# US 2010/0169255 A1 · Coating color database creating method, search method using the database, their system, program, and recording medium

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2010/0169255 A1
- Document ID: US20100169255A1
- Status: Published application
- Kind: A1
- Application number: 12/663,364
- Filed: 2008-06-19
- Published: 2010-07-01
- Earliest claimed: 2007-06-20
- Assignees: Kansai Paint Co., Ltd.
- Inventors: Tsukasa Fujieda; Keisuke Sai; Yutaka Masuda
- Web page: https://patentgrep.com/patent/US20100169255A1

## CPC

- G06V 10/56: relating to colour
- C23C 16/52: Controlling or regulating the coating process {(C23C16/45557, C23C16/279 take precedence)}
- G01J 3/46: Measurement of colour; Colour measuring devices, e.g. colorimeters (measuring colour temperature G01J5/60)
- G01J 3/462: {Computing operations in or between colour spaces; Colour management systems}
- G06N 3/0499: Feedforward networks
- G06N 3/09: Supervised learning
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- C23C: COATING METALLIC MATERIAL; COATING MATERIAL WITH METALLIC MATERIAL; SURFACE TREATMENT OF METALLIC MATERIAL BY DIFFUSION INTO THE SURFACE, BY CHEMICAL CONVERSION OR SUBSTITUTION; COATING BY VACUUM EVAPORATION, BY SPUTTERING, BY ION IMPLANTATION OR BY CHEMICAL VAPOUR DEPOSITION, IN GENERAL (making metal-coated products by extrusion B21C23/22; covering with metal by connecting pre-existing layers to articles, see the relevant places, e.g. B21D39/00, B23K; metallising of glass C03C; metallising mortars, concrete, artificial stone, ceramics or natural stone C04B41/00; enamelling of, or applying a vitreous layer to, metals C23D; treating metal surfaces or coating of metals by electrolysis or electrophoresis C25D; single-crystal film growth C30B; by metallising textiles D06M11/83; decorating textiles by locally metallising D06Q1/04)
- G01J: MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY (light sources F21, H01J, H01K, H05B; investigating properties of materials by optical means G01N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

The subject invention provides a method of creating a database for searching for a paint color having a desired texture, a search method using the database, and systems, programs, and recording mediums for carrying out the method and the search. The method for creating a database includes a step (S11) for storing spectral reflectance data and micro-brilliance data of a plurality of paint colors after associating each spectral reflectance data and each micro-brilliance data with a paint color code; a step (S13) for storing texture evaluation values of sample paint colors after associating the each texture evaluation value with the paint color code; a step (S14) for calculating characteristic quantities of the paint colors expressing textures using the spectral reflectance data and the micro-brilliance data, and storing the characteristic quantities after associating the each characteristic quantity with the paint color code; a step (S15) for carrying out a process for training a neural network using the characteristic quantities and the texture evaluation values of the sample paint colors as training data; and a step (S16) for inputting characteristic quantities of the paint colors other than the sample paint colors into the neural network after the training process, and storing output data after associating the each output data with the paint color code.

## Claims

### Claim 1

A method for creating a paint color database, comprising:
a first step of storing spectral reflectance data and micro-brilliance data expressing particle feeling of a plurality of paint colors in a recording unit after associating each spectral reflectance data and each micro-brilliance data with a code for specifying each paint color;
a second step of storing texture evaluation values of sample paint colors selected from the plurality of paint colors in the recording unit after associating the each texture evaluation value with the code;
a third step of calculating characteristic quantities of the paint colors expressing textures using the spectral reflectance data and the micro-brilliance data, and storing the characteristic quantities in the recording unit after associating the each characteristic quantity with the code;
a fourth step of carrying out a process for training a neural network having input units corresponding to the characteristic quantities and output units corresponding to the texture evaluation values, using the characteristic quantities and the texture evaluation values of the sample paint colors as training data; and
a fifth step of inputting characteristic quantities of the paint colors other than the sample paint colors into the neural network after the training process, and storing output data in the recording unit after associating the each output data with the code.

### Claim 2 (depends on 1)

The method for creating a paint color database according to claim 1,
wherein:
the first step comprises a step of obtaining image data of each paint color using an imaging device and calculating micro-brilliance data from the image data,
the micro-brilliance data are HG, HB, HBL and SB,
the HG has a condition such that:
in cases where IPSL≧0.32: HG=500·IPSL−142.5
in cases where 0.32>IPSL≧0.15: HG=102.9·IPSL−15.4
in cases where 0.15>IPSL: HG=0
the HB has a condition such that:
HB=(BV−50)/2
the HBL is an HB found from the image data taken under a condition where GL=14, and the SB is an HG found from the image data taken under a condition where GL=125,
wherein GL represents an average gray level of image data, V and A respectively represent gross luminance volume and gross luminance area greater than a threshold=“GL+32” binarized by the threshold, L represents an average particle diameter binarized by a threshold=“GL+24”, PHav and PSav respectively represent an average peak height and an average peak skirt of luminance image that satisfy: PHav=3V/A, PSav=L/PHav and BV=PHav+350 PSav,
IPSL satisfies:
IPSL=∫0N∫02πP(ν,θ)dνdθ/P(0,0)
wherein ν represents spatial frequency, θ represents angle, P(ν,θ) represents power spectrum found from the image data, and 0 to N are spatial frequency regions for the particle feeling.

