US 5,699,491 AGrant
Printer Driver Having Gamut-Mapped Colors
Issue Date:1997-12-16
•39 Claims
•27 Drawing Sheets
Abstract
A color printer driver for a printer includes a reproduction look-up table (LUT) which stores CMY values for colors both in-gamut and out-of-gamut of the printer. For close-in out-of-gamut colors, the reproduction LUT stores perceptually matched CMY values whereas for far-out-of-gamut colors, the reproduction LUT stores saturation matched CMY values. CMY values for the out-of-gamut colors are obtained by a dual-cone gamut mapping technique in which an actual cone of N-cone arranged surfaces is defined so as to mimic the actual gamut of the printer, an ideal cone of M-cone arranged surfaces is defined so as to mimic an ideal gamut such as the gamut of a typical color monitor, a target out-of-gamut color is projected onto the actual cone and de-saturated until an in-gamut CMY value is obtained, the same target color is projected onto the ideal cone to obtain a fully-saturated CMY value, and a weighted average is obtained of the in-gamut CMY value and the fully-saturated CMY value, the weight being based on a measure of distance between the out-of-gamut target color and its projection onto the actual cone.
Metadata
Assignee
- Canon Information Systems, Inc.
Inventor
- Ron Barzel
Application Information
Application Number:US 4905571
Filing Date:1995-06-15
Priority Date:1995-06-15
Art Unit:268
Classifications
IPC:
G06K 1500G03F 308H04N 146
Field of Search:
395358345382109;101;115520;519;518;523-525;515;504;501199;154167
Patent Drawings (27 sheets)
Description
Background of the Invention
1. Field of the Invention
The present invention pertains to a method and apparatus for building and using, in a printer driver or color management system, color look-up tables which determine the colors that the printer driver sends to a color printer in response to requests to print specific colors. The specific colors requested to be printed may include colors that are not printable, i.e., are not in the gamut of the color printer. For those colors in particular, the invention provides for gamut mapping of the unprintable colors to printable colors in such a way as to preserve both perceptual color reproduction and saturation color reproduction.
2. Description of the Related Art
Recently, as the availability of color monitors and color printers has increased, it is more and more commonplace for a computer user to create a full color image on a color monitor, such as by using desktop publishing software, and then to request a full color printout of that image on a color printer.
However, color printers and color monitors form color images differently. Specifically, a color monitor is a light emitting device; colors are formed on color monitors by additive light processing in which light from three color primaries, generally red, green and blue, are added. Printed images, on the other hand, simply reflect ambient light; colors are formed by subtractive light processing according to the way ambient light is affected by three subtractive primaries, generally cyan, magenta and yellow (and sometimes black).
Additive light processing and subtractive light processing are fundamentally different. As a result, the range of colors displayable on a monitor is different from the range of colors printable by a printer. FIG. 1 is the CIE 1931 chromaticity diagram showing the range (or "gamut") of colors displayable by a monitor (area "A") and the range (or "gamut") of colors printable by a printer (area "B"). As seen in FIG. 1, the range of colors displayable on a monitor is generally greater than the range of colors printable by a printer. This is because a monitor is a light emitting device and is able to display colors with greater saturation. There are, however, some low saturation areas such as at area 10 where a printed image, which is light-subtractive, has greater color range than a monitor.
Because of the difference between the gamuts of printable and displayable colors, it has heretofore been difficult to print color images which are perceived as faithful reproductions of displayed color images. Specifically, it is simply not possible to print a color in areas like out-of-gamut area 11 which are outside the range "B" of printable colors. Accordingly, even though those colors may be seen on color monitors, they cannot be printed on a color printer.
U.S. Pat. No. 5,299,291, which is assigned to the assignee of the present invention and the contents of which are incorporated herein by reference as if set forth in full, describes a system which maps out-of-gamut colors to printable colors. In the system of the aforesaid U.S. Pat. No. 5,299,291, out-of-gamut colors are adjusted to printable colors within the printer gamut by building printer tables which produce colors which vary smoothly in the out-of-gamut regions and which exhibit monotonic increases in lightness. A warping process is utilized so as to compensate for the so-called "Abney effect", in which two identically-hued colors are perceived by a viewer as colors with different hues because of differences in color saturation. The system in this aforesaid U.S. Pat. No. 5,299,291 produces good results from a perceptual standpoint, meaning that viewers who view a printed color and a displayed color will generally perceive those colors as being the same. Accordingly, the gamut mapping system in U.S. Pat. No. 5,299,291 is sometimes referred to as "perceptual matching".
Room for improvement, however, is needed in an area known as "saturation matching". Specifically, especially in business-oriented computer-generated graphics, fully-saturated colors are often used to depict graphics such as bar charts, pie graphs, emphasis lines, and the like. Using perceptual matching techniques will generally print a color that is perceived to be the same as that displayed, but the printed color will no longer be saturated. Ordinarily, this is undesired since the viewer, in such business-graphic applications, generally desires to print a fully-saturated color even if that color is not exactly the same hue as that of the color that is displayed. Such as process is known as "saturation matching", meaning that saturated colors on a display should be printed as saturated colors, even if the color hues are not exactly the same.
While perceptual matching techniques have been proposed, and saturation matching techniques have also been proposed, since the goals of those two matching techniques are different, it has heretofore been difficult to realize both of those goals in the same system. Accordingly, a need exists for a system which provides good perceptual matching for relatively unsaturated colors, while at the same time providing good saturation matching for highly saturated colors.
Summary of the Invention
The present invention addresses the above-noted need by providing a system which provides good perceptual matching for relatively unsaturated colors, while at the same time providing good saturation matching for highly saturated colors. According to the invention, for an out-of-gamut target color, the target color is projected onto two different surfaces in a color coordinate system such as HLS coordinates. The first surface is referred to as an "actual cone" and mimics the actual gamut of a printer. The second surface is referred to as an "ideal cone" and mimics an ideal gamut such as the gamut of a color monitor. The projection onto the actual cone is de-saturated until an in-gamut CMY value is obtained; this value represents a good "perceptual match" between the out-of-gamut target color and the printable color. The projection onto the ideal cone gives an in-gamut CMY value which represents a good "saturation match" between the target color and the printable color. A weighted average between the two CMY values is then obtained based on the distance from the target color to its projection onto the actual cone, thereby resulting in a printable color which is a good perceptual match for relatively unsaturated colors and which is a good saturation match for relatively saturated colors.
