Systems and methods for texture assessment of a coating formulation

Patent metadata

Field	Value
Publication number	US-10147043-B2
Application number	US-201313835675-A
Country	US
Kind code	B2
Filing date	Mar 15, 2013
Priority date	Mar 15, 2013
Publication date	Dec 4, 2018
Grant date	Dec 4, 2018

Abstract

A computer implemented method. The method includes identifying, using a processor, a texture in a target coating, wherein identifying comprises applying a Bayesian process, and assigning, using the processor, a texture value adapted for use by one of a search engine and a formulation engine.

First claim

What is claimed is:

1. A computer implemented method for formulating a matching coating for a target sample, comprising:
generating multiple sets of Bayesian decision points, wherein each set of Bayesian decision points comprises a feed-forward configuration of a plurality of conditional decisions corresponding to a probability that a particular effect pigment is present in a known coating given a presence of one or more reflectance values, the one or more reflectance values comprising at least one chroma value;
wherein one set of Bayesian decision points based on one chroma value measured at an angle differs from another set of Bayesian decision points based on another chroma value taken at the same angle;
storing the generated sets of Bayesian decision points for the known coatings;
receiving, from a spectrophotometer, spectrophotometric information taken at a plurality of different angles of a target coating applied to a target sample;
calculating, using the spectrophotometric information and without a camera, each of the generated sets of Bayesian decision points simultaneously to determine a probability that the target coating comprises a texture type corresponding to at least one of (i) one of multiple coarseness categories, or (ii) one of multiple sparkle categories; and
for each target coating, sending the determined texture type and corresponding probability thereof to a formulation engine, wherein the formulation engine compares the calculated texture type and corresponding probability with one or more texture types determined for the plurality of known coatings to determine a probability that the target coating contains one or more effect pigments found in one of the known coatings.

2. The method as recited in claim 1 , further comprising determining, for each known coating, and for each target coating, a probability value corresponding to the determined coarseness category.

3. The method as recited in claim 1 , further comprising determining, for each known coating, and for each target coating, a probability value corresponding to the determined sparkle category.

4. The method as recited in claim 1 , wherein each texture type is characterized by both: (i) at least one of the multiple coarseness categories, and (ii) at least one of the multiple sparkle categories.

5. The method as recited in claim 1 , further comprising:
the formulation engine determining a pigment concentration for the target coating based on the determined texture type and corresponding probability thereof.

6. The method as recited in claim 1 , wherein:
a first Bayesian decision point comprises a logical set of decisions that emphasize color when the at least one chroma value is greater than a particular value at the angle; and
a second Bayesian decision point comprises a logical set of decisions that emphasize lightness and darkness when the at least one chroma value is less than the particular value at the angle.

7. The method as recited in claim 3 , wherein a particular chroma value at a first angle is 10, such that the at least one chroma value that correlates with the another set of Bayesian decision points is less than 10.

8. The method as recited in claim 1 , further comprising:
calculating, simultaneously, multiple different sets of Bayesian decision points derived from data taken at multiple angles using the another spectrophotometer of the target coating to determine a probability that the target coating comprises at least one effect pigment.

9. The method as recited in claim 1 , further comprising:
calculating the determined sparkle category for the determined texture category of each known and each target sample by calculating each of the following from the corresponding reflectance data:
(i) intensity of sparkle;
(ii) concentration of sparkle within a given radius; and
(iii) similarity indices.

10. The method as recited in claim 1 , further comprising:
determining a contribution to texture of one or more individual effect pigments at various concentrations to (i) determine pigment selection, and (ii) pigment concentration in the target coating.

11. A computer system comprising computer-executable instructions stored thereon that, when executed cause one or more processors in the computer system to perform a method for generating a known coating database for use in formulating a matching coating for a target sample, comprising:
measuring a plurality of known coatings at multiple angles with a multi-angle spectrophotometer, each known coating correlating to a texture category based on the presence of one or more effect pigments in the known coatings, wherein the texture categories are characterized by at least one of (i) one of multiple coarseness categories, or (ii) one of multiple sparkle categories;
determining that each of the known coatings comprise one of two categories: (i) coatings that comprise an effect pigment or (ii) coatings that do not comprise an effect pigment;
associating a set of Bayesian decision points with each determined category, wherein the Bayesian decision points comprise a feed-forward configuration of conditional decisions corresponding to a probability that a coating does or does not contain any effect pigment based on reflectance values corresponding at least to Chroma, Lightness, and Hue;
generating, for the known coatings determined to comprise an effect pigment, at least one of (i) a set of multiple coarseness values, or (ii) a set of multiple sparkle values;
receiving spectral data from the multi-angle spectrophotometer for a target sample;
calculating, using the spectral data and without a camera, all of the set of Bayesian decision points with respect to the received spectral data simultaneously to determine that the target sample comprises an effect pigment; and
storing the spectral data, texture category, coarseness value, and sparkle value corresponding to the new known coating in the database.

12. The method as recited in claim 11 , further comprising determining, for each known coating, a probability value corresponding to the determined coarseness category.

13. The method as recited in claim 11 , further comprising determining, for each known coating, a probability value corresponding to the determined sparkle category.

14. The method as recited in claim 11 , wherein each texture category is characterized by both: (i) at least one of the multiple coarseness categories, and (ii) at least one of the multiple sparkle categories.

15. The method as recited in claim 11 , wherein:
a first Bayesian decision point comprises a logical set of decisions that emphasize color when the at least one chroma value is greater than 10 at the measured angle; and
a second Bayesian decision point comprises a logical set of decisions that emphasize lightness and darkness when the at least one chroma value is less than 10 at the measured angle.

16. The method as recited in claim 11 , further comprising:
calculating the determined sparkle category for the determined texture category of each known coating by calculating each of the following from the corresponding reflectance data:
(iv) intensity of sparkle;
(v) concentration of sparkle within a given radius; and
(vi) similarity indices.

17. The method as recited in claim 11 , further comprising:
determining a contribution to texture of one or more individual pigments in each known coating at various concentrations to (i) determine pigment selection and (ii) pigment concentration in the target sample.

18. A computer system comprising computer-executable instruction

Assignees

• Ppg Ind Ohio Inc

Inventors

• Beymore Paul M

Classifications

• G01J3/463 Primary

  Colour matching · CPC title

• G06N7/01 Primary

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06T7/41

  based on statistical description of texture · CPC title

• G06T2207/10024

  Color image · CPC title

• G06N7/005 Primary

  Physics · mapped topic