Color image reproduction system having color analysis function performed with a neural network system

• US 5,506,696 A
• Filed: 08/10/1994
• Issued: 04/09/1996
• Est. Priority Date: 08/10/1994
• Status: Expired due to Fees

• Current Assignee: TOYO Ink Manufacturing Company Limited (artience Co. Ltd.)
• Original Assignee: TOYO Ink Manufacturing Company Limited (artience Co. Ltd.)
• Inventors: Nakano, Yoshiki
• Primary Examiner(s): Rogers, Scott A.
• Application Number:
  US08/288,494
• Time in Patent Office: 608 Days
• Field of Search: 358/504, 358/518, 358/519, 358/520, 358/523, 358/515, 382/167, 382/162, 382/156, 395/21
• US Class Current: 358/504
• CPC Class Codes: H04N 1/6055 using test pattern analysis

Abstract

A color separation section for converting colorimetric values into color separation values adopts a neural network. When a color image output device for outputting a color image on the basis of color separation value signals is used, the color image output device to be used outputs a standard color sample having known color separation values. The color sample is colorimetrically measured by a colorimetry device to obtain colorimetric values of the color sample. The neural network executes learning to have conversion characteristics for converting the colorimetric values into corresponding color separation values. An object to be reproduced which has a required color is colorimetrically measured by the colorimetry device. Colorimetric values obtained by the colorimetry processing are converted into color separation values using the neural network of the color separation section. The color image output device outputs a target color based on the converted color separation values.

Claims

1. A color image reproduction system which comprises colorimetry means for generating a colorimetric value signal by colorimetrically measuring a color sample, and image output means for outputting a color image corresponding to a color separation value signal and for outputting a standard color sample in correspondence with predetermined color separation value signals, comprising:

conversion means, constituted by a neural network, for converting the colorimetric value signal generated by said colorimetry means into a color separation value signal in accordance with setting of a parameter;

learning control means for controlling said neural network of said conversion means so that colorimetric value signals of the standard color sample received from said colorimetry means are converted into the predetermined color separation value signals in order to adjust the parameter of said neural network corresponding to said image output means; and

image reproduction control means for converting a colorimetric value signal of an object to be reproduced, which signal is generated by said colorimetry means, into a color separation value signal in response to said conversion means, and supplying the converted color separation value signal to said image output means to output a color image of the object to be reproduced, said color image reproduction system comprising a plurality of image output means of different types.

2. A system according to claim 1, wherein the colorimetric value signal is a signal representing L*a*b* values defined in a CIE L*a*b* space.

3. A system according to claim 1, wherein the colorimetric value signal is a signal representing XYZ values defined in an XYZ space.

4. A system according to claim 1, wherein the colorimetric value signal is a signal representing Yxy values defined in a Yxy space.

5. A system according to claim 1, wherein the colorimetric value signal is a signal representing L*u*v* values defined in a L*u*v* space.

6. A system according to claim 1, wherein the colorimetric value signal is a signal representing HVC Munsell values defined in a HVC Munsell space.

7. A system according to claim 1, wherein the color separation value signal is a signal of CMY values representing ratios of cyan (C), magenta (M), and yellow (Y).

8. A system according to claim 1, wherein the color separation value signal is an RGB value signal representing ratios of red (R), green (G), and blue (B).

9. A system according to claim 1, wherein said neural network of said conversion means has a layered structure equal to or more than three-layered, and each layers of said neural network include units equal to or more than three units.

10. A system according to claim 1, wherein each of the parameters includes a type of the image output device, the layer structure of said neural network, the number of units of each layer in the layer structure, and a weight associated with each unit.

11. A color image reproduction system comprising:

colorimetry means for generating a colorimetric value signal by colorimetrically measuring a color sample;

a plurality of image output means of different types, each having a function of outputting a color image corresponding to a color separation value signal and a function of outputting a standard color sample in correspondence with predetermined color separation value signals;

conversion means, constituted by a neural network, for converting the colorimetric value signal generated by said colorimetry means into a color separation value signal in accordance with setting of a parameter;

storage means for storing the parameter and the predetermined color separation value signals;

input means for inputting an instruction of the image output means to be used in color image reproduction processing of said plurality of image output means; and

a processor for controlling said color image reproduction system,

said processor having a function of executing processing for performing, for each of said plurality of image output means, learning of said neural network of said conversion means so as to convert the colorimetric value signals of the standard color sample which are colorimetrically measured by said colorimetry means into color separation value signals stored in said storage means, and an image reproduction function of reading out the parameter corresponding to the image output means designated by said input means from said storage means, setting the readout parameter in said conversion means, inputting a colorimetric value signal of an object to be reproduced, which signal is generated by said colorimetry means, to said conversion means, and outputting a color separation value signal converted by said conversion means to the image output means designated by said input means to output a color image corresponding to the colorimetric value signal of the object to be reproduced.