Thus, according to one aspect of the invention, a color printer driver for a printer includes a reproduction look-up table (LUT) which, for each of plural colors both in-gamut and out-of-gamut of the printer,-stores color component values for printing by the printer. Means, responsive to a command to print a color, are provided for outputting color component values stored in the reproduction LUT in correspondence to the commanded color. For close-in out-of-gamut colors, the reproduction LUT stores perceptually matched color component values while for far-out-of-gamut colors, the reproduction LUT stores saturation matched color component values. For out-of-gamut colors between close-in out-of-gamut colors and far-out-of-gamut colors, the reproduction LUT stores weighted averages between a perceptual match and a saturation match.
In a further aspect of the invention, a color printer driver which outputs color component values for printing by a printer includes a reproduction look-up table (LUT) which, for each of plural colors both in-gamut and out-of-gamut of the printer, stores color component values for printing by the printer. Means, responsive to a command to print a color, are provided for outputting color component values stored in the reproduction LUT in correspondence to the commanded color. For out-of-gamut colors, the reproduction LUT stores color component values obtained by dual-cone gamut mapping in which an actual cone of N-cone arranged surfaces is defined which mimics an actual gamut of the printer, in which an ideal cone of M-cone arranged surfaces is defined which mimics an ideal gamut, in which a target out-of-gamut color is projected onto the actual cone and de-saturated until in-gamut color component values are obtained, in which the same target color is projected onto the ideal cone to obtain fully-saturated color component values, and in which a weighted average is obtained of the in-gamut color component values and the fully-saturated color component values, weight being based on the distance between the projection onto the actual cone and the out-of-gamut target color.
This brief summary has been provided so that the nature of the invention may be understood quickly. A more complete understanding of the invention can be obtained by reference to the following detailed description of the preferred embodiment thereof in connection with the attached drawings.
Brief Description of the Drawings
FIG. 1 is a chromaticity diagram showing how the gamut of colors printable on a printer is related to the gamut of colors displayable on a monitor.
FIG. 2 is a view showing the outward appearance of representative computing equipment which incorporates a printer driver according to the invention.
FIG. 3 is a detailed block diagram showing the internal construction of FIG. 2.
FIG. 4 is a functional block diagram which stresses functional connectivity of the FIG. 3 block diagram.
FIG. 5 is a flow diagram showing printing by a printer driver according to the invention.
FIG. 6 is a flow diagram showing how to construct a reproduction LUT for use in a printer driver according to the invention.
FIG. 7 is a flow diagram for explaining how to construct a forward mapping for a typical printer.
FIG. 8 is a view for explaining how to smooth a forward mapping.
FIG. 9 is a view for explaining dual-cone gamut mapping.
FIGS. 10 through 14 are views showing how to project a target out-of-gamut color onto actual cones and ideal cone in accordance with the dual-cone gamut mapping of the invention.
FIG. 15 is a flow diagram for explaining how to smooth a reproduction LUT.
FIGS. 16 through 25 are figures referred to in the Appendix to this application.
Detailed Description of the Preferred Embodiment
FIG. 2 is a view showing the outward appearance of representative computing equipment which incorporates a printer driver according to the invention. Shown in FIG. 2 is computing equipment 20 such as an IBM PC or PC-compatible computer having a windowing operating system such as a Microsoft Windows operating system. Computing equipment 20 is provided with a display monitor 23 having a display screen 22 on which computing equipment 20 displays images to the user. Computing equipment 20 is further provided with a floppy disk drive 24 with which removable floppy disk media may be read or written, fixed disk drive 25 for storing data files and application program files, a keyboard 26 for permitting input of text data and manipulation of objects displayed on display screen 22, a pointing device 27 such as a mouse or the like which is also provided to permit manipulation of objects on display screen 22, and a combined speaker/microphone 29. A conventional color printer 30, such as a color bubble jet printer, is also provided. Also provided are connections to a network 31 or to an ordinary voice telephone line 32 for sending and receiving voice and/or facsimile messages.
While a bubble jet printer is presently preferred, any color printer which forms full color images by mixing color primaries in amounts set by corresponding color component values, such as a color laser beam printer or the like, is suitable in the practice of the invention.
In accordance with operator instructions, and under control of the windowing operating system, stored application programs such as graphics application programs, drawing application programs, desktop publishing application programs and the like, are selectively activated to process and to manipulate data. Also in accordance with operator instructions, and based on those stored application programs, commands are issued to display images on monitor 23 and to print images appearing on monitor 23, and those images are then printed on printer 30 as described more fully hereinbelow.
FIG. 3 is a detailed block diagram showing the internal construction of computing equipment 20. As shown in FIG. 3, computing equipment 20 includes a central processing unit (CPU) 40 such as programmable microprocessor interfaced to a computer bus 41. Also interfaced to computer bus 41 is speaker/microphone interface 42, display interface 44, network interface 45, fax/modem/telephone interface 46, printer interface 47, and floppy disk drive interface 49.
Main memory 51 such as random access memory (RAM) interfaces to computer bus 41 so as to provide CPU 40 with access to memory storage. In particular, when executing stored application program instruction sequences such as those associated with application programs stored on disk 25, CPU 40 loads those instruction sequences from disk 25 (or other storage media such as media accessed via network 31) into main memory 51 and executes those stored program instruction sequences out of main memory 51.
ROM (read only memory) 52 is provided for storing invariant instruction sequences such as start-up instruction sequences or basic input/output operating system (BIOS) sequences for operation of keyboard 26.
As shown in FIG. 3, and as previously mentioned, fixed disk 25 stores program instruction sequences for the windowing operating system and for various application programs such as a graphics application program, a drawing application program, a desktop publishing application program, and the like. In addition, stored on fixed disk 25 are image files such as are displayed on monitor 23 or printed on printer 30 under control of a designated application program. Fixed disk 25 also stores a monitor driver which controls how RGB color primary values are provided to display interface 44, and printer driver 34 which controls how CMY color component values are provided to printer interface 47 for printout by printer 30. Other device drivers are also stored on fixed disk 25, for providing appropriate signals to the various devices (such as the speaker and the microphone and the network) connected in computing equipment 20.