### Claim 3 (depends on 1)

The method for creating a paint color database according to claim 1,
wherein:
the characteristic quantities include:
a characteristic quantity L*15, a characteristic quantity a*15, and a characteristic quantity b*15, which are L* value, a* value, and b* value calculated from a spectral reflectance at an observation angle of 15° in the L*a*b* color space, a characteristic quantity L*45, a characteristic quantity a*45, and a characteristic quantity b*45, which are L* value, a* value, and b* value calculated from a spectral reflectance at an observation angle of 45° in the L*a*b* color space, a characteristic quantity L*75, a characteristic quantity a*75, and a characteristic quantity b*75, which are L* value, a* value, and b* value calculated from a spectral reflectance at an observation angle of 75° in the L*a*b* color space;
a characteristic quantity FF(15,25), a characteristic quantity FF(25,45), a characteristic quantity FF(45,75), a characteristic quantity FF(75,110) and a characteristic quantity FF(15,45), which are found according to 2×(Y15−Y25)/(Y15+Y25), 2×(Y25−Y45)/(Y25−Y45), 2×(Y45−Y75)/(Y45+Y75), 2×(Y75−Y110)/(Y75+Y110), and 2×(Y15−Y45)/(Y15+Y45), respectively, wherein Y15 represents a Y value calculated from a spectral reflectance at an observation angle of 15° in the XYZ color space, Y25 represents a Y value calculated from a spectral reflectance at an observation angle of 25° in the XYZ color space, Y45 represents a Y value calculated from a spectral reflectance at an observation angle of 45° in the XYZ color space, Y75 represents a Y value calculated from a spectral reflectance at an observation angle of 75° in the XYZ color space, Y110 represents a Y value calculated from a spectral reflectance at an observation angle of 110° in the XYZ color space;
a characteristic quantity c*15 that is a c* value calculated from a spectral reflectance at an observation angle of 15° in the L*C*h* color space, a characteristic quantity c*45 that is a c* value calculated from a spectral reflectance at an observation angle of 45° in the L*C*h* color space, and a characteristic quantity c*75 that is a c* value calculated from a spectral reflectance at an observation angle of 75° in the L*C*h* color space;
a characteristic quantity cFF(15,25) that is found according to 2×(c*15−c*25)/(C*15+c*25), a characteristic quantity cFF(25,45) that is found according to 2×(c*25−c*45)/(c*25+c*45), a characteristic quantity cFF(45,75) that is found according to 2×(c*45−c*75)/(c*45+c*75), a characteristic quantity cFF(75,110) that is found according to 2×(c*75−c*110)/(c*75+c*110), and a characteristic quantity cFF(15,45) that is found according to 2×(c*15−c*45)/(c*15+c*45), wherein c*25 represents a c* value calculated from a spectral reflectance at an observation angle of 25° in the L*C*h* color space, and c*110 represents a c* value calculated from a spectral reflectance at an observation angle of 110° in the L*C*h* color space; and
characteristic quantities HG, HB, HBL and SB which are the micro-brilliance data.

### Claim 4

A method for creating a paint color database, comprising:
a first step of storing spectral reflectance data of a plurality of paint colors in a recording unit after associating each spectral reflectance data with a code for specifying each paint color;
a second step of storing color category evaluation values of sample paint colors selected from the plurality of paint colors in the recording unit after associating the each color category evaluation value with the code;
a third step of calculating characteristic quantities of the paint colors expressing color categories using the spectral reflectance data, and storing the characteristic quantities in the recording unit after associating the each characteristic quantity with the code;
a fourth step of carrying out a process for training a neural network having input units corresponding to the characteristic quantities and output units corresponding to the color category evaluation values, using the characteristic quantities and the color category evaluation values of the sample paint colors as training data; and
a fifth step of inputting characteristic quantities of the paint colors other than the sample paint colors into the neural network after the training process, and storing output data in the recording unit after associating the each output data with the code.

