Method of processing image signal

• US 6,278,533 B1
• Filed: 11/28/1997
• Issued: 08/21/2001
• Est. Priority Date: 11/29/1996
• Status: Expired due to Fees

• Current Assignee: Fujifilm Corporation (Fujifilm Holdings Corporation)
• Original Assignee: Fuji Photo Film Co Limited (Fujifilm Holdings Corporation)
• Inventors: Takemoto, Fumito
• Primary Examiner(s): Au, Amelia M.
• Assistant Examiner(s): Dastouri, Mehrdad
• Application Number:
  US08/980,367
• Time in Patent Office: 1,362 Days
• Field of Search: 382/162, 382/167, 358/515, 358/518, 358/520, 358/523, 358/534, 358/521
• US Class Current: 358/521
• CPC Class Codes: H04N 1/6022 Generating a fourth subtractive colour signal, e.g. under colour removal, black masking
• CPC Class Codes: H04N 1/6033 using test pattern analysis H04N1/6055 takes precedence
• CPC Class Codes: H04N 1/6058 Reduction of colour to a range of reproducible colours, e.g. to ink- reproducible colour gamut

Abstract

R, G, B signals outputted from a digital camera or a color scanner are converted into colorimetric signals by a colorimetric converter having a colorimetric conversion matrix and a colorimetric conversion table. Colorimetric signals are processed for setup by a colorimetric setup unit. Colorimetric signals processed for setup are converted by a cmyk converter into c, m, y, k signals that are half tone dot % signals for being supplied to an image output unit such as a film producing device, a color printer, or the like.

Claims

1. A method of converting arbitrary color signals in a first color space into color signals in a second color space, comprising:

reading a color chart having a plurality of color patches representing typical colors with different chromas, lightnesses, and hues with a device which is to output color signals in the first color space;

determining first-order and second order terms of the read color signals in the first color space as a group of first explanatory variables;

subjecting the group of first explanatory variables to a principal component analysis for conversion into a group of second explanatory variables which cross each other;

using colorimetric values of the color patches of the color chart as a group of criterion variables, and subjecting the group of criterion variables and the group of second explanatory variables to a multiple regression analysis to determine a matrix of partial regression coefficients; and

subjecting arbitrary color signals in the first color space to the matrix of partial regression coefficients for conversion into color signals in the second color space.

2. A method according to claim 1, further comprising:

after the matrix of partial regression coefficients is determined, selecting explanatory variables at a predetermined significance level among the group of second explanatory variables, and establishing selected explanatory variables as usable explanatory variables.

3. A method according to claim 1, wherein when the matrix of partial regression coefficients is determined, a selected color of said color chart is weighted to increase the accuracy with which to convert said selected color.

4. A method according to claim 3, wherein said selected color comprises a default color of gray.

5. A method according to claim 3, wherein said selected color comprises a color determined by statistically processing color signals in the first color space of an original scene.

6. A method according to claim 5, wherein the color signals in the first color space are statistically processed by extracting a color of skin, a color of blue, or a color of green from the first color space of the original scene.

7. A method according to claim 5, wherein the color signals in the first color space are statistically processed by dividing the first color space into regions around the colors of the patches of said color chart, and determining the frequencies of pixels, present in the divided regions, of the color signals in the first color space of the original scene, and wherein said particular color is weighted depending on the frequencies.

8. A method according to claim 1, wherein said first color space comprises an RGB color space, and said second color space comprises a device-independent colorimetric color space.

9. A method according to claim 1, wherein said first color space comprises an RGB color space, and said second color space comprises a device-independent colorimetric color space, and wherein the first color space is divided into regions around the colors of the patches of the color chart, R, G, B signals in the first color space of an original scene are converted into colorimetric signals by said matrix of partial regression coefficients which are not weighted, the frequencies of pixels, present in the regions, of the colorimetric signals are determined, and a particular color of the color chart is weighted to increase the accuracy with which to convert said particular color.

10. A method according to claim 1, wherein said color chart comprises a Macbeth color checker.

11. A method of processing an image signal, comprising:

converting device-dependent image signals produced from a non-reversal film subject into colorimetric signals of an original;

converting said colorimetric signals into R, G, B signals in a RGB color system and thereafter logarithmically converting ratios of said R, G, B signals to a basic stimulus into colorimetric equivalent-neutral-density signals;

generating rough data of said colorimetric equivalent-neutral-density signals;

setting up automatically a gradation conversion curve based on said rough data;

converting gradations of said colorimetric equivalent-neutral-density signals based on said gradation conversion curve;

inversely logarithmically converting the colorimetric equivalent-neutral-density signals whose gradations have been converted into R, G, B signals; and

converting the R, G, B signals into colorimetric signals;

said RGB color system comprising a range including reference color stimuli and containing a color reproduction range of a reversal film, or

a range including reference color stimuli and containing a color reproduction range of said non-reversal film subject.