Ordinarily, application programs and drivers stored on disk 25 need first to be installed by the user onto disk 25 from other computer readable media on which those programs and drivers are initially stored. For example, it is customary for a user to purchase a floppy disk or other computer readable media on which a copy of printer driver 34 is stored. The user would then install printer driver 34 onto disk 25 by inserting the purchased floppy disk into floppy disk drive 24 and by commanding CPU 40 to copy printer driver 34 from the floppy disk onto disk 25. It is also possible for the user, via modem interface 46, to download printer driver 34 from a computerized bulletin board to which the drivers had previously been uploaded.
FIG. 4 is a functional block diagram showing how computer 20 interacts with monitor 23 and printer 30. Shown in FIG. 4 are computer 20 with monitor driver 33, printer driver 34, CPU 40, RAM 51 and ROM 52, all arranged functionally rather than in the structural arrangement shown in FIG. 3.
As mentioned above, using keyboard 26, an operator can cause CPU 40 to execute stored program instructions which cause color images to be displayed on monitor 23 and which cause corresponding color images to be printed on color printer 30. Specifically, and in cooperation with the stored program instructions in the application program stored on disk 25, CPU 40 derives a color image for display on monitor 23. CPU provides the color image to monitor driver 33 which in turn derives RGB values for each pixel in the monitor 23. The RGB values are provided via display interface 44 to monitor 23 where those values are displayed.
Upon request, CPU 40 also feeds a color image to printer driver 34 for printing by color printer 30. Printer driver 34 derives CMY values for each pixel of the color image based on the color values provided from CPU 40. The CMY values are determined in accordance with a reproduction look-up table (LUT) 35. As will be described more fully hereinbelow, reproduction LUT 35 can accept colors which are both in-gamut for printer 30 as well as out-of-gamut for printer 30. The purpose of reproduction LUT 35 is to provide CMY values which are printable by printer 30 whether the input colors are in-gamut or out-of-gamut.
Printer driver 34 feeds the CMY values via printer interface 46 to printer 30 where they are stored in bitmap memory 37 within printer 30. The CMY values may be altered by printer 30 so as to provide a black (hereinafter "K" value) in accordance with standard under color removal ("UCR") or gray component replacement ("GCR") techniques. Alternatively, it is possible for printer driver 34 to provide an appropriate K value.
In any event, bitmap memory 37 may store a full bitmap image of the printed image, or it may store only a band or partial bitmap image. When sufficient color data, namely CMYK data, is stored in bitmap memory 37, a color print head 36 reciprocates across a platen adjacent a sheet of paper. In the presently preferred embodiment, print head 36 includes 32 ink jet nozzles arranged in a four column by eight row pattern. The nozzles in the first column all eject droplets of cyan ink; the nozzles in the second column all eject droplets of magenta ink; the nozzles in the third column all eject droplets of yellow ink; and the nozzles in the fourth column all eject droplets of black ink. The nozzles are controlled independently in accordance with the color data in bitmap memory 37 such that in one reciprocation of print head 36 across the platen, eight rows of pixels are printed.
FIG. 5 is a flow diagram showing how printer driver 35 selects CMY values from the color data provided by CPU 40. In step S501, printer driver 34 receives RGB values for a location (x,y) in bitmap memory 37. In step S502, printer driver 34 derives device independent color coordinates from the RGB value. Preferably, the device independent coordinates are L*a*b* coordinates in CIELAB space. This is because the coordinates in CIELAB space can enclose the gamut of any printer or monitor, and also because both tri-linear and tetra-linear interpolation techniques work well in this space.
Flow then advances to step S503 which looks up corresponding CMY values stored in reproduction LUT 35 at the derived L*a*b* coordinates. In step S504, interpolation is performed so as to obtain an interpolated CMY value in a case where the derived L*a*b* coordinates are not identically found in reproduction LUT 35. Appropriate interpolation techniques include, as mentioned above, both trilinear interpolation and tetrahedral interpolation.
In step S505, the CMY values are stored in bitmap memory 37 at location (x,y). If desired, the CMY values may be modified before storage, for example, by appropriate UCR or GCR techniques so as to obtain a K value.
In step S506, printer driver 34 determines whether the bitmap memory has been completed, or if a needed part or band of bitmap memory 37 has been completed. If the bitmap memory has not been completed, then flow returns to step S501 in which the next RGB value is received for the next location (x,y) in bit map memory. On the other hand, if the bitmap memory has been completed, or if a sufficient area of the bitmap memory has been completed (such as an eight row long band corresponding to the eight rows of ink jet nozzles in head 36), then flow advances to step S507 where gamma correction is performed. Gamma correction corrects the CMY values in bitmap memory so as to achieve a uniform distribution of luminanos. Gamma correction may be carried out either by printer driver 34 or by printer 30, or by a combination thereof, as appropriate. In step S508, UCR or GCR is performed, if not already performed, so as to derive a black value for location (x,y) in bitmap memory. UCR in the present embodiment may be performed by the szmple expedient of selecting the minimum value of CMY and assigning that value to the black value. Then, each of the CMY values is adjusted by subtracting the assigned black value.
The order of steps S507 and S508 is not critical, and those steps may be switched, for example, to accommodate a particular color printing technique such as continuous tone, dither or error diffusion.
In step S509, color printing is initiated using the CMYK values stored in bitmap memory 37.
FIG. 6 is a detailed flow diagram showing the steps, including the gamut-mapping steps, by which CMY values are stored in reproduction LUT 35. As will be described in further detail below, those steps generally include construction of a forward mapping of CMY input to L*a*b* output for each particular printer, such as by printing a large number of colors printable by each different printer 30 thereby to measure the gamut for each different printer 30, testing each individual L*a*b* value in reproduction LUT 35 to determine if it is in-gamut or out-of-gamut, for in-gamut colors storing a colorimetrically-matched CMY value in reproduction LUT 35, while for out-of-gamut colors applying a dual-cone gamut mapping technique so as to obtain a gamut-mapped CMY value, repeating the process for each individual L*a*b* value in reproduction LUT 35 until the reproduction LUT is full, smoothing the reproduction LUT, and storing the reproduction LUT in printer driver 34.
The process steps of FIG. 6 result in a reproduction LUT, which is included in printer driver 34, in which for close-in out-of-gamut colors a perceptually matched CMY value is stored and in which for far-out-of-gamut colors a saturation matched CMY value is stored.
