JP2005210551A - Image processing algorithm - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize color reproduction that a user considers preferable.
In a printing apparatus having a color conversion table that maps to a color reproduction space, the color conversion table passes through a host computer that drives the printing apparatus connected to the printing apparatus {R, G, B } = {0, 255, 0}, {R, G, B} = {0, k, 0} () with respect to the angle θ formed by the colorimetric value in the printer color reproduction space obtained by printing the signal In the case of k ≦ C, 32 ≦ C ≦ 112), closer to cyan than θ, {R, G, B} = {0, k, 0} (C <k ≦ 255, 32 ≦ C ≦ 112) In some cases, the mapping is determined so as to be biased toward yellow rather than θ.
[Selection] Figure 7

Description

  The present invention relates to mapping between a printer color gamut and a monitor color gamut, so-called color correction.

  When outputting a color image from an image output unit based on a three-dimensional color signal input from an image input unit in an image output apparatus such as an ink jet printer, it is necessary to maintain good color reproducibility. As a technique for this purpose, there is a so-called color gamut mapping technique in which a range where color reproduction cannot be performed by the image output unit is compressed and mapped to a range where color reproduction is possible.

  As this color gamut mapping technique, when performing color gamut mapping (color gamut conversion) by performing color conversion using a conventional three-dimensional lookup table (LUT), an output color space (color reproduction conversion system and output) The space in which the color signal is formed with the system) is divided into the area where the color gamut mapping process is performed and the area where the color gamut mapping process is not performed, and the input color space (the color signal is formed between the input system and the color reproduction conversion system) There is a technique for determining a clipping region by mapping to a space to perform color gamut mapping (for example, Japanese Patent Laid-Open No. 5-284347).

  When mapping is performed using this method and a method similar to this method, the mapping is generally faithful to the input R, G, and B signals. For example, in R, G, B signals constituting electronic data obtained from a digital image input device such as a digital camera or a scanner, {R, G, B} = {i, j, i} (0 ≦ i ≦ 255, 0 ≦ j ≦ 255 (always j ≧ i) indicates all green colors that can be expressed regardless of brightness and saturation. In order to faithfully print a plane extending in the sR, G, and B spaces. When the color of the printed material is measured, it becomes a plane of the same hue in the printer color reproduction space.

  By the way, with the recent development of printing technology, the demand for printing apparatuses that print not only faithful to input signals but also so-called memory colors that emphasize psychological factors such as impression / expectation before printing has increased. ing. This is a color correction for the purpose of making the printing result equal to the color displayed on the display, while the mapping faithful to the input signal is compared with the printing of the memory color, the user feels that the printed matter itself is preferable. This is a color correction method.

  However, in color correction faithful to the input signal using the above technology, the memory color often depends on psychological factors such as the user's expectation and impression, so that there is a problem that color reproduction that the user feels is not easily performed. there were.

  The present invention has been made in view of the above problems, and an object thereof is to obtain a user's favorite color reproduction that cannot be achieved by color correction faithful to an input signal.

  The means for solving / realizing the problem concerning the present invention is to display the input green {R, G, B} = {i, j, i} (0 ≦ i ≦ 255, 0 ≦ j ≦ 255, but always When a signal j ≧ i) is input, the surface where the colorimetric value obtained by measuring the print result in the printer color reproduction space is not a plane faithful to the input signal, {R, G , B} = {0, 255, 0}, {R, G, B} = {0, k, 0 with respect to the angle θ formed by the colorimetric value in the printer color reproduction space obtained by printing the signal } (K ≦ C, 32 ≦ C ≦ 112), closer to cyan than θ, {R, G, B} = {0, k, 0} (C <k ≦ 255, 32 ≦ C ≦ 112) In the case of), LUTs are created that are mapped in a biased manner toward yellow rather than θ.

  According to the present invention, it is possible to obtain high support when printing a natural image containing a lot of green using, for example, an ink jet printer using four color materials of CMYK.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings. In the following description, a CIE L * a * b * uniform color space will be described as an example.

