Apparatus and method for color calibration

• US 5,285,297 A
• Filed: 08/19/1992
• Issued: 02/08/1994
• Est. Priority Date: 06/25/1991
• Status: Expired due to Fees

• Current Assignee: Scitex Corporation Ltd. (Suny Cellular Communication Ltd.)
• Original Assignee: Scitex Corporation Ltd. (Suny Cellular Communication Ltd.)
• Inventors: Grossman, Tal, Rose, Oded
• Primary Examiner(s): BRINICH, STEPHEN M
• Application Number:
  US07/931,846
• Time in Patent Office: 538 Days
• Field of Search: 358/530-531, 358/518-523
• US Class Current: 358/518
• CPC Class Codes: H04N 1/6019 using look-up tables H04N1/6025 takes precedence
• CPC Class Codes: H04N 1/6033 using test pattern analysis H04N1/6055 takes precedence

Abstract

A method and apparatus for constructing, training and utilizing an artificial neural network (also termed herein a "neural network", an ANN, or an NN) in order to transform a first color value in a first color coordinate system into a second color value in a second color coordinate system.

Claims

1. A method for modifying a neural network, operative to transform first color values from a first set of color values to second color values from a second set of color values, to fit at least one pair of first and second color values, the method comprising the steps of:

superimposing, on the neural network, a local transformation from the first set of color values to the second set of color values characterized in that the local transformation, superimposed on the neural network, is operative to transform the first color value in the pair to the second color value in the pair without substantially affecting the operation of the neural network on first color values which are distant from the first color value in the pair.

2. A method according to claim 1 and also comprising, prior to said step of superimposing, the step of:

combining a plurality of first color values respectively paired with a plurality of second color values, thereby to generate an individual first color value and an individual second color value to be paired therewith.

3. A method according to claim 2 wherein said step of combining comprises the step of computing an average of said plurality of first color values and an average of said plurality of second color values.

4. A method according to claim 3 wherein said averages each comprise a weighted average.

5. A method according to claim 1 and wherein said first color value in the pair comprises a memorial color.

6. A method according to claim 1 and wherein said sample first color value in the pair comprises a crucial point in the first set of color values.

7. A method according to claim 6 wherein the crucial point comprises a white point.

8. A method according to claim 6 wherein the crucial point comprises a dark point.

9. Apparatus for modifying a neural network, operative to transform first color values from a first set of color values to second color values from a second set of color values, to fit at least one pair of first and second color values, the apparatus comprising:

a local transformation superimposer operative to generate a local transformation from the first set of color values to the second set of color values and to modify the neural network by superimposing the local transformation thereupon, wherein the local transformation, superimposed on the neural network, is operative to transform the first color value in the pair to the second color value in the pair without substantially affecting the operation of the neural network on first color values which are distant from the first color value in the pair.

10. Apparatus for providing a neural network including:

a neural network structure including a plurality of neurons for receiving a first color value from among a first set of color values which first color value is to be transformed into a corresponding second color value from among a second set of color values and for providing an output indication of the corresponding second color value; and

a neural network training system operative to receive a plurality of ordered pairs and to train the neural network structure on the plurality of ordered pairs, each ordered pair comprising a first color value from among the first set of color values and a corresponding second color value from among the second set of color values, one or both of the plurality of first values and the plurality of second color values being a systematic representation of the corresponding one of the first and second sets of color values.

11. Apparatus for constructing a look-up table relating a first multiplicity of values to a second multiplicity of values, the apparatus comprising:

an artificial neural network relating the first multiplicity of values to the second multiplicity of values; and

a LUT representation of the artificial neural network comprising a memory operative to store a plurality of LUT addresses and to store, in respective association therewith, a plurality of output values generated by passing the plurality of LUT addresses through the neural network.

12. A system for constructing apparatus for sampling the color processing characteristics of a color processing device, the color processing device being operative to convert a first representation of a color image to a second representation thereof, the system comprising:

a first representation of a color image comprising a first multiplicity of first color values and a second representation of the color image comprising a second multiplicity of second color values corresponding to the first multiplicity of first color values and having been generated by processing said first representation with the color processing device;

an artificial neural network which, when operated on each individual one of the first multiplicity of first color values, gives a value substantially equal to the corresponding one of said second multiplicity of second color values; and

a neural network operating unit operative to pass the first representation of the color image through the neural network.