This is
especially true if the original image has vivid red or blue colors. However, if an image
does look good in a color anaglyph, it looks real and vivid.
Gray . The two original images are converted to grayscale and the same crucial
elements are painted red and blue. A gray anaglyph image (Figure 3.50b) is therefore
seen in grayscale, but its depth information is normally easy to perceive.
Pure . The right-eye image is entirely converted to shades of red. The left-eye
image is treated as in the color anaglyph method. The combined image looks reddish