Facsimile and static presentation processing – Static presentation processing – Attribute control
Reexamination Certificate
1998-08-26
2002-11-19
Grant, II, Jerome (Department: 2624)
Facsimile and static presentation processing
Static presentation processing
Attribute control
C358S529000, C358S003030, C382S167000
Reexamination Certificate
active
06483606
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to color printing, and in particular to color printing in which a small set of colorants, such as inks, are placed on a sheet in a manner to create a large range of apparent colors. The present invention is most useful in digital printing devices, such as xerographic or ink-jet printers.
Digital halftoning is the process of converting a continuous tone image to bi-level. Many output devices, including many printers and some cathode ray tube (“CRT”) and liquid crystal display (“LCD”) based devices are intrinsically bi-level. In other words, the process only prints or displays a dot or the process does not print or display a dot. Thus, a variety of geometrical patterns are created such that a group of dots and blank areas represent the continuous tone image as closely as possible. Because the halftoned image is only an approximate representation of the continuous tone image, there are differences between the continuous tone image and the halftone image. Those areas of the halftone pattern which do not match the original image are noise or error. An objective of much research in digital halftoning is reducing the amount of visible noise.
There are two broad classifications of digital halftoning: conventional, passive halftoning and active halftoning. Conventional, passive halftoning yields an appearance that is similar to that provided by classical analogue processes developed before digital techniques were available. It is most appropriate for devices that cannot display isolated pixels. Active halftoning typically implements some type of error diffusion to produce an image having a more pleasing appearance. Images created using error diffusion tend to have noise at higher and, hence, less visible spatial frequencies. Error diffusion is most appropriate for devices that can display isolated pixels.
Digital printing devices implementing halftoning technology, such as electrophotographic laser printers and ink-jet printers, are well known. In color digital printing devices, a small set of primary colorants, typically black, yellow, magenta, and cyan, are selectably placed in different areas on a print sheet. Small areas of each primary colorant are then optically blended together to create a large range (i.e., gamut) of colors which would be apparent to an observer.
Error diffusion allows a large gamut of colors to be obtained from a small number of colorants in a halftoned image. In order to print a small area of a desired source color using error diffusion, the source color is located in color space relative to the locations of the primary colorants, such as cyan, magenta, yellow, and black. The colorant which is closest to the target color in color space is then selected for one pixel. However, the error, essentially meaning the Euclidean distance in color space between the source color and the selected colorant, is recorded, and is in effect distributed or diffused to the image data of subsequent neighboring pixels. In brief, this diffusion of each error to neighboring pixels influences the decision of which primary colorant to use in those neighboring pixels. The overall effect is, over a reasonably large number of pixels, an optical blend resulting in the desired source color.
Recently, particularly in the technology of ink-jet printing, there has been developed a hardware option in which selectably available colorants are provided beyond the usual pure primary colorants of cyan, magenta, yellow, and black. For example, some designs may include colorants of additive colors, such as red, blue, and green, in addition to the subtractive colors of cyan, magenta, and yellow. Other designs may include colorants representing a lighter or diluted version of another primary color, such as a light-cyan, which is 50% lighter than regular-cyan. The use of such additional colorants can enhance and/or enlarge the available gamut associated with a particular apparatus. One particular additional colorant, which will be the subject of the embodiment of the present invention described below, is, in addition to a pure black K colorant, a grey LK colorant. The grey LK colorant is a 50% dilution of black K ink. Selectable use of the grey LK ink will, of course, be helpful in the creation of monochrome halftones, such as black-and-white photographs, and also for the creation of non-saturated colors. Other additional colorants which will be discussed include light-cyan LC, light-magenta LM, and light-yellow LY.
These additional colorants, such as light-cyan or grey, are considered intra-gamut colorants. More specifically, the pure colorants such as cyan or magenta define a gamut and are, therefore, disposed at the corners of a cube
10
representing a color space of the gamut. The intra-gamut colorants, on the other hand, are disposed within the color space of the gamut. In other words, the intra-gamut colorants are along the edges, on the faces, or within the cube
10
. While this is useful for obtaining accurate representations of colors, such as pastels, which are near the white or grey areas of a gamut, use of such intra-gamut colorants can interfere with the error-diffusion colorant selection process.
For example, an error-diffusion selection process may occasionally decide that an intra-gamut colorant such as grey LK, or a mixture of on-edge colorants such as light-cyan LC plus light-magenta LM (i.e., a color on one of the faces of the color cube), is desirable for a particular pixel in an image. In either of these situations, the resulting error from selection of the intra-gamut color, when diffused to influence the selection of colorants for neighboring pixels, may require the selection of colors which are out of the gamut. These out-of-gamut colors would be physically impossible to obtain with the available colorants. In other words, selection of a colorant either inside the gamut or on the face of the gamut may lead to errors which require selection of colorants outside the gamut, which are not available. Consequently, artifacts from large errors may occur.
To illustrate the drawback of the prior art system,
FIG. 1
illustrates a cube
10
representing a color space CS
1
. The color space characterizes a gamut indicating a three-dimensional volume including every combination of the various primary colorants available for a particular printing apparatus. Starting at the lower corner W of
FIG. 1
, a white value indicates no colorant is placed for a particular pixel in an image to be printed. Three (3) axes extend from the lower corner W. Numeric values are associated with points along each of the three axes. More specifically, the value of zero (0) is assigned to the lower corner W and the value of 255 is assigned to the point of full color saturation along the different directions. These three directions represent contributions of three primary colorants, yellow Y, cyan C, and magenta M. The more a particular colorant is apparent, the farther along any particular axis and, consequently, the higher the value (up to 255) of the particular color along the axis in the color space.
A color desired to be printed in an image is indicated as a source color at the location marked X. The color X is disposed near the inside surface of the center of the face within the gamut CS
1
formed by the points marked C, W, M, B. Therefore, the color X is closest to the location in color space of grey LK. The particular problem addressed by the present invention occurs when selecting which colorants to use in order to obtain this desired source color X by a combination of, in this case, cyan C, grey LK, magenta M, light-cyan LC, and light-magenta LM.
Under the basic well known technique of error diffusion to select colorants, a source color X desired to be printed is located in color space, and the colorant selected for a particular pixel is the colorant which is closest, by Euclidean distance, to the source color X in color space. Once the particular colorant is selected, the distance, here indicated as d, is calculated in terms of both magnitude an
Klassen R. Victor
Waldron Brian L.
Fay Sharpe Fagan Minnich & McKee LLP
Grant II Jerome
Xerox Corporation
LandOfFree
Error diffusion on moderate numbers of output colors does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Error diffusion on moderate numbers of output colors, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Error diffusion on moderate numbers of output colors will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2926064