12. A method of processing an image signal, comprising:

converting device-dependent image signals produced from an original scene into a plurality of colorimetric signals of said original scene;

converting said colorimetric signals into colorimetric equivalent-neutral-density signals;

generating rough data of said colorimetric equivalent-neutral-density signals;

setting up automatically a gradation conversion curve based on said rough data;

converting gradations of said colorimetric equivalent-neutral-density signals based on said gradation conversion curve;

converting the colorimetric equivalent-neutral-density signals whose gradations have been converted into a plurality of colorimetric signals; and

converting said converted colorimetric signals into c, m, y, k signals whose colors are colorimetrically saved.

14. A method according to claim 12, wherein said device-dependent image signals comprise R, G, B signals generated by a digital camera.

13. A method of processing an image signal, comprising:

converting device-dependent image signals produced from an original scene into a plurality of colorimetric signals of said original scene;

changing said colorimetric signals into colorimetric signals under a light source different from a light source under which said device-dependent image signals have been produced;

converting said colorimetric signals under the different light source into colorimetric equivalent-neutral-density signals;

generating rough data of said colorimetric equivalent-neutral-density signals;

setting up automatically a gradation conversion curve based on said rough data;

converting gradations of said colorimetric equivalent-neutral-density signals based on said gradation conversion curve;

converting the colorimetric equivalent-neutral-density signals whose gradations have been converted into a plurality of colorimetric signals; and

converting said converted colorimetric signals into c, m, y, k signals whose colors are colorimetrically saved.

15. A method according to claim 13, wherein said device-dependent image signals comprise R, G, B signals generated by a digital camera.

16. A method of processing an image signal, comprising:

converting device-dependent image signals produced from an original scene into colorimetric signals of said original scene;

converting said colorimetric signals into colorimetric equivalent-neutral-density signals;

generating rough data of said colorimetric equivalent-neutral-density signals;

setting up automatically a gradation conversion curve based on said rough data;

converting said colorimetric signals into dye density signals on a reversal subject;

processing said dye density signals under standard image processing conditions including a gradation conversion process, a color correction process, an under color removal process, and a K-plate generating process for conversion into c, m, y, k signals.

18. A method according to claim 16, wherein said step of processing said dye density signals under standard image processing conditions comprises processing a regular array of color patches of colorimetric values or signals convertible into the colorimetric values on the reversal subject under said standard image processing conditions to generate a cmyk image, generating a data set representing an association between patches reproduced on said cmyk image and said colorimetric values or said signals convertible into the colorimetric values, generating a standard image processing condition table from said data set, and converting said dye density signals into the c, m, y, k signals according to said standard image processing condition table and an interpolation process.

17. A method of processing an image signal, comprising:

converting image signals of a reversal subject read by an image reader into a plurality of colorimetric signals on said reversal;

converting said colorimetric signals into colorimetric equivalent-neutral-density signals;

generating rough data of said colorimetric equivalent-neutral-density signals;

setting up automatically a gradation conversion curve based on said rough data;

converting gradations of said colorimetric equivalent neutral density signals based on said gradation conversion curve;

inversely logarithmically converting the colorimetric equivalent neutral density signals whose gradations have been converted into dye density signals; and

processing said dye density signals under standard image processing conditions of said image reader including a gradation conversion process, a color correction process, an under color removal process, and a K-plate generating process for conversion into c, m, y, k signals.

19. A method according to claim 17, wherein said step of processing said dye density signals under standard image processing conditions comprises processing a regular array of color patches of colorimetric values or signals convertible into the colorimetric values on the reversal subject under said standard image processing conditions to generate a cmyk image, generating a data set representing an association between patches reproduced on said cmyk image and said colorimetric values or said signals convertible into the colorimetric values, generating a standard image processing condition table from said data set, and converting said dye density signals into the c, m, y, k signals according to said standard image processing condition table and an interpolation process.

20. A method of converting arbitrary color signals in a first color space into color signals in a second color space, comprising:

reading a color chart having a plurality of color patches representing typical colors with different chromas, lightnesses, and hues with a device which is to output color signals in the first color space;

determining first order and second order terms of the read color signals in the first color space as a group of first explanatory variables;

subjecting the group of first explanatory variables to a principal component analysis for conversion into a group of second explanatory variables, wherein each of said second explanatory variables are not correlated to each other to avoid a multicollineation effect during multiple regression analysis;

using colorimetric values of the color patches of the color chart as a group of criterion variables, and subjecting the group of criterion variables and the group of second explanatory variables to a multiple regression analysis to determine a matrix of partial regression coefficients; and

subjecting arbitrary color signals in the first color space to the matrix of partial regression coefficients for conversion into color signals in the second color space.

21. A method according to claim 20, wherein when the matrix of partial regression coefficients is determined, a selected color of said color chart is weighted to increase the accuracy with which to convert the selected color.

22. A method according to claim 21, wherein said selected color comprises a color determined by statistically processing color signals in the first color space of an original scene.

23. A method according to claim 22, wherein the color signals in the first color space are statistically processed by dividing the first color space into regions around the colors of the patches of said color chart, and determining the frequencies of pixels, present in the divided regions, of the color signals in the first color space of the original scene, and wherein said particular color is weighted depending on the frequencies.