  FIG. 1 is a diagram schematically showing a configuration of an image processing system according to a preferred embodiment of the present invention. This image processing system includes a digital image input device 100 such as a digital steel camera or a scanner, a computer 110 such as a personal computer, an image display device 170 such as a display, and a printing apparatus 150.

  In the computer 10, a software program for decoding and displaying data obtained from the image input device 100, an application 120 such as a document processing, a spreadsheet, an Internet browser, and the application 120 are issued to the operating system 130. A printer driver 140 that processes various drawing command groups (image drawing command, text drawing command, graphics drawing command, etc.) related to the print command to be generated and generates print data, and various command groups issued by the application 120 to the operating system And a software program such as a monitor driver 160 for displaying the image on the image display device 170 is incorporated.

  Although not explicitly shown in the drawing, the computer 110 includes various hardware (operational processors such as a CPU or MPU, hardware constituting a known computer such as RAM, HDD, and ROM) capable of operating these software. doing.

  As a specific configuration example of the image processing system shown in FIG. 1, a personal computer of an IBM compatible computer of IBM, which is generally popular, a display and a printing device are connected to the personal computer, and the personal computer In addition, Windows 95 (registered trademark) 95, 98, me, NT, 2000, XP, etc. of Microsoft Corporation are installed as an operating system, and application software having a printing function is installed.

  In the computer 110, document data classified into text such as characters by the application 120 based on the display image displayed on the monitor, graphics data classified into graphics such as graphics, image data classified into natural images, etc. Output image data is generated. When printing an image based on the output image data, the application 120 issues a print output request to the operating system 130 and issues a drawing command group corresponding to the type of output image data to the operating system 130.

  In response to the output request of the application 120, the operating system 130 issues a drawing command group to the printer driver 140 corresponding to the output printer. The printer driver 140 processes a print request input from the operating system 130, creates print data that can be printed by the printing apparatus 150, and transmits the print data to the printing apparatus 150. When the printing apparatus 150 is a raster printer, the printer driver 140 sequentially performs image processing in response to a drawing command from the operating system 130, and the printing apparatus 150 can print the contents of the R, G, B 24-bit page memory. The data is converted into a data format (for example, CMYK data) and transmitted to the printing apparatus 150.

  FIG. 2 is a diagram illustrating the configuration of the printer driver 140. The image correction processing unit 141 performs image correction processing on the color information included in the drawing command group from the operating system 130. In this image correction process, R, G, B color information is converted into luminance / color difference signals, exposure correction is performed on the luminance signals, and the corrected luminance / color difference signals are inversely converted into R, G, B color information. .

  The printing device correction processing unit 142 rasterizes a drawing / drawing command based on the R, G, B color information subjected to the image correction process, generates a raster image on the R, G, B 24-bit page memory, and performs a color reproduction space mapping process. Then, color separation into CMYK and gradation correction processing are performed, and CMYK data depending on the color reproducibility of the printer is generated for each pixel and transmitted to the printing apparatus 150.

  FIG. 3 is a diagram illustrating the configuration of the correction processing unit 142 for the printing apparatus. Hereinafter, the processing in the printing correction processing unit 142 will be described with reference to FIG. In the following description, a color space usually expressed in three dimensions will be described with reference to a diagram (for example, FIG. 4) schematically expressed in two dimensions for convenience of explanation.

  First, 8-bit image data of each of R, G, and B, whose brightness, gradation, and color tone are adjusted by the image correction processing unit 141 (see FIG. 2), is sent to the image signal input unit 200 of the correction processing unit 142 for the printing apparatus. Entered. The R, G, B 8-bit image data input to the image signal input unit 200 corresponds to colors reproduced on the image display device 170 (see FIG. 1), and is, for example, a uniform color space. In the coordinate values of the L * a * b * color system, colors of (L_display, a_display, b_display) are represented.