### Claim 5 (depends on 4)

The method for creating a paint color database according to claim 4,
wherein:
the characteristic quantities include:
a characteristic quantity L*15, a characteristic quantity a*15, and a characteristic quantity b*15, which are L* value, a* value, and b* value calculated from a spectral reflectance at an observation angle of 15° in the L*a*b* color space, a characteristic quantity L*45, a characteristic quantity a*45, and a characteristic quantity b*45, which are L* value, a* value, and b* value calculated from a spectral reflectance at an observation angle of 45° in the L*a*b* color space, a characteristic quantity L*110, a characteristic quantity a*110, and a characteristic quantity b*110, which are L* value, a* value, and b* value calculated from a spectral reflectance at an observation angle of 110° in the L*a*b* color space;
a characteristic quantity c*15 that is a c* value calculated from a spectral reflectance at an observation angle of 15° in the L*C*h* color space, a characteristic quantity c*45 that is a c* value calculated from a spectral reflectance at an observation angle of 45° in the L*C*h* color space, and a characteristic quantity c*110 that is a c* value calculated from a spectral reflectance at an observation angle of 110° in the L*C*h* color space;
a characteristic quantity sin(h15); a characteristic quantity cos(h15), a characteristic quantity sin(h45), a characteristic quantity cos(h45), a characteristic quantity sin(h110) and a characteristic quantity cos(h110), wherein h15 represents a hue angle at an observation angle of 15°, h45 represents a hue angle at an observation angle of 45°, h110 represents a hue angle at an observation angle of 110°, which are found from the a*15, b*15, a*45; b*45, a*110, and b*110 according to h=tan−1(b*/a*); and
a characteristic quantity FF(15,45), which is found according to 2×(Y15−Y45)/(Y15+Y45), wherein Y15 represents a Y value calculated from a spectral reflectance at an observation angle of 15° in the XYZ color space, and Y45 represents a Y value calculated from a spectral reflectance at an observation angle of 45° in the XYZ color space.

### Claim 6 (depends on 1)

A method for searching for a paint color in a paint color database created by the method for creating a paint color database according to claim 1, comprising:
a first step of receiving, as a search query, texture and an evaluation value that denotes a degree of presence of the texture;
a second step of retrieving an evaluation value corresponding to the texture from the database, and determining whether the evaluation value is a value indicating the presence of the texture; and
a third step of, if determining that the evaluation value is a value indicating the presence of the texture in the second step, giving a corresponding paint color code as a search result.

### Claim 7 (depends on 4)

A method for searching for a paint color in a paint color database created by the method for creating a paint color database according to claim 4, comprising:
a first step of receiving, as a search query, a color category and an evaluation value that denotes a degree of attribution to the color category;
a second step of retrieving an evaluation value corresponding to the color category from the database, and determining whether the evaluation value is a value indicating the attribution to the color category; and
a third step of, if determining that the evaluation value is a value indicating the attribution to the color category in the second step, giving a corresponding paint color code as a search result.

### Claim 8

A system for creating a paint color database, comprising:
an arithmetic unit having a recording unit;
a spectrophotometer; and
an imaging device,
wherein:
the arithmetic unit measures spectral reflectance data for each of a plurality of paint colors using the spectrophotometer,
the arithmetic unit obtains image data of the paint colors using the imaging device, and calculates micro-brilliance data expressing particle feeling of the paint colors from the image data,
the arithmetic unit stores the spectral reflectance data and the micro-brilliance data of the paint colors in the recording unit after associating each spectral reflectance data and each micro-brilliance data with a code for specifying each paint color,
the arithmetic unit stores texture evaluation values of sample paint colors selected from the paint colors in the recording unit after associating the each texture evaluation value with the code,
the arithmetic unit calculates characteristic quantities of the paint colors expressing textures using the spectral reflectance data and the micro-brilliance data, and stores the characteristic quantities in the recording unit after associating the each characteristic quantity with the code,
the arithmetic unit carries out a process for training a neural network having input units corresponding to the characteristic quantities and output units corresponding to the texture evaluation values, using the characteristic quantities and the texture evaluation values of the sample paint colors as training data; and
the arithmetic unit inputs characteristic quantities of the paint colors other than the sample paint colors into the neural network after the training process, and stores output data in the recording unit after associating the each output data with the code.

### Claim 9

A system for creating a paint color database, comprising:
an arithmetic unit having a recording unit; and
a spectrophotometer,
wherein:
the arithmetic unit measures spectral reflectance data for each of a plurality of paint colors using the spectrophotometer,
the arithmetic unit stores the spectral reflectance data of the paint colors in the recording unit after associating each spectral reflectance data with a code for specifying each paint color,
the arithmetic unit stores color category evaluation values of sample paint colors selected from the paint colors in the recording unit after associating the each color category evaluation value with the code,
the arithmetic unit calculates characteristic quantities of the paint colors expressing color categories using the spectral reflectance data, and stores the characteristic quantities in the recording unit after associating the each characteristic quantity with the code,
the arithmetic unit carries out a process for training a neural network having input units corresponding to the characteristic quantities and output units corresponding to the color category evaluation values, using the characteristic quantities and the color category evaluation values of the sample paint colors as training data; and
the arithmetic unit inputs characteristic quantities of the paint colors other than the sample paint colors into the neural network after the training process, and stores output data in the recording unit after associating the each output data with the code.

