measureColor

Measure color reproduction using test chart

Syntax

colorValues = measureColor(chart)

[colorValues,colorCorrectionMatrix] = measureColor(chart)

colorValues = measureColor(im,roiPositions,refLAB)

colorValues = measureColor(im,roiPositions,refLAB,Name=Value)

[colorValues,colorCorrectionMatrix] = measureColor(___)

Description

esfrChart or colorChecker Object

Use an esfrChart or a colorChecker object when you want to automatically detect the color regions of interest (ROIs) of a test chart. The esfrChart object supports the Enhanced or Extended version of the Imatest^® eSFR test chart [1]. The esfrChart object supports the Calibrite ColorChecker^® Classic test chart [2].

colorValues = measureColor(chart) measures the color values at all color ROIs of an Imatest eSFR test chart or a Calibrite ColorChecker Classic test chart.

[colorValues,colorCorrectionMatrix] = measureColor(chart) also returns a color correction matrix computed using a linear least squares fit.

Test Chart Image (since R2024a)

Use a test chart image for other types of test charts that are not supported by the esfrChart or colorChecker objects. You must identify the positions of the color ROIs.

colorValues = measureColor(im,roiPositions,refLAB) measures the color values at all color ROIs at positions roiPositions for test chart image im. The reference L*a*b* values of each ROI are given by refLAB.

colorValues = measureColor(im,roiPositions,refLAB,Name=Value) optionally specifies the color space and white point using name-value arguments.

[colorValues,colorCorrectionMatrix] = measureColor(___) also returns a color correction matrix computed using a linear least squares fit, using any combination of input arguments from previous syntaxes.

Examples

Measure Color Accuracy of eSFR Chart

Read an image of an eSFR chart into the workspace.

I = imread("eSFRTestImage.jpg");

Create an esfrChart object, then display the chart with ROI annotations. The 16 color patch ROIs are labeled with white numbers.

chart = esfrChart(I);
displayChart(chart,displayEdgeROIs=false, ...
    displayGrayROIs=false,displayRegistrationPoints=false)

Measure the color in all color patch ROIs.

colorTable = measureColor(chart)

colorTable=16×8 table
    ROI    Measured_R    Measured_G    Measured_B    Reference_L    Reference_a    Reference_b    Delta_E
    ___    __________    __________    __________    ___________    ___________    ___________    _______

     1         67            57            58          38.586           7.541         7.0812       15.15 
     2        156           127           122          62.182          13.225         13.826      9.8745 
     3         73            95           152          49.369        -0.51463        -20.062      18.455 
     4         62            79            58          43.926         -6.8587         17.278      14.849 
     5        104           109           171          53.415           9.457        -22.822       12.99 
     6        118           175           187           69.95         -20.889       -0.21752      13.123 
     7        214           192            69          78.643          1.8052         67.091      9.2399 
     8        154            73           138          46.853          41.998        -17.056      5.1282 
     9         62           120           182           51.05         -15.166        -22.416      22.813 
    10         55            80           185          40.811          8.7346        -44.265      22.782 
    11         79           135            72          55.716         -23.419         28.839      9.4826 
    12        152            53            77          42.759          44.167         7.9536      5.4168 
    13        169            91            52          58.211           27.58         47.578      15.988 
    14        142            63            87          47.012           39.15         8.5453      11.651 
    15         91            67           102          40.591          17.951         -9.525      10.897 
    16        152           183            80          70.505         -16.318         49.811      10.253 

Display the color accuracy measurements. Each square color patch is the measured color, and the thick surrounding border is the reference color for that ROI. Each color accuracy measurement is displayed as Delta_E, the Euclidean distance between measured and reference colors in the CIE 1976 L*a*b* color space. More accurate colors have a smaller Delta_E.

figure
displayColorPatch(colorTable)

For an alternative representation of the color accuracy measurements, plot the measured and reference colors in the CIE 1976 L*a*b* color space on a chromaticity diagram. Red circles indicate the reference color. Green circles indicate the measured color of each color patch. The chromaticity diagram does not portray the brightness of color.

figure
plotChromaticity(colorTable)

ROIs with a shorter distance between the reference and measurement points have smaller differences in chromaticity, which can contribute to a smaller value of Delta_E. However, brightness also contributes to the value of Delta_E. For example, even though the reference and measurement points for ROI 13 are near each other on the chromaticity diagram, they have a large Delta_E because of their large difference in brightness.

Measure Color of ColorChecker Chart

Read an image of a ColorChecker® chart into the workspace.

I = imread("colorCheckerTestImage.jpg");

Create a colorChecker object, then display the chart with ROI annotations.

chart = colorChecker(I);
displayChart(chart)

Measure the color in each color patch ROI.

colorTable = measureColor(chart)

colorTable=24×9 table
    ROI         Color          Measured_R    Measured_G    Measured_B    Reference_L    Reference_a    Reference_b    Delta_E
    ___    ________________    __________    __________    __________    ___________    ___________    ___________    _______