  However, as can be seen from FIG. 4, the color reproduction space of the monitor and the color reproduction space of the printer are different in size on a uniform color space such as an L * a * b * space. That is, even if the image data expressed on the monitor is output to the printer via the post-stage color signal conversion unit 220 (R, G, B-CMYK conversion) and the gradation correction unit 230 in FIG. (Area on the color space that cannot be expressed on the printer). Therefore, on the printer side, the pseudo color (L * a * b * value), that is, the monitor's reproducible color area (hatched area in FIG. 4), which is reproducible and not reproducible by the printer. It is necessary to generate and print a pseudo color different from color development.

  In the following, the L * a * b * space is assumed to be a uniform color space, and cylindrical coordinates H (= atan (b / a)), S (= (a * ^ 2 + b * ^ 2) ^ 1/2), V (= A description will be given of an example in which gamut compression is performed using a predetermined compression formula for one S for the entire space on the HSV space copied to (L *).

  The pre-stage color signal conversion unit in FIG. 3-210 compresses the monitor color reproduction space (monitor gamut) for the area in the color space that cannot be reproduced by the printer, and associates it with points in the printer color reproduction space. That is, the input R, G, B8-bit image signal processed by the preceding color signal conversion unit in FIG. 3-210 is converted into R′G′B′8-bit data corresponding to a point on the color reproduction space of the printer.

  Specifically, in general, the pre-stage color signal conversion process performs conversion from the monitor R, G, B color space to the printer R′G′B ′ color space according to the following procedure.

  The color reproduction space represents the color signal intensity of three axes R, G, B (red, green, blue), and the value that each color signal can take is when the intensity data of the color signal is an 8-bit configuration. , “0”, “1”, “2”,..., “255”. Accordingly, the colors that can be expressed using the 8-bit R, G, B color signal intensities are 16777216 colors, but it is not realistic to obtain L * a * b * values calorimetrically for all these colors. Therefore, in this embodiment, the color signal intensities of R, G, and B are divided into nine as follows, and processing is performed, and R, G, and B values located between the following grids are interpolated and used. ing. In the present embodiment, a cubic interpolation method is used as a method for obtaining a corresponding L * a * b * value from the R, G, and B values.

  Image data (patch) as shown in FIG. 5 is output by each device, and the color reproduction space (gamut) by each device is associated with the coordinate point of the uniform color space, respectively in the R, G, B space. Expressed by a three-dimensional lattice.

  Accordingly, the coordinates of the grid points are considered such that the signal values of the respective colors can be represented by combinations of “0”, “16”, “32”,..., “224” and “225”. As a result, the number of grid points is 93 = 729. Here, the color signal intensity is measured using, for example, a color measuring device such as SPM100-II manufactured by GRETAG, and the absolute colors of the input and output color signals are compared by obtaining, for example, L * a * b * values. It becomes possible.

  Instead of using the colorimetric value, an output result prediction method described in Japanese Laid-Open Patent Publication No. 5-287347 may be used. CMYK values can be obtained directly from L * a * b * values by using an approximate expression applying a neural network for prediction. In this embodiment, any R, G, B can be obtained from CMYK values at 729 lattice points. When the value is obtained, it is necessary to interpolate. However, if the approximate expression is used, the interpolation process can be omitted and the efficiency can be improved. Various methods are known as the prediction method, and an appropriate method may be used in consideration of the required element accuracy.

  The pre-stage color signal conversion unit in FIG. 3-210 holds, for example, lightness L * so that the gamuts by the monitors R, G, and B are inside the printer R′B′G ′ gamut in the L * a * b * space. The input device gamut is mapped to the output device gamut by reducing the saturation S (= (a ** a * + b ** b *) ^ 1/2) or the like as it is (FIG. 6). That is, by this operation, a set of L * a * b * values in the printer gamut corresponding to the monitor R, G, and B values is obtained.

