Browse Prior Art Database

Procedure for Optimum Choice of a Small Number of COLORS From a Large COLOR Palette for COLOR Imaging

IP.com Disclosure Number: IPCOM000061581D
Original Publication Date: 1986-Aug-01
Included in the Prior Art Database: 2005-Mar-09
Document File: 5 page(s) / 56K

Publishing Venue

IBM

Related People

Braudaway, GW: AUTHOR

Abstract

The present procedure selects a small number of colors, to be used for image presentation on a cathode ray tube (CRT) display, from among those available in a very large color palette. The selected colors produce an image having minimum entropy. Minimum image entropy means that the per pel average color error is a minimum. 1. THE COLOR-CUBE MODEL A "color-cube" will be used in explaining the algorithms that, as a whole, comprise the present procedure. Colors in a three-phosphor CRT can be represented as three-component vectors. The magnitude of each of the three components, ranging from zero to one, represents the intensity or saturation of one of the three primary phosphor colors. Each of the three color components can be laid along an edge of a cube radiating and increasing in intensity from a common corner (Fig. 1).

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 25% of the total text.

Page 1 of 5

Procedure for Optimum Choice of a Small Number of COLORS From a Large COLOR Palette for COLOR Imaging

The present procedure selects a small number of colors, to be used for image presentation on a cathode ray tube (CRT) display, from among those available in a very large color palette. The selected colors produce an image having minimum entropy. Minimum image entropy means that the per pel average color error is a minimum. 1. THE COLOR-CUBE MODEL A "color-cube" will be used in explaining the algorithms that, as a whole, comprise the present procedure. Colors in a three-phosphor CRT can be represented as three-component vectors. The magnitude of each of the three components, ranging from zero to one, represents the intensity or saturation of one of the three primary phosphor colors. Each of the three color components can be laid along an edge of a cube radiating and increasing in intensity from a common corner (Fig. 1). The common corner where all three components are zero is the black corner of the cube. Each of the three edges represents all possible shades of one of the primary colors. They are designated the red, green, and blue edges. Thus, any hue or shade representable by a three phosphor system corresponds to a point within or on the edge of the color cube. The human eye, with a normal pathology, is capable of distinguishing approximately 350,000 different colors [1]. Altogether about 128 hues are distinguishable. Except for spectral extremes, the wave- lengths of the distinguishable hues lie within three nanometers of those of their spectral neighbors. If the colors vary only in saturation, the eye can distinguish from 16 (for yellow) and 23 (for red and violet) intensity levels. All of these measurements were made with natural sunlight as illumination. The three phosphors of color CRTs are not capable of producing all hues and saturations available in natural sunlight. So for them an approximation of the human visual limit can be produced employing no more than six bits of resolution per primary. This yields 64 discrete intensity levels per primary and 262,144 distinct colors. 2. MAPPING A COLOR IMAGE INTO THE COLOR CUBE Each color image can be thought of as a two-dimensional array of rectangular pels, with each pel having a finite size. The color associated with each pel is the average color contained in the corresponding small rectangular element of the image. The color of each pel can be decomposed or separated into an additive combination of the three primary phosphor colors. The intensities of the three primary colors range from zero to one. Each color-separated image can therefore be represented by three distinct two-dimensional arrays, one corresponding to each of the primary colors. Each element of each array corresponds to a pel of the image. The value of each element is the magnitude of the primary color component of the corresponding pel. The entire color image can be mapped pel- by-pel into the color cube. Cor...