Facsimile and static presentation processing – Natural color facsimile – Color correction
Reexamination Certificate
1998-09-30
2001-07-24
Williams, Kimberly A. (Department: 2722)
Facsimile and static presentation processing
Natural color facsimile
Color correction
C382S162000
Reexamination Certificate
active
06266165
ABSTRACT:
TECHNICAL FIELD
The present invention relates generally to color printing equipment and is particularly directed to transforming an input color space into an output color space which is more suitable for a particular style of color printer. The invention is specifically disclosed as a scattered data transformation method that uses training points and training vectors to define the effect on input points existing in the input color space in a manner that localizes the transforming effect on each of the input points due to the training points and training vectors.
BACKGROUND OF THE INVENTION
Color image devices have become quite common, not only in televisions and computers CRT displays, but also in color printers such as ink jet printers. Color data operations and manipulations vary widely between the different types of color imaging devices. To ensure a quality color image, different approaches to color rendering are required based on the physics and operating characteristics of the device being characterized, and different imaging technologies utilize different primary colors. For example, devices which emit light to create color (e.g., televisions and CRT displays) use the additive color mixing primaries red, green, and blue (“RGB”) to create all of the colors within their color gamut (i.e., the range of all displayable colors). In addition, devices which use reflected light to create color (e.g., printers) use the subtractive color mixing primaries cyan, magenta, and yellow (CMY) to obtain a wide gamut (in this case, the range of all printable colors) of colors. For an RGB image displayed on a CRT to look as good as a CMY print made by a color ink jet printer, a color conversion is required to convert the RGB input data to CMY output data. Oftentimes, conventional color data transformations required for the color conversion are mathematically and/or computationally complex.
It is common to refer to the gamut of colors associated with an imaging device as a “color space.” The color space associated with digital color imaging devices is generally specified as a three-dimensional entity (because the human visual system is trichromatic). The coordinates of a color space can vary from input RGB data, to output CMY data, or to coordinate systems which mimic the response of the human visual system. The most common of these coordinate systems is CIELAB, which utilizes three coordinates (L*, a*, and b*) to characterize any color. The L* coordinate represents the lightness of a color. L* values range from 0 to 100, where zero (0) represents black and 100 represents white. The second coordinate, a*, represent the redness or greenness of a color. An a* value of zero (0) corresponds to a color that is neutral between red and green. A positive a* value represent a redish color while a negative a* is greenish. The third coordinate, b*, is similar to a*, except that it represents a tradeoff between yellowness and blueness, rather than redness and greenness.
Developers of today's color imaging systems commonly transform color information between many color spaces in the process of characterizing a printing or displaying device. Generally, the mathematical manipulation of color information is performed in a so-called “device independent color” space, such as CIELAB, to ensure that the operations are being performed in a way consistent with the way in which people see. The original data can be considered as existing in an “input color space” and, after the data transformation, the color space can be called an “output color space.”
In situations where one type of physical device (such as a color monitor) has received a set of color data that corresponds to its format, and then that color data is desired to be utilized with a different type of physical device (such as a color printer), then it will be obvious that the color gamut will have to be significantly modified when transforming the color data from one format to the other. In other situations, a color space may be defined for a particular type of physical device (such as a color printer), however, improvements are desired for certain portions of the color space, and again, a transformation is required to create a new look-up table of colors to be used with the physical output for the device of interest. In general, a “warping” or “morphing” algorithm is used to transform either two-dimensional or three-dimensional data to create the new, improved color space. Such warping/morphing algorithms have been used in the past, with varied results.
One standard approach is known as “polynomial warping” which has certain drawbacks. For example, polynomial warping requires inversion of a large ill conditioned matrix which is numerically and computationally impossible due to the size of the matrix and floating point precision problems. In addition, if a “training point” (a point that is hand-picked by the color space system designer to have a certain input coordinate and a certain corresponding output coordinate) also happens to be an “input point” (i.e., a point in the input color space that will be transformed into the output color space), then the output point corresponding to that input point is generally not exactly defined by that training point. Furthermore, the effects of a training point are not localized to that training point within the color space, and in fact can have a significant effect over the entire transformed output color space.
Another standard approach is known as an inverse distance interpolation (also known as “Shephard's Method”). In the inverse distance interpolation method, if the training points are “do nothing” training points, the transform does not degenerate into an identity transform, in that the transform will actually modify the non-training point even though the training points are such that input equals output. Moreover, the effects of a training point are not generally localized to that training point.
Other conventional methods for modifying color space have been patented, including U.S. Pat. No. 5,289,295 (by Yumiba), which discloses a color adjustment apparatus that includes a color space conversion circuit that obtains a chromaticity signal in a rectangular coordinate system of a plane representing hue and saturation components. Instead of using the uniform color space expression of color, L*a*b*, this Yumiba invention uses the CIE 1976 uniform observer color space coordinates, L*u*v*. Using a two-dimension coordinate system, the “original chromaticity signal” having the values (u*, v*) are converted to a “corrected chromaticity signal” called (u
0
*, v
0
*). This corrected chromaticity signal is used to create a “target color chromaticity signal” (u
0h
*, v
0h
*) which is used to convert the entire color plane. A “pointed color chromaticity signal setting circuit” creates the converted chromaticity signal (u
0
*, v
0
*). A “target color chromaticity signal setting circuit” is used to create the target color chromaticity signal (u
0h
*, v
0h
*).
A particular area of color space can be affected by a “target color chromaticity signal” which affects certain points to a greater extent if these points are spaced closely to a particular target point that is chosen by the “pointed color chromaticity signal setting circuit.” In other words, a single rectangular coordinate point is chosen as the center of the target area, and the color adjustment area has all of its multiple points affected to a certain degree, but the points most closely spaced to the center of that adjustment area are most affected. The Yumiba device can be used as an instantaneous color adjustment device (e.g., for a color TV or a color monitor).
U.S. Pat. No. 5,583,666 (by Ellson) discloses a method for transforming an input color space to an output color space where each of a plurality of specified colors or color regions are constrained to be transformed by one or more explicitly specified color calibration or color enhancement strategies. The constraints are applied to a subset of the points in the input color space that specifies t
Huang Xuan-Chao
Nystrom Brant Dennis
Reel Richard Lee
Gribbell Frederick H.
Lexmark International Inc.
Williams Kimberly A.
LandOfFree
Method for morphing n-dimensional data using a scattered... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for morphing n-dimensional data using a scattered..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for morphing n-dimensional data using a scattered... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2552475