The process steps shown in FIG. 6 need only be performed once, for each different kind of printer 30, to derive a reproduction LUT 35 specifically tuned to each different kind of printer, ordinarily, these steps are performed once by a printer manufacturer, to be used in a printer driver which is sold to end users on a floppy disk as part of a printer. It is also possible for a printer manufacturer to upload the printer driver to, for example, a computer bulletin board, from which end users can download the printer driver to their disk 25.
With specific reference to FIG. 6, step S601 constructs a forward mapping of CMY input values to L*a*b* output values, so as to measure the gamut of printer 30. Preferably, this is achieved by printing a very large subset of all colors printable by printer 30. For example, in the printer used in the presently preferred embodiment shown in FIG. 2, a subset of about nine different values for each of CMY values gives 9.sup.3 =729 different CMY color combinations. Based on these 729 color combinations, a forward mapping from CMY input to L*a*b* output is constructed and stored in a table, as described more fully in connection with FIG. 7.
Step S602 opens a blank reproduction LUT 35 in which will be stored CMY values to be printed based on L*a*b* input values corresponding to both in-gamut and out-of-gamut colors. Once filled with CMY values, the blank reproduction LUT will be included in printer driver 34 so as to map L*a*b* input to CMY output, as described above in connection with FIG. 5.
Steps S603 through S606 store CMY values in each L*a*b* location in reproduction LUT 35, and those steps are repeated until the reproduction LUT is full. Thus, in step S603, a particular L*a*b* location in reproduction LUT is tested to determine if it is an in-gamut color, based on the gamut measured in step S601. One preferred method for determining whether a particular L*a*b* value is an in-gamut color is to attempt an inversion, using numerical techniques such as the well-known Newton's inversion method, to determine whether it is possible to obtain a CMY value based on the forward mapping obtained in step S601. If the inversion attempt fails to converge to a numerically printable CMY value, then step S603 determines that the particular L*a*b* value is out-of-gamut. On the other hand, if the inversion attempt converges to a numerically printable CMY solution, then step S603 determines that the particular L*a*b* value is an in-gamut color.
If the L*a*b* value is an in-gamut color, then flow advances to step S604 in which the printable CMY value for the L*a*b* value, which is a colorimetrically matched value, is stored in reproduction LUT 35.
If step S603 determines that a particular L*a*b* value is an out-of-gamut color, then flow branches to step S605 so as to gamut-map the L*a*b* target color into a printable CMY value. As detailed more fully below in connection with FIG. 9, gamut mapping is performed by applying a dual-cone gamut mapping technique in which the target L*a*b* color is first projected onto an actual cone and then projected onto an ideal cone, the projection onto the actual cone is de-saturated until an in-gamut CMY value is found, and a weighted average is taken of this in-gamut CMY value with the CMY of the projection onto the ideal cone, the weighting being based on the distance between the projection onto the actual cone and the target L*a*b* color. The gamut-mapped printable CMY value is then stored, as shown in step S604, for the L*a*b* value in reproduction LUT 35.
Step S606 determines whether reproduction LUT 35 is full. If the reproduction LUT is not yet full, then flow returns to step S603 until CMY values have been stored in each L*a*b* location in the reproduction LUT.
When the reproduction LUT is full, flow advances to step S607 in which the reproduction LUT is smoothed. Smoothing is described more fully below in connection with FIG. 15 in which a local relaxation and a shift is applied to the reproduction LUT.
Flow then advances to step S608 in which the smoothed reproduction LUT is stored in printer driver 34. As mentioned above, the printer driver, with reproduction LUT 35, is then stored on computer readable medium, such as by being stored onto a floppy disk or being uploaded to a computer bulletin board, for distribution to end users in connection with a printer.
FIG. 7 is a flow diagram showing how to construct a forward mapping of CMY input to L*a*b* output by printing out various combinations of CMY values so as to measure the printable gamut of colors. As detailed more fully hereinbelow, the steps shown in FIG. 7 involve measuring, in L*a*b* coordinates, an evenly spaced grid of CMY colors, applying a white point adjustment to the measured L*a*b* values, applying a black point adjustment to the measured values, applying a gray axis adjustment to the measured values, and then smoothing the adjusted measurements.
Specifically, in step S701, a regular grid of CMY values is printed and then measured, using, for example, a Gretag SMP100 spectrophotometer with 2.degree. observer and D65 luminance. From a standpoint of accuracy and completeness, it is of course always preferable to measure as many CMY color combinations as possible, but for an eight-bit per component printer, (2.sup.8).sup.3 =16,777,216 possible such combinations exist. Accordingly, it is presently preferred to measure nine regularly spaced CMY values, resulting in 9.sup.3 =729 color combinations of CMY values. Together these points form a grided box in CMY space, each point being addressed by the printed CMY value and the content of each point being the measured L*a*b* value.
While it is possible to use the measured L*a*b* values as the forward mapping in step S601, it is preferable to apply a number of adjustments (such as white point, black point, gray axis, and smoothing adjustments) to the measured L*a*b* values before using them. Thus, in step S702, a white point adjustment is made to the measurements so that the white paper on which the CMY color combinations were printed ends up with an L*a*b* value of (100,0,0). Preferably, a Yon Kries white point adjustment is made to the measured value by first transforming the measurements to XYZ coordinate space using the illuminant, here, D65, as reference white and then transforming back to L*a*b* values using the measured value of paper as reference white.
In step S703, a black point adjustment is applied to the measured value so as to ensure that they cover the entire L* range, that is, from 0 to 100. The black point adjustment is made by designating the L*a*b* measured value for CMY=(100%,100%,100%) ink after white point adjustment. Then, for each other measured L*a*b* value, the following adjustment is made, where (1.sub.min, a.sub.min, b.sub.min) is the L*a*b* values for the black point: ##EQU1##
Thereafter, in step S704, a gray axis adjustment is made to the measurements. Specifically, most color printers use black as well as cyan, magenta and yellow, but it is not always necessary for printer driver 34 to produce a black color component. This is because the black color component is often generated by UCR or GCR performed by printer 30 itself. Most such printers typically use some variation of a 100% GCR replacement technique wherein equal amounts of CMY are replaced by K. Thus, so long as reproduction LUT 35 stores equal amounts of CMY in locations corresponding to a gray color, such a reproduction LUT would produce pleasing grays. However, since the reproduction LUT described so far depends on actual measured values, without gray axis adjustment it is not possible to guarantee that equal amounts of CMY are stored at locations corresponding to gray colors. Specifically, without gray scale adjustment, it is possible for a gray scale image to be produced in which some tones would appear neutral, some would have a slight reddish cast, some would have a slight greenish cast, etc., all of which would be disturbing to a viewer who expected a pure gray scale image.
