Browse Prior Art Database

Efficient Reduction Algorithm

IP.com Disclosure Number: IPCOM000110678D
Original Publication Date: 1992-Dec-01
Included in the Prior Art Database: 2005-Mar-25
Document File: 2 page(s) / 70K

Publishing Venue

IBM

Related People

Aghili, H: AUTHOR

Abstract

Disclosed are two efficient algorithms for reducing (compressed) digital binary monochrome images by integer reduction factors. One algorithm produces high quality gray-scale output, whereas the other offers a better performance but generates lower quality binary monochrome output.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Efficient Reduction Algorithm

       Disclosed are two efficient algorithms for reducing
(compressed) digital binary monochrome images by integer reduction
factors.  One algorithm produces high quality gray-scale output,
whereas the other offers a better performance but generates lower
quality binary monochrome output.

      The reduction operator is a useful primitive for building image
systems.  It is often needed because input devices, such as scanners,
typically have much higher resolution (>300 dots per inch) than
output devices like displays (<100 dots per inch).  The modern
display devices, on the other hand, can render multiple shades of
gray, typically <256 shades.  This capability may be capitalized by
the reduction routines to compensate for some of the quality which is
lost as a result of the resolution modification.

      Efficiency of the reduction operators is often a major concern.
In such cases, one typically deals with millions of bits of data, and
slight carelessness can result in disastrous response time.  The
algorithms which are disclosed here are devised with the above as the
primary design point.  A central observation is that standard digital
computers are very efficient for processing integer numbers, but fall
short when manipulating bits.  As a result, implementations of
reduction algorithms which take a bitmap for input images do not
typically offer acceptable performance.

      To alleviate the above problem, the disclosed algorithms assu...