### Claim 10 (depends on 1)

A system for searching for a paint color in a paint color database created by the method for creating a paint color database according to claim 1, comprising:
an arithmetic unit having a recording unit storing the database,
wherein:
the arithmetic unit receives, as a search query, texture and an evaluation value that denotes a degree of presence of the texture,
the arithmetic unit retrieves an evaluation value corresponding to the texture from the database, and determines whether the evaluation value is a value indicating the presence of the texture; and
if determining that the evaluation value is a value indicating the presence of the texture in the second step, the arithmetic unit gives a corresponding paint color code as a search result.

### Claim 11 (depends on 4)

A system for searching for a paint color in a paint color database created by the method for creating a paint color database according to claim 4, comprising:
an arithmetic unit having a recording unit storing the database,
wherein:
the arithmetic unit receives, as a search query, a color category and an evaluation value that denotes a degree of attribution to the color category,
the arithmetic unit retrieves an evaluation value corresponding to the color category from the database, and determines whether the evaluation value is a value indicating the attribution to the color category; and
if determining that the evaluation value is a value indicating the attribution to the color category in the second step, the arithmetic unit gives a corresponding paint color code as a search result.

### Claim 12

A program for creating a paint color database by causing a computer to realize:
a first function for storing spectral reflectance data and micro-brilliance data expressing particle feeling of a plurality of paint colors in a recording unit after associating each spectral reflectance data and each micro-brilliance data with a code for specifying each paint color;
a second function for storing texture evaluation values of sample paint colors selected from the plurality of paint colors in the recording unit after associating the each texture evaluation value with the code;
a third function for calculating characteristic quantities of the paint colors expressing textures using the spectral reflectance data and the micro-brilliance data, and storing the characteristic quantities in the recording unit after associating the each characteristic quantity with the code;
a fourth function for carrying out a process for training a neural network having input units corresponding to the characteristic quantities and output units corresponding to the texture evaluation values, using the characteristic quantities and the texture evaluation values of the sample paint colors as training data; and
a fifth function for inputting characteristic quantities of the paint colors other than the sample paint colors into the neural network after the training process, and storing output data in the recording unit after associating the each output data with the code.

### Claim 13

A program for creating a paint color database by causing a computer to realize:
a first function for storing spectral reflectance data of a plurality of paint colors in a recording unit after associating each spectral reflectance data with a code for specifying each paint color;
a second function for storing color category evaluation values of sample paint colors selected from the plurality of paint colors in the recording unit after associating the each color category evaluation value with the code;
a third function for calculating characteristic quantities of the paint colors expressing color categories using the spectral reflectance data, and storing the characteristic quantities in the recording unit after associating the each characteristic quantity with the code;
a fourth function for carrying out a process for training a neural network having input units corresponding to the characteristic quantities and output units corresponding to the color category evaluation values, using the characteristic quantities and the color category evaluation values of the sample paint colors as training data; and
a fifth function for inputting characteristic quantities of the paint colors other than the sample paint colors into the neural network after the training process, and storing output data in the recording unit after associating the each output data with the code.

### Claim 14 (depends on 1)

A program for searching for a paint color in a paint color database created by the method for creating a paint color database according to claim 1, by causing a computer to realize:
a first function of receiving, as a search query, texture and an evaluation value that denotes a degree of presence of the texture;
a second function of retrieving an evaluation value corresponding to the texture from the database, and determining whether the evaluation value is a value indicating the presence of the texture; and
a third function of, if determining that the evaluation value is a value indicating the presence of the texture in the second step, giving a corresponding paint color code as a search result.

### Claim 15 (depends on 4)

A program for searching for a paint color in a paint color database created by the method for creating a paint color database according to claim 4, by causing a computer to realize:
a first function of receiving, as a search query, a color category and an evaluation value that denotes a degree of attribution to the color category;
a second function of retrieving an evaluation value corresponding to the color category from the database, and determining whether the evaluation value is a value indicating the attribution to the color category; and
a third function of, if determining that the evaluation value is a value indicating the attribution to the color category in the second step, giving a corresponding paint color code as a search result.

### Claim 16 (depends on 12)

A computer-readable recording medium storing the program for creating a paint color database according to claim 12.

### Claim 17 (depends on 13)

A computer-readable recording medium storing the program for creating a paint color database according to claim 13.

### Claim 18 (depends on 14)

A computer-readable recording medium storing the program for searching for a paint color according to claim 14.

### Claim 19 (depends on 15)

A computer-readable recording medium storing the program for searching for a paint color according to claim 15.