  When the monitor gamut after compression comes to fit in the printer gamut, for example, ΔE (= ((L * ′ − L *) ^ 2+ (a * ′ − a *) ^ 2+ (b * ′ − b *) ^ 2)) Determine the set of monitor (R, G, B) and printer (R ′, G ′, B ′) using L * a * b * value as a key so that ^ 1/2 is minimized. Thus, the printer R′G′B ′ values corresponding to the monitor R, G, and B values can be obtained.

  By the preceding color conversion processing, the monitor color signal (R, G, B) and the printer color signal (R ′, G) are reproduced with faithful color reproduction using ΔE as an index in an area common to both the monitor gamut and the printer gamut. It is possible to determine the set of ', B').

  In the conventional embodiment, a set of the monitor (R, G, B) and printer (R ′, G ′, B ′) is stored in a memory as a previous stage color signal conversion table, and sent to the driver. By performing conversion on the image signal, a process equivalent to the color gamut conversion process is performed on each image signal.

  FIG. 7 shows the colorimetric values in the L * a * b * uniform color space obtained by printing the signal {R, G, B} = {0, 255, 0}, which is the greatest feature of the present invention. When {R, G, B} = {0, k, 0} (k ≦ C, 32 ≦ C ≦ 112) with respect to angle θ, {R, G, B} is closer to cyan than θ. = {0, k, 0} (C <k ≦ 255, 32 ≦ C ≦ 112), printing / colorimetry of FIG. 6 using LUTs that are offset and mapped to yellow rather than θ Is two-dimensionally expressed by projecting onto the a * -b * plane in the L * a * b * uniform color space.

  This preferred hue plane is obtained as follows. First, a general test method for realizing a preferable color reproduction, a monitor test consisting of a large number of randomly extracted subjects was conducted to determine whether a large number of support could be obtained.

  For this monitor test, the preceding color signal conversion table obtained as described above and the conversion table are configured as appropriate means (color correction processing using software / hardware and the preceding color signal conversion table are configured. Create a number of tables in which the colorimetric results are different by a few degrees for (a *, b *) in L * a * b * uniform color space coordinates. A print of image data such as a natural image containing many signals in the vicinity of the green was prepared.

  Specifically, as shown in FIG. 8, the output result of the printing apparatus 150 is evaluated through the colorimetric device 240, the previous color signal change table is corrected based on the evaluation result, and the corrected table is displayed in the previous stage of FIG. A print sample was prepared by replacing the table used in the color signal changing section.

  Specifically, for the same image, the signal {R, G, B} = {i, j, i} (0 ≦ i ≦ 255, 0 ≦ j ≦ 255, but always j ≧ i) is faithfully reproduced. LUT (A) and the angle θ formed by the colorimetric values in the L * a * b * uniform color space obtained by printing the signal {R, G, B} = {0, 255, 0} , LUT (B) in which the signal is uniformly biased to the cyan side, LUT (C) in which the signal is uniformly biased to the yellow side, and {R, G, B} = {0 , K, 0} (in the case of k ≦ C, 32 ≦ C ≦ 112, more yellow than θ, {R, G, B} = {0, k, 0} (C <k ≦ 255, 32 ≦ In the case of C ≦ 112), the LUT (D) mapped to be biased toward cyan rather than θ and the signal {R, G, B} = {0, k, 0} (k ≦ C, 32) ≦ C ≦ 1 In the case of 2), it is closer to cyan than θ, and in the case of {R, G, B} = {0, k, 0} (C <k ≦ 255, 32 ≦ C ≦ 112), it is closer to yellow than θ. A plurality of prints were made using LUTs (E) that were mapped to each other.

  Furthermore, the monitor test subjects ranked the printed samples from 1st to 5th.

  Here, from the colorimetric values obtained by printing the image data (FIG. 5) with the LUT (A) described above, the green hue plane stretched in the L * a * b * uniform color space is represented by a in the coordinate system. When projected onto the * -b * plane, it has a linear shape that faithfully reflects the input signal as shown in FIG.