     1     {'DarkSkin'    }       160           129           120           37.54          14.37          14.92       20.193 
     2     {'LightSkin'   }       229           200           191           64.66          19.27           17.5       22.587 
     3     {'BlueSky'     }       146           191           241           49.32          -3.82         -22.54       27.312 
     4     {'Foliage'     }       130           161           117           43.46         -12.74          22.72       20.404 
     5     {'BlueFlower'  }       175           187           248           54.94           9.61         -24.79       23.073 
     6     {'BluishGreen' }       155           232           226           70.48         -32.26          -0.37       18.284 
     7     {'Orange'      }       255           161            99           62.73          35.83           56.5       16.114 
     8     {'PurplishBlue'}       130           164           254           39.43          10.75         -45.17       28.889 
     9     {'ModerateRed' }       252           146           160           50.57          48.64          16.67       23.601 
    10     {'Purple'      }       139           118           175            30.1          22.54         -20.87       24.672 
    11     {'YellowGreen' }       187           226           110           71.77         -24.13          58.19        15.21 
    12     {'OrangeYellow'}       241           194            76           71.51          18.24          67.37       14.148 
    13     {'Blue'        }        96           131           255           28.37          15.42          -49.8        33.34 
    14     {'Green'       }       118           209           130           54.38         -39.72          32.27       22.461 
    15     {'Red'         }       234           116           114           42.43          51.05          28.62        21.87 
    16     {'Yellow'      }       241           227           105            81.8           2.67          80.41       23.495 
      ⋮

Measure Color of Custom Test Chart

Read and display an image of a custom test chart with color patch ROIs. This example simulates a custom test chart image by cropping an Imatest eSFR test chart.

I = imread("RGBColorPatches.jpg");
imshow(I)

Draw ROIs for the red, green, and blue color patches.

numROIs = 3;
roiPos = zeros(numROIs,4);
for cnt = 1:numROIs
    hrect = drawrectangle;
    roiPos(cnt,:) = hrect.Position;
end  

Load the reference L*a*b* values of the color patches, which are saved in a file called RGBColorPatches_refLab.mat. The white point of the reference values is the CIE standard illuminant D50.

load("RGBColorPatches_refLab.mat");

Calculate the color reproduction values for these selected edges, specifying the white point.

colorValues = measureColor(I,roiPos,refLab,ReferenceLABWhitePoint="d50");

On a color patch diagram, display the measured and reference colors and the color error.

displayColorPatch(colorValues)

Input Arguments

chart — Test chart

esfrChart object | colorChecker object

Test chart, specified as an esfrChart object or a colorChecker object.

im — Test chart image

RGB image

Since R2024a

Test chart image, specified as an RGB image.

Data Types: single | double | uint8 | uint16

roiPositions — ROI positions

n-by-4 numeric matrix

Since R2024a

ROI positions, specified as an n-by-4 numeric matrix, where n is the number of ROIs.

refLAB — Reference L*a*b* values

n-by-3 numeric array

Since R2024a

Reference L*a*b* values for each of the selected ROIs, specified as an n-by-3 numeric matrix, where n is the number of ROIs.

Data Types: single | double

Name-Value Arguments

Specify optional pairs of arguments as Name1=Value1,...,NameN=ValueN, where Name is the argument name and Value is the corresponding value. Name-value arguments must appear after other arguments, but the order of the pairs does not matter.

Example: colorValues = measureColor(im,roiPositions,refLAB,InputColorSpace="adobe-rgb-1998") specifies the Adobe RGB 1998 color space.

InputColorSpace — Input color space

"srgb" (default) | "adobe-rgb-1998" | "prophoto-rgb"

Since R2024a

Input color space of test chart image im, specified as "srgb", "adobe-rgb-1998", or "prophoto-rgb". You can specify this name-value argument only when you use a test chart image im.

If you specify a chart object, chart, then the measureColor function uses the sRGB color space.

ReferenceLABWhitePoint — White point

"d65" (default) | 1-by-3 numeric vector

Since R2024a

White point used for the reference L*a*b* values, specified as a 1-by-3 numeric vector or one of the CIE standard illuminants listed in the table. You can specify this name-value argument only when you use a test chart image, im.

Value	White Point
"a"	CIE standard illuminant A, [1.0985, 1.0000, 0.3558]. Simulates typical, domestic, tungsten-filament lighting with correlated color temperature of 2856 K.
"c"	CIE standard illuminant C, [0.9807, 1.0000, 1.1822]. Simulates average or north sky daylight with correlated color temperature of 6774 K. Deprecated by CIE.
"e"	Equal-energy radiator, [1.000, 1.000, 1.000]. Useful as a theoretical reference.
"d50"	CIE standard illuminant D50, [0.9642, 1.0000, 0.8251]. Simulates warm daylight at sunrise or sunset with correlated color temperature of 5003 K. Also known as horizon light.
"d55"	CIE standard illuminant D55, [0.9568, 1.0000, 0.9214]. Simulates mid-morning or mid-afternoon daylight with correlated color temperature of 5500 K.
"d65"	CIE standard illuminant D65, [0.9504, 1.0000, 1.0888]. Simulates noon daylight with correlated color temperature of 6504 K.
"icc"	Profile Connection Space (PCS) illuminant used in ICC profiles. Approximation of [0.9642, 1.000, 0.8249] using fixed-point, signed, 32-bit numbers with 16 fractional bits. Actual value: [31595,32768, 27030]/32768.

• If you specify chart as an esfrChart object, then the measureColor function uses the CIE standard illuminant D65.

• If you specify chart as a colorChecker object, then the measureColor function uses the CIE standard illuminant D50. The reference L*a*b* values are for the "After November 2014" version of the ColorChecker chart.

Output Arguments

colorValues — Color values

table

Color values in each color patch ROI, returned as a table. The table has one