Gray axis adjustment according to step S704 proceeds as follows. For each measured L*a*b* value, its corresponding CMY value is inspected, and if C=M=Y, then a* and b* are set to zero. Otherwise, a* and b* are left unchanged.
Step S705 applies smoothing so as to eliminate non-smooth characteristics of the forward mapping. Such non-smooth characteristics may be introduced from a variety of sources, such as measurement error, errors introduced by the halftone or dithering or GCR processes used by printer 30, and the above gray axis adjustments. Thus, step S705 smooths values with a smoothing technique which is described below and illustrated in FIG. 8.
Specifically, the smoothing technique is parameterized by two Values, an integer k and a real number z in which k is an integral divisor of n-1, wherein n is the size of the look-up table L after gray axis adjustment in step S704. In the presently preferred embodiment, n=9, meaning that there are nine values for each of C, M, and Y, all arranged in a three-dimensional grid as described above.
FIG. 8 shows a simplified situation in which, for look-up table L, designated at 60, n=5 although only one layer, such as Y=30%, is shown. Next, a new look-up table of L', designated at FIG. 8 at 61, is formed with size m.sup.3 by sub-sampling L, wherein ##EQU2##
A weighted average of L and L' is obtained so as to form smoothed look-up table L", designated at 62 in FIG. 8. Look-up table L" is of size n.sup.3. Weighting between L and L" is performed as follows. Let (c,m,y) be an entry in L, let (l,a,b).sub.L be its corresponding measurement, let (l,a,b).sub.L' be the result of interpolating (c,m,y) in L'. Then, the L" entry for (c,m,y) is given by: ##EQU3##
With the foregoing arrangement, with increasingly large k, the smoothness of L' increases but its accuracy decreases. With increasingly large z, more weight is given to the sub-sampled L' look up table in computing the smoothed L" look-up table. L" will be identical to L if z=0 or k=1.
Referring again to FIG. 7, in step S706, the smoothed look-up table L" is stored for the forward mapping look-up table used in connection with step S601.
FIG. 9 is a flow diagram for explaining dual cone gamut mapping by which step S605 obtains gamut-mapped CMY values for out-of-gamut L*a*b* colors. As explained more fully hereinbelow, gamut mapping shown in FIG. 9 proceeds by defining an "actual cone" of n-cone arranged surfaces which mimic the actual gamut of printer 30, defining an "ideal cone" of m-cone arranged surfaces which mimic an ideal gamut such as the gamut of monitor 23, projecting a target color in L*a*b* coordinates onto the actual cone and de-saturating the projected target color until an in-gamut CMY value is obtained, projecting the same target color onto the ideal cone and obtaining fully-saturated CMY values, and obtaining a weighted average of the in-gamut CMY and the fully-saturated CMY, weighting proceeding in accordance with the distance between the projection onto the actual cone and the target color.
Specifically, referring to FIG. 9, step S901 specifies defining an "actual cone" of N-cone arranged surfaces which mimic the actual gamut of printer 30. A suitable actual cone is shown in FIG. 10. As seen there, actual cone H.sub.A, which is designated by reference numeral 70, is defined by N chromatic vertices such as vertices 71, 72 . . . 76, in a case where N=6. The N upper surfaces of the N-cone are triangles formed by white point (100,0,0) and each of the chromatic vertices, while the N lower surfaces are triangles formed by black point (0,0,0) and each of the chromatic vertices.
The LCH values of each of chromatic vertices 71 through 76 are chosen by selecting N combinations of CMY and using the forward model obtained in step S601 to compute their LCH values. (The ensuing discussion freely mixes LCH values with L*a*b* values. As will be readily apparent to those skilled in the art, LCH values and L*a*b* values refer to the same physical CIELAB color space; it is simply that LCH values are cylindrical coordinate values whereas L*a*b* values are rectangular coordinate values.) For N=6, typical CMY combinations used are:
(C,M,Y).sub.0 =(100%,0%,0%)
(C,M,Y).sub.1 =(100%,100%,0%)
(C,M,Y).sub.2 =(0%,100%,0%)
(C,M,Y).sub.3 =(0%,100%,100%)
(C,M,Y).sub.4 =(0%,0%,100%)
(C,M,Y).sub.5 =(100%,0%,100%)
While the values specified above ensure that each of the chromatic vertices lies exactly on the gamut of printer 30, and that the resulting N-cone therefore closely mimics the actual gamut of printer 30, it is not necessary to use the values specified above. Specifically, in some instances, it is possible to specify other values for CMY, values which do not lie identically on the surface of the printer gamut, such as in situations where it is desired to even-out spacing of chromatic variations. In such cases, as specified in the appendix, CMY values of (80%,80%,0) and the like may be more appropriate. Appropriate values of CMY for use as chromatic vertices may readily be obtained by those skilled in the art based on the instant specification and the appendix hereto.
Reverting to FIG. 9, step S902 involves defining an "ideal cone" of M-cone arranged surfaces which mimic an ideal gamut such as the gamut of a typical RGB monitor. A suitable ideal cone is shown in FIG. 11. As seen there, ideal cone H.sub.I, designated generally with reference numeral 80, includes M chromatic vertices such as vertices 81 . . . 86, for M=6. The M upper surfaces of the M cone are triangles formed by white point (100,0,0) and each of the chromatic vertices. The lower M surfaces are triangles formed by black point (0,0,0) and each of the chromatic vertices.
Suitable values for the chromatic vertices for ideal M-cone 80 are obtained by setting M=6 and computing the vertices from a monitor's RGB primaries and secondaries. In this case, a standard CIE-defined RGB.fwdarw.CIEXYZ.fwdarw.LCH transformation for a typical RGB monitor is used. At each vertex of H.sub.I, a CMY combination is chosen that is subjectively deemed to be the best saturation match for the RGB color. For example, it might be decided that RGB fully-saturated blue is best reproduced by CMY=(100%,50%,0), or that RGB fully-saturated green is best reproduced by CMY=(60%,0,100%), and so forth. Other suitable values for CMY at each of the chromatic vertices of the ideal hex cone 80 are described in the appendix to this application. In this regard, it is noted that in some instances, M and N are best not equal to each other and that it may be desirable for M to be greater than N, such as M=8. In those instances, the additional chromatic vertices for ideal hex cone 80 enable the resulting gamut-mapped colors to reproduce highly saturated in-gamut colors like CMY=(100%,100%,0) or CMY=(100%,0,100%). Appropriate values of CMY for use as chromatic vertices may readily be obtained by those skilled in the art based on the instant specification and the appendix hereto.