12. A system according to claim 11, wherein the colorimetric value signal is a signal representing L*a*b* values defined in a CIE L*a*b* space.

13. A system according to claim 11, wherein the colorimetric value signal is a signal representing XYZ values defined in an XYZ space.

14. A system according to claim 11, wherein the colorimetric value signal is a signal representing Yxy values defined in a Yxy space.

15. A system according to claim 11, wherein the colorimetric value signal is a signal representing L*u*v* values defined in a L*u*v* space.

16. A system according to claim 11, wherein the colorimetric value signal is a signal representing HVC Munsell values defined in a HVC Munsell space.

17. A system according to claim 11, wherein the color separation value signal is a signal of CMY values representing ratios of cyan (C), magenta (M), and yellow (Y).

18. A system according to claim 11, wherein the color separation value signal is an RGB value signal representing ratios of red (R), green (G), and blue (B).

19. A color image reproduction system comprising:

a transmission path for connecting a plurality of units;

a first unit comprising first colorimetry means for generating a colorimetric value signal by colorimetrically measuring a color sample, and transmission means for transmitting the colorimetric value signal generated by said first colorimetry means via said transmission path; and

a second unit comprising second colorimetry means for generating a colorimetric value signal by colorimetrically measuring a color sample, image output means having a function of outputting a color image corresponding to a color separation value signal and a function of outputting a standard color sample in correspondence with predetermined color separation value signals, first conversion means, constituted by a neural network, for converting the colorimetric value signals generated by said first and second colorimetry means into color separation value signals in accordance with setting of a parameter, storage means for storing the parameter and data represented by the predetermined color separation value signals, learning control means for generating the parameter and storing the generated parameter in said storage means by executing learning of said neural network of said first conversion means, so that colorimetric value signals of the standard color sample, which signals are generated by said second colorimetry means, are converted into the color separation value signals of the data stored in said storage means, and output means for setting the parameter in said conversion means, converting a colorimetric value signal of an object to be reproduced, which signal is sent from said first unit via said transmission path, into a color separation value signal using said conversion means, and supplying the converted color separation value signal to said image output means to output a color image corresponding to the colorimetric value signal of the object to be reproduced.

20. A system according to claim 19, wherein said second unit comprises means for transmitting a colorimetric value signal of a second object to be reproduced, which signal is generated by said second colorimetry means, to said first unit via said transmission path, and

said first unit comprises second image output means having a function of outputting a color image corresponding to a color separation value signal and a function of outputting a standard color sample in correspondence with predetermined color separation value signals, second conversion means, constituted by a neural network, for converting the colorimetric value signals generated by said first and second colorimetry means into color separation value signals in accordance with setting of a parameter, second storage means for storing the parameter and data represented by the predetermined color separation value signals, second learning control means for generating the parameter and storing the generated parameter in said second storage means by executing learning of said neural network of said second conversion means, so that colorimetric value signals of the standard color sample, which signals are generated by said first colorimetry means, are converted into the color separation value signals of the data stored in said second storage means, and second output means for setting the parameter stored in said second storage means in said second conversion means, converting a colorimetric value signal of the second object to be reproduced, which signal is sent from said second unit via said transmission path, into a color separation value signal using said second conversion means, and supplying the converted color separation value signal to said second image output means to output a color image corresponding to the colorimetric value signal of the second object to be reproduced.

21. A color image reproduction system comprising:

a transmission path for connecting a plurality of units;

a first unit comprising colorimetry means for generating a colorimetric value signal by colorimetrically measuring a color sample, first image output means having a function of outputting a color image corresponding to a color separation value signal and a function of outputting a standard color sample in correspondence with predetermined color separation value signals, first conversion means, constituted by a neural network, for converting the colorimetric value signal generated by said colorimetry means into a color separation value signal in accordance with setting of a parameter, storage means for storing the parameter and data represented by the predetermined color separation value signals, learning control means for generating the parameter and storing the generated parameter in said storage means by executing learning of said neural network of said first conversion means, so that colorimetric value signals of the standard color sample, which signals are generated by said colorimetry means, are converted into the color separation value signals of the data stored in said storage means, and transmission means for generating color information on the basis of a colorimetric value signal of an object to be reproduced, which signal is generated by said colorimetry means, and the parameter stored in said storage means, and transmitting the color information onto said transmission path; and

a second unit comprising second image output means which outputs a color image in correspondence with a color separation value signal and is of the same type as said first image output means, second conversion means, constituted by a neural network, for converting a colorimetric value signal into a color separation value signal in accordance with setting of a parameter, and output means for setting the parameter included in the color information sent from said first unit in said second conversion means, converting a colorimetric value signal of an object to be reproduced, which signal is included in the color information, into a color separation value signal using said second conversion means, and supplying the converted color separation value signal to said second image output means to output a color image corresponding to the colorimetric value signal of the object to be reproduced.