Browse Prior Art Database

Fast color conversion algorithm allowing non-uniform table sampling.

IP.com Disclosure Number: IPCOM000021029D
Original Publication Date: 2003-Dec-17
Included in the Prior Art Database: 2003-Dec-17
Document File: 3 page(s) / 26K

Publishing Venue

Motorola

Related People

Tom Schuneman: AUTHOR [+3]

Abstract

A method for performing fast conversion between color spaces is presented that allows for non-uniform sampling of a sparse color-look-up-table (CLUT). This method is useful in applications where a rapid conversion of, or correction to, colors is required; and the relationship of two color spaces can be represented by a sparsely populated CLUT. Applications include color scanning, printing, and display devices. This invention improves upon what has been previously presented in the literature by allowing the CLUT to be non-uniformly sampled.

This text was extracted from a Microsoft Word document.
This is the abbreviated version, containing approximately 57% of the total text.

Title:

Fast color conversion algorithm allowing non-uniform table sampling.

Authors:

Tom Schuneman

Ronald Recene

Peter Barada

Abstract:

A method for performing fast conversion between color spaces is

presented that allows for non-uniform sampling of a sparse

color-look-up-table (CLUT). This method is useful in

applications where a rapid conversion of, or correction to, colors

is required; and the relationship of two color spaces can be

represented by a sparsely populated CLUT. Applications include

color scanning, printing, and display devices. This invention

improves upon what has been previously presented in the literature

by allowing the CLUT to be non-uniformly sampled.

The technique uses a set of pre-computed tables that define the

relationships between an input code value, the surrounding sample

locations, and the distances from the input value to the surrounding

locations. These tables are used in conjunction with a

pseudo-random number generator to successively select among one of

the surrounding values. When iterated many times, the statistical

average result matches that which would have resulted using a

conventional interpolation. Many fewer addresses need to be

calculated, many fewer computations performed, and many fewer

accesses to the CLUT memory need to be made.

Background:

Color conversion is the process of converting color data captured,

created, or reference to on one device to data that are appropriate

for use on another device. Color conversion is used in printing

applications, photography, television, paints and pigments, grading

of food products, and many other applications.

Color conversion started out as a manual process performed by

artists. This evolved into photographic processes. With the advent

of electronics, systems were developed that implemented color

conversion in analog electronics. Color conversion is now performed

using digital electronics.

In many practical applications the color conversion transformation

is a highly non-linear vector function that is not accurately

represented as a set of polynomial, or other equations. Most

practical applications make use of a color-look-up-table (CLUT) to

represent the required transformation. To fully sample the input

color space for the CLUT would require more memory than is

practical. Most implementations sample the input space sparsely and

perform interpolation to compute the entries that are not discretely

represented in the table.

Color Conversion Technique:

This color conversion technique converts N input colors to M o...