For both actual cone 70 and ideal cone 80, and for each chromatic vertex in those two cones, two additional parameters are defined: an attenuation factor "a" and a slope factor .delta.. The attenuation factor "a" is used in connection with weighting between an in-gamut CMY value and a fully-saturated CMY value, and the slope factor .delta. is used in connection with projecting a target color onto the surface of each of the actual cone 70 and the ideal cone 80. Both of these processes are described more fully below.
Specifically, FIGS. 12(A), 12(B) and 12(C) are perspective, side and top views, respectively, showing projection of an out-of-gamut target color onto a surface of an N or an M cone. Specifically, as seen in these figures, an out-of-gamut target color 90 is projected to a point 91 on one surface 92 of the M or N cone-arranged surfaces. The surface onto which out-of-gamut target color 90 is projected is selected to be the surface with chromatic vertices which are clockwise-closest in hue to the hue of target color 90 and counterclockwise-closest in hue to the hue of target color 90. In FIGS. 12a through 12c, the clockwise-closest chromatic vertex is designated as 93 and the counterclockwise-closest vertex is designated as 94. The slope m.sub.P of projection line 95 from target color 90 to projection point 91 depends on the slope factor .delta. and on the difference between the luminance of target color 90 and the luminance of a point 96 (called the "crux") which lies at the same hue as that of the target color on a line 97 that lies between chromatic vertices 93 and 94. Thus, the slope m.sub.P is set as follows: ##EQU4## and where m.sub.S is the slope of the line 98 between the white or black origin and the crux (point 96), and where .delta. is the slope factor.
In practice, selecting a value for the slope factor .delta. depends on both the hue region in question and on the gamut of the particular printer 30 in question. The appendix to this specification gives suitable values for a wide variety of situations. Here, it will be noted that a value of .delta.=1 gives a slope for the projection line of m.sub.P =0 so that the projection line has a slope of zero regardless of the luminance of target color 90. On the other hand, a value of .delta.=0 gives a slope m.sub.P of the projection line which has a value of zero when the target luminance is equal to the crux luminance, and which has a value which makes the projection line 95 orthogonal to the surface as the target luminance approaches either zero or 100. This latter situation is depicted in FIG. 13 for plural out-of-gamut target colors (each of which are indicated by a "dot") at various luminance values.
It should be noted that the slope factor .delta. used in the above equation is an interpolated slope factor which depends upon the position of crux point 96 between chromatic vertex 93 and 94, and on the stored values of the slope factors for each chromatic vertex 93 and 94.
Having now completed a description of how to project an out-of-gamut target color 90 onto the surface of the actual cone or the ideal cone, attention is redirected to step S903 of FIG. 9 so as to complete description of how to apply dual cone gamut mapping so as to obtain a gamut-mapped printable CMY value for an out-of-gamut color. As specified at step S903, the out-of-gamut target color is projected onto actual cone 70 so as to obtain a projected point on a surface of actual cone 70. This is shown in side view in FIG. 14 in which out-of-gamut target color 101 is projected to a point P.sub.A on the surface of actual cone H.sub.A. Thereafter, as further shown in FIG. 14, and as specified at step S903, the projected point P.sub.A is de-saturated until an in-gamut CMY value, such as that shown at point r is obtained.
Then, in step S904, the same target color 101 is projected onto ideal cone H.sub.I to obtain a fully saturated CMY value, which is shown in FIG. 14 at 102. Finally, in step S905, a weighted average is obtained between the in-gamut CMY color specified at point r and the fully-saturated CMY color specified at point 102. The weight of the weighted average is determined so that for largely-saturated out-of-gamut target colors 101, fully saturated color 102 dominates, whereas for less-saturated out-of-gamut colors 101, (i.e., those colors for which target color 101 is close to H.sub.A), the in-gamut CMY value at point r dominates. In the present situation, the weight w of the weighted average is given by ##EQU5## where a.sub.I is the attenuation factor of point 96 (see FIG. 12) as interpolated between each of chromatic vertices 93 and 94, and .DELTA.E is the Euclidean distance in L*a*b* space between the target out-of-gamut color 101 and point P.sub.A on the surface of actual cone H.sub.A.
The so-calculated weight w is used to calculate the weighted average as follows: ##EQU6##
In calculating the above weight, w, it is preferred to use Euclidean distance in L*a*b* space as the measure of the distance between the target color and its projection onto actual cone H.sub.A. However, other measures of this distance may also be employed. For example, a so-called "city block" distance function can be employed, and it is possible also to employ measures of distance which emphasize one component (such as saturation) at the expense or even total exclusion of another component (such as luminance). In this latter regard, since in the preferred embodiment the target color is projected onto the actual cone H.sub.A at the same hue angle as the target color, the difference in hue between the target color and its projection onto H.sub.A is exactly zero; thus, in the preferred embodiment, the Euclidean distance between the target color and its projection onto H.sub.A depends only on differences in saturation and luminance.
Finally, so as to account for the possibility that the weighted CMY value has a component greater than 100%, the following equation is employed so as to limit each of the CMY values to 100%. ##EQU7##
By virtue of the arrangement described above, for out-of-gamut target colors far from the printer gamut (and consequently far from actual cone H.sub.A) the resulting CMY color is determined primarily by CMY values for ideal cone H.sub.I. As out-of-gamut target colors approach the actual cone H.sub.A, the direction of projection becomes more and more influenced by the actual cone. Thus, gamut mapping according to the technique described in FIG. 9 has the dual advantages that target colors close into the actual gamut of printer 30 are reproduced with a perceptual match, whereas target colors with large saturation are reproduced with a saturation match.
