pamsistoaglyph

Updated: 5 April 2009

Table Of Contents

NAME

pamsistoaglyph - convert a single-image stereogram to a red/cyan anaglyphic image

SYNOPSIS

pamsistoaglyph [--invert] [--sep=number] [--minsep=number] [--gray=number] [in_netpbmfile

All options can be abbreviated to their shortest unique prefix. You may use either white space or an equals sign between an option name and its value.

DESCRIPTION

This program is part of Netpbm.

pamsistoaglyph reads a Netpbm image as input and produces a Netpbm image as output.

pamsistoaglyph takes a single-image stereogram (SIS) such as those produced by pamstereogram and converts it to a red/cyan anaglyphic image such as those produced by ppm3d. Many people have trouble tricking their eyes into focusing beyond the image in front of them and are therefore unable to perceive the 3-D shape hidden within a single-image stereogram. Anaglyphic stereograms are easier to perceive in 3-D but require a pair of red/cyan glasses such as those often used to watch 3-D movies. The goal of pamsistoaglyph is to help people who have trouble viewing single-image stereograms see the intriguing 3-D effect.

pamsistoaglyph can convert single-image random-dot stereograms (SIRDS), wallpaper stereograms, and even dual-image stereograms to anaglyphic images.

OPTIONS

For most images, no command-line options need to be specified. The following options are available, however, for unusual circumstances:

--invert
Swap the left- and right-eye images. pamsistoaglyph assumes that its input represents a wall-eyed stereogram and generates the anaglyphic image accordingly. If the generated image appears to recede into the page where it should pop out of the page (and vice versa), this typically implies that the input image represents a cross-eyed stereogram. Use --invert to correct the image depth.
--sep=number
Specify the distance in pixels between the left- and right-eye images. Essentially, this corresponds to the distance between repetitions of the background pattern. The --sep option should rarely be necessary as pamsistoaglyph is fairly good at determining automatically the eye-separation distance.
--minsep=number
This option is similar to --sep but constrains pamsistoaglyph only to a minimum eye-separation distance. Any distance larger than number is acceptable. The --minsep option should rarely be necessary as pamsistoaglyph is fairly good at determining automatically the eye-separation distance. The default value for the minimum eye-separation distance is 10% of the image width; this value seems to work well in practice.
--gray=number
Limit the number of gray levels to use when searching for the optimal eye-separation distance. Because pamsistoaglyph looks for repeated patterns, it is vulnerable to being confused by slight variations in color. By reducing the input image to grayscale and capping the number of gray levels, pamsistoaglyph ameliorates the effects of unintentional color variations (such as those caused by conversion from a low-quality JPEG image, for example). The default of 63 seems to work well so the --gray option should rarely be necessary.

NOTES

The registration algorithm used by pamsistoaglyph was developed specifically for this program. As far as the author knows, there are no existing algorithms for converting stereograms to anaglyphs. The algorithm works as follows:

  1. Convert the image to grayscale to increase the ability to identify matches.
  2. Count the number of pixels that match N pixels ahead in the image for all N in [1, width/2].
  3. Maintain a running mean (μ) and standard deviation (σ) of the number of matched pixels.
  4. Store the N corresponding to each spike in the number of matched pixels. A spike is defined as a tally that exceeds the mean plus one, two, or three standard deviations. Only the first spike of a given standard-deviation multiplier is stored.
  5. If a tally greater than μ+3σ was encountered, return the corresponding N. If not, then if a tally greater than μ+2σ was encountered, return the corresponding N. If not, then if a tally greater than μ+σ was encountered, return the corresponding N. If not, then return the N that produces the minimum average distance between matched pixels (i.e., #matches divided by #pixels). If no such N exceeds the minimum allowable eye-separation value, return zero to indicate failure.
  6. If the algorithm returned zero, rerun the algorithm independently on each row of the input image and return the median of all N that exceed the minimum allowable eye-separation value. If no such N exists, abort with an error message.

HISTORY

Scott Pakin wrote pamsistoaglyph in April 2009. It first appeared in Netpbm in Release 10.47 (June 2009).

AUTHOR

Copyright (C) 2009 Scott Pakin, scott+pbm@pakin.org.

SEE ALSO


Table Of Contents