Facsimile and static presentation processing – Static presentation processing – Attribute control
Reexamination Certificate
1999-11-29
2004-04-20
Lee, Thomas D. (Department: 2722)
Facsimile and static presentation processing
Static presentation processing
Attribute control
C358S518000, C358S525000, C382S167000
Reexamination Certificate
active
06724500
ABSTRACT:
This application is related to the field of color reproduction, and more specifically, to methods and systems for mapping from an additive color-space associated with a display to a subtractive color-space associated with a printer. The application describes an invention for use with digital color printers such as xerographic and ink-jet printers.
BACKGROUND
At the very least, a color printer is expected to reproduce on paper the colors that appear on a display monitor. A fundamental difficulty inherent in this task is that a display monitor generates color by illuminating combinations of red, green, and blue phosphors on a screen, whereas a printer generates color by depositing selected amounts of cyan, magenta, and yellow inks on paper. To accurately reproduce a screen color on paper, it is therefore necessary to determine, for each location on the display, what mixture of inks on paper corresponds to a particular combination of phosphors on the display. This process of determining how to represent the same color on two different media is referred to as a transformation from one “color space” to another.
A display monitor generates colors by forming a linear combination of three primary colors. For most display monitors, the three primary colors are red, green, and blue. However, any three colors can be used as primary colors provided that no primary color can be generated by a combination of the other two primary colors.
The color space associated with a display monitor is thus an additive color-space. An additive color-space is roughly analogous to a linear vector space in which the three primary colors form a basis and the color is a vector represented in terms of that basis. Because additive color-spaces are analogous to linear vector spaces, given the representation of a color in terms of one basis, one can easily represent the color in terms of a different basis by performing a matrix transformation.
Throughout this specification, we refer to the color space associated with the display as a “source color-space.” This color space can be an RGB color-space. However, it is common to transform the RGB color-space into a perceptually uniform space in which the distance between any two colors is indicative of the subjective perceptual difference between those two colors. An example of such a color space is the CIEL*a*b* color-space.
It would appear, based on the foregoing, that the problem of determining the ink mixture corresponding to a particular color in the source color-space is no more difficult than performing a matrix transformation. However, this assumes that the color space associated with the printer is also an additive color-space. It is not.
In contrast to the phosphors of a display monitor, an ink generates color by taking away selected colors that are already present. This is easily demonstrated by comparing the result of placing a spot of yellow ink on a black piece of paper with the result of placing a spot of ink on a white piece of paper. In the latter case, the white paper reflects all colors present in that light and the yellow ink suppresses the reflection of all but the yellow component of the incident light. In the former case, the black paper does not reflect any of the colors in the incident light. Hence, there is nothing for the yellow ink to suppress.
Unlike the additive source color-space associated with the display monitor, the color space associated with the printer is a subtractive color-space. An additional difficulty is introduced by the fact that printer inks absorb colors other than their nominal colors. In general, the transformation from an additive color-space to a subtractive color-space is non-linear and cannot be accomplished by a matrix transformation.
Conventional methods for performing a color transformation between colors in an additive color-space and corresponding colors in a subtractive color-space rely on a color-transformation table. In such methods, one determines the additive color combination to be reproduced, locates that combination on the table, and reads the appropriate mixture of inks that best reproduces that color. Since it is generally impractical to include every possible combination of additive primary colors in a color-transformation table, there are many instances in which a particular combination of additive primary colors combination is not listed. When this occurs, the conventional method interpolates the appropriate ink mixture on the basis of the ink mixtures associated with nearby combinations of additive primary colors.
In order to generate a color-transformation table, it is necessary to make colorimetric measurements for each ink mixture listed on the table. Since color-transformation tables generally include over a thousand distinct colors, the creation of a table is a daunting task. In most cases, a color-transformation table is generated only once, with a typical printer operating under normal conditions. Such a table cannot easily accommodate small manufacturing differences between individual printers, changes in performance of a printer as it ages, changes in the properties of ink over time, and the changes in the environmental conditions in which the printer operates. These changes result in subtle variations in color that are not readily accommodated by a single table. Although, in principle, these variations can be accommodated by generating a new color-transformation table, the number of colorimetric measurements required to do so makes this an impractical option.
The range of colors, or “gamut,” that can be printed defines a three-dimensional solid in the CMY color-space. The resulting solid is referred to as the printer gamut. The source gamut does not, in general, define a cube in the source color-space. Nevertheless, a conventional color-transformation table samples the printer gamut within a regular grid of points in the CMY color-space. This sampling scheme which would be appropriate if the printer gamut were a cube, results in a quantization error at the printer gamut boundary when the printer gamut is an irregular solid in the source color-space. This error results in clipping, or truncation, for colors outside the quantitized printer gamut boundary.
In some cases, the color image to be reproduced has a greater dynamic range than the color printer is capable of printing. This may occur, for example, when the image is a photograph. A related problem arises when the gamut of colors to be reproduced is different from that provided by the printer. For example, a computer monitor can display many bright secondary colors that cannot be reproduced in a subtractive color-space. In both these cases, the image to be reproduced must have its gamut compressed so that all colors to be reproduced are within the printer gamut. This need to perform gamut compression raises the question of how to select a target color within the printer gamut to correspond to an out-of-gamut color in the source image. Known methods for performing gamut compression agree that it is desirable for the target color in the printer gamut to have the same hue angle as the source color in the source image. Many known methods also agree that it is preferable to reduce lightness before reducing chroma. Other known methods of performing gamut compression map the out-of-gamut color to the closest point that lies within the printer gamut. These methods rely on precise knowledge of the gamut boundary. However, in modeling the geometry of a printer gamut, these methods do not take into account the highly irregular geometry of the printer gamut.
SUMMARY
The method and system of the invention relies on the observation that although the mapping from an additive color-space to a subtractive color-space is non-linear, the mapping from a limited portion of the additive color-space to a corresponding limited portion of the subtractive color-space is approximately linear. In effect, the mapping between the two color spaces is globally non-linear but piecewise linear. As a result, it is possible to partition the printer gamut into a plurali
Buckley Robert R.
Hains Charles M.
Fay, Sharpe. Fagan, Minnich & McKee, LLP
Lee Thomas D.
Xerox Corporation
LandOfFree
Piecewise color transformation by gamut partitioning does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Piecewise color transformation by gamut partitioning, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Piecewise color transformation by gamut partitioning will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3202296