FIG. 15 is a flow diagram showing in more detail the smoothing process mentioned at step S607 in FIG. 6. Specifically, as shown in FIG. 15, to smooth reproduction LUT 35, a local relaxation is applied in step S1501, and a local shift is applied in step S1502. The local relaxation and shift are parameterized by the following values:
n=the neighborhood to use for relaxation
(L,a,b).sub.c =the center of the local relaxation area
r=a radius of the local relaxation area
w=the weight of the smoothing.
While the invention has been described with respect to an embodiment of a printer driver, it is to be understood that the invention can be embodied in any suitable color correction or printing system. For example, it is possible to embody the invention in a color management system, which is used in conjunction with a conventional printer driver, and which pre-processes color images in accordance with the invention and then feeds the processed color images to the conventional printer driver, thereby achieving the beneficial result of the invention. It is further possible to embody the invention in a color correction system which emulates color errors introduced by the printing process by processing an original color image in accordance with the invention and then displaying the processed color image on a color monitor, so as to emulate the effect of the color printing process. An operator can then compare the processed color image with what is desired, and based on the comparison can change the original color image so as to achieve some desired effect.
Thus, the invention has been described with respect to particular illustrative embodiments. It is to be understood that the invention is not limited to the above described embodiments and that various changes and modifications may be made by those of ordinary skill in the art without departing from the spirit and scope of the appended claims. ##SPC1##
Claims
What is claimed is:
1. A color printer driver for a printer, comprising: a reproduction look-up table (LUT) which, for each of plural colors both in-gamut and out-of-gamut of the printer, stores color component values for printing by the printer; and means, responsive to a command to print a color, for outputting color component values stored in said reproduction LUT in correspondence to the commanded color; wherein for close-in out-of-gamut colors, said reproduction LUT stores perceptually matched color component values, for far-out-of-gamut colors, said reproduction LUT stores saturation matched color component values, and for intermediate out-of-gamut colors, said reproduction LUT stores color component values corresponding to a smooth blending of perceptually matched color component values and saturation matched color component values.
2. An apparatus according to claim 1, wherein for in-gamut colors said reproduction LUT stores colorimetrically matched color component values.
3. An apparatus according to claim 1, wherein for intermediate out-of-gamut colors said reproduction LUT stores a weighted average of perceptually matched color component values and saturation matched color component values.
4. An apparatus according to claim 3, wherein the weighted average is weighted in accordance with a measure of distance of the out-of-gamut color from a gamut of the printer.
5. An apparatus according to claim 1, wherein the printer driver is stored on a computer readable media.
6. An apparatus for color printing comprising: a color printer for printing color images in response to color component values, said color printer having a gamut of printable colors; a color monitor for displaying colors within a gamut of displayable colors; a processing unit including a computer for executing stored program process steps, said processing unit including a color monitor interface for providing color primary values to said color monitor and a color printer interface for providing color component values to said color printer; and a memory for storing process steps for execution by said processing unit; said process steps including (a) steps which together comprise an application program for deriving a color image and for displaying and printing the derived color image, (b) steps which together comprise a monitor driver for providing color primary values to said color monitor interface in response to display of the color image by said application program, and (c) steps which together comprise a color printer driver for providing color component values to said color printer interface in response to printing by said application program, wherein for in-gamut colors colorimetrically matched color component values are provided, for close-in-out-of-gamut colors perceptually matched color component values are provided, for far-out-of-gamut colors saturation matched color component values are provided, and for intermediate out-of-gamut colors color component values corresponding to a smooth blending of perceptually matched color component values and saturation matched color component values are provided.
7. An apparatus according to claim 6, wherein said printer driver includes a reproduction look-up table (LUT) in which for in-gamut colors said reproduction LUT stores the colorimetrically matched color component values.
8. An apparatus according to claim 6, wherein said printer driver includes a reproduction look-up table (LUT) in which for intermediate out-of-gamut colors said reproduction LUT stores a weighted average of perceptually matched color component values and saturation matched color component values.
9. An apparatus according to claim 8, wherein the weighted average is weighted in accordance with a measure of distance of the out-of-gamut color from a gamut of the color printer.
10. An apparatus according to claim 6, wherein the color printer driver is stored on a computer readable media.
11. A color printer driver which outputs color component values for printing by a printer, said color printer driver comprising: a reproduction look-up table (LUT) which, for each of plural colors both in-gamut and out-of-gamut of the printer, stores color component values for printing by the printer; and means, responsive to a command to print a color, for outputting color component values stored in said reproduction LUT in correspondence to the commanded color; wherein for out-of-gamut colors said reproduction LUT stores color component values obtained by dual-cone gamut mapping in which an actual cone of N-cone arranged surfaces is defined which mimics an actual gamut of the printer, in which an ideal cone of M-cone arranged surfaces is defined which mimics an ideal gamut, in which a target out-of-gamut color is projected onto the actual cone and de-saturated until in-gamut color component values are obtained, in which the same target color is projected onto the ideal cone to obtain fully-saturated color component values, and in which a weighted average is obtained of the in-gamut color component values and the fully-saturated color component values, weight being based on a measure of distance between the out-of-gamut target color and its projection onto the actual cone.
12. An apparatus according to claim 11, wherein for in-gamut colors said reproduction LUT stores colorimetrically matched color component values.
13. An apparatus according to claim 11, wherein projection of the out-of-gamut target color is at a slope which varies, the variation in slope being based at least on a difference in luminance between the target color and a crux of the cone-arranged surfaces.
14. An apparatus according to claim 13, wherein projection of the out-of-gamut target color is also based on a slope factor which is varied based on chromatic vertices of the cone-arranged surfaces and hue of the target color.
15. An apparatus according to claim 11, wherein weighting is based on Euclidean distance between the out-of-gamut target color and the target color projected onto the actual cone.
16. An apparatus according to claim 15, wherein weighting is also based on an attenuation factor which is varied based on chromatic vertices of the cone-arranged surfaces and hue of the target color.
