Anti-aliasing algorithms based on self-similar multitudes

Abstract

In computer graphics, anti-aliasing is a software technique for diminishing jaggies -- stairstep-like lines that should be smooth. Jaggies occur because the output device, the monitor or printer, doesn't have a high enough resolution to represent a smooth line. The standard anti-aliasing reduces the prominence of jaggies by surrounding the stair-steps with intermediate shades of gray. Although this reduces the jagged appearance of the lines, it also makes them fuzzier. The suggested anti-aliasing algorithm uses the interpolation based on self-similar multitudes to remove the jaggies. In contrast to the well-known anti-aliasing algorithms, this approach changes neither brightness nor colors of a magnified image. This anti-aliasing algorithm was implemented in the special software for low vision people -- the L&H Magnifier that is being developed by CMS, Ukraine and Lernaut & Hauspie, Belgium. The preliminary tests confirmed that the developed technology improves the quality of zoomed images much better than the standard algorithms but it needs a large number of the computer operation. So, it's reasonable to use the anti- aliasing algorithm based on self-similar multitudes, when the magnification level is 4 and higher.