  As a result of performing a monitor test using a plurality of print samples prepared as described above, it was possible to obtain favorable support from the user using the LUT (E) (FIG. 10).

  At this time, it is also possible to use a LUT in which the θ is mapped so as to be 150 degrees when projected onto the a * -b * plane in the L * a * b * uniform color space as shown in FIG. , Suitable support was obtained from the user.

  With respect to the angle θ formed by the colorimetric values in the L * a * b * uniform color space obtained by printing the signal {R, G, B} = {0, 255, 0} in this way, { When R, G, B} = {0, k, 0} (k ≦ C, 32 ≦ C ≦ 112), closer to cyan than θ, {R, G, B} = {0, k, 0 } (C <k ≦ 255, 32 ≦ C ≦ 112), it was possible to obtain a high level of support from the user by using ones that were biased and mapped closer to yellow than θ.

  Further, the angle θ formed by the colorimetric values in the L * a * b * uniform color space obtained by printing the signal {R, G, B} = {0, 255, 0} is equal to L * a * b *. When projected onto the a * -b * plane in the color space, the measured color value is 140 degrees ≦ ξ ≦ 160 degrees when arctangent (b * / a *) = ξ in the printer color reproduction space. Also, by using the LUT mapped in such a way, a favorable support can be obtained from the user.

  In this example, the case where four color materials of CMYK are used is taken as an example. However, the same result can be obtained by using a printing device using green, red, blue and light inks having different dilution rates, so-called light inks. Needless to say.

  Furthermore, in this example, the L * a * b * uniform color space has been described as an example, but the same effect can be obtained even if another color space coordinate system is used.

1 is a diagram showing a schematic configuration of an image processing system according to a preferred embodiment of the present invention. FIG. 6 is a diagram illustrating processing performed by a printer driver. FIG. 3 is a schematic diagram illustrating color processing performed by a printer driver. It is a figure which represents typically the comparison with the color reproduction range of a display, and the color reproduction range of a printer. Chart used when determining the gamut of a printer / display, etc. (By outputting this and measuring the color, it is possible to associate R, G, B color signal intensities with L * a * b * values. ). It is a figure which illustrates notionally the gamut compression to a saturation direction in a (lightness-saturation) plane. FIG. 6 was printed / colorimetrically projected onto the a * -b * plane in the L * a * b * uniform color space using the LUT set to collect high support realized by the present invention. Is. FIG. 4 is a schematic diagram illustrating color processing performed by a printer driver, realized by the present invention. FIG. 6 is a graph obtained by printing / colorimetrically printing FIG. 6 using the LUT realized by the conventional technique and projecting it on the a * -b * plane in the L * a * b * uniform color space. It is the result of the monitor test. This is a preferable color reproduction realized by the present invention.

Claims (2)

In a printing apparatus having a color conversion table that maps to a color reproduction space,
The color conversion table is a printer color reproduction obtained by printing a signal {R, G, B} = {0, 255, 0} via a host computer that drives the printing apparatus connected to the printing apparatus. With respect to the angle θ formed by the colorimetric value in the space, if {R, G, B} = {0, k, 0} (k ≦ C, 32 ≦ C ≦ 112), the angle is closer to cyan than θ. , {R, G, B} = {0, k, 0} (C <k ≦ 255, 32 ≦ C ≦ 112) A printing apparatus. In a printing apparatus having a color conversion table that maps to a color reproduction space,
The color conversion table is obtained when a signal {R, G, B} = {0, 255, 0} is input to the printing apparatus through a host computer that drives the printing apparatus connected to the printing apparatus. In addition, when the colorimetric value is arctangent (b * / a *) = ξ in the printer color reproduction space, it is determined to perform mapping with such a characteristic that 140 degrees ≦ ξ ≦ 160 degrees. A printing apparatus characterized by that.

Images