17. An apparatus according to claim 11, wherein the printer driver is stored on a computer readable media.
18. An apparatus for color printing comprising: a color printer for printing color images in response to color component values, said color printer having a gamut of printable colors; a color monitor for displaying colors within a gamut of displayable colors; a processing unit including a computer for executing stored program process steps, said processing unit including a color monitor interface for providing color primary values to said color monitor and a color printer interface for providing color component values to said color printer; a memory for storing process steps for execution by said processing unit; said process steps including (a) steps which together comprise an application program for deriving a color image and for displaying and printing the derived color image, (b) steps which together comprise a monitor driver for providing color primary values to said color monitor interface in response to display of the color image by said application program, and (c) steps which together comprise a color printer driver for providing color component values to said color printer interface in response to printing by said application program, said color printer driver including a reproduction look-up table (LUT) which, for each of plural colors both in-gamut and out-of-gamut of the printer, stores color component values for printing by said printer, wherein for out-of-gamut colors said reproduction LUT stores color component values obtained by dual-cone gamut mapping in which an actual cone of N-cone arranged surfaces is defined which mimics the actual gamut of the printer, in which an ideal cone of M-cone arranged surfaces is defined which mimics the gamut of the color monitor, in which a target out-of-gamut color is projected onto the actual cone and de-saturated until in-gamut color component values are obtained, in which the same target color is projected onto the ideal cone to obtain fully-saturated color component values, and in which a weighted average is obtained of the in-gamut color component values and the fully-saturated color component values, weight being based on a measure of distance between the out-of-gamut target color and its projection onto the actual cone.
19. An apparatus according to claim 18, wherein for in-gamut colors said reproduction LUT stores colorimetrically matched color component values.
20. An apparatus according to claim 18, wherein projection of the out-of-gamut target color is at a slope which varies, the variation in slope being based at least on a difference in luminance between the target color and a crux of the cone-arranged surfaces.
21. An apparatus according to claim 20, wherein projection of the out-of-gamut target color is also based on a slope factor which is varied based on chromatic vertices of the cone-arranged surfaces and hue of the target color.
22. An apparatus according to claim 18, wherein weighting is based on Euclidean distance between the out-of-gamut target color and the target color projected onto the actual cone.
23. An apparatus according to claim 18, wherein weighting is also based on an attenuation factor which is varied based on chromatic vertices of the cone-arranged surfaces and hue of the target color.
24. An apparatus according to claim 18, wherein the color printer driver is stored on a computer readable media.
25. A method for storing color component values into a reproduction look-up table (LUT), the color component values for printing by a color printer in response to commands to print both in-gamut and out-of-gamut colors, the method comprising the steps of: measuring colors printed by the printer in response to a suitably large number of various combinations of the color component values; constructing a forward mapping which maps color component values into printed colors; opening a blank reproduction LUT which, for each of plural different input colors both in-gamut and out-of-gamut, will map the input color to a printable color component value; testing whether each color in the blank reproduction LUT is in-gamut or out-of-gamut; for in-gamut colors storing in the blank-reproduction LUT a color component value derived from the forward mapping; for out-of-gamut colors applying a dual-cone gamut mapping so as to obtain a gamut-mapped color component value, and storing the gamut-mapped color component value into the blank reproduction LUT; and smoothing the reproduction LUT.
26. A method according to claim 25, wherein said step of applying a dual-cone gamut mapping so as to obtain a gamut-mapped color component value further comprises the steps of: defining an actual cone of N-cone arranged surfaces which mimic an actual gamut of the printer; defining an ideal cone of M-cone arranged surfaces which mimic an ideal gamut; projecting the out-of-gamut target color onto the actual cone and de-saturating until an in-gamut color component value is obtained; projecting the same target color onto the ideal cone so as to obtain a fully-saturated color component value; and obtaining a weighted average of the in-gamut color component value and the fully-saturated color component value, weighting being based on a measure of distance between the out-of-gamut target color and its projection onto the actual cone.
27. A method according to claim 25, wherein for in-gamut colors said reproduction LUT stores colorimetrically matched color component values.
28. A method according to claim 25, wherein projection of the out-of-gamut target color is at a slope which varies, the variation in slope being based at least on a difference in luminance between the target color and a crux of the cone-arranged surfaces.
29. A method according to claim 28, wherein projection of the out-of-gamut target color is also based on a slope factor which is varied based on chromatic vertices of the cone-arranged surfaces and on hue of the target color.
30. A method according to claim 25, wherein weighting is based on Euclidean distance between the out-of-gamut target color and the target color projected onto the actual cone.
31. A method according to claim 30, wherein weighting is also based on an attenuation factor which is varied based on chromatic vertices of the cone-arranged surfaces and on hue of the target color.
32. A method according to claim 25, wherein said step of constructing a forward mapping further comprises the steps of: applying a white point adjustment to the measured colors; applying a black point adjustment to the measured colors; applying a gray axis adjustment to the measured colors; and smoothing the measured colors.
33. A method according to claim 32, wherein said step of smoothing the measured colors comprises the steps of: sub-sampling the measured colors; and obtaining a weighted average between the measured colors and the sub-sampled colors.
34. A method according to claim 25, wherein said step of smoothing the reproduction LUT comprises the steps of: applying local relaxation to the reproduction LUT; and applying a local shift.
35. A dual-cone gamut mapping technique for mapping an out-of-gamut target color to color component values printable by a printer, comprising the steps of: defining an actual cone of N-cone arranged surfaces which mimic an actual gamut of the printer; defining an ideal cone of M-cone arranged surfaces which mimic an ideal gamut; projecting the out-of-gamut target color onto the actual cone and de-saturating until an in-gamut color component value is obtained; projecting the same target color onto the ideal cone so as to obtain a fully-saturated color component value; and obtaining a weighted average of the in-gamut color component value and the fully-saturated color component value, weighting being based on a measure of distance between the out-of-gamut target color and its projection onto the actual cone.
36. A method according to claim 35, wherein projection of the out-of-gamut target color is at a slope which varies, the variation in slope being based at least on a difference in luminance between the target color and a crux of the cone-arranged surfaces.
37. A method according to claim 36, wherein projection of the out-of-gamut target color is also based on a slope factor which is varied based on chromatic vertices of the cone-arranged surfaces and on hue of the target color.
38. A method according to claim 35, wherein weighting is based on Euclidean distance between the out-of-gamut target color and the target color projected onto the actual cone.
39. A method according to claim 38, wherein weighting is also based on an attenuation factor which is varied based on chromatic vertices of the cone-arranged surfaces and on hue of the target color.
Patent Citations (7)
Non-Patent Literature (2)
- M. Stone, et al., "Color Gamut Mapping and the Printing of Digital Color Images", ACM Transactions on Graphics, vol. 7, No. 4, Oct. 1988, pp. 249-292.
- K. Spaulding, et al., "UltraColor: A new gamut mapping strategy" SPIE vol. 2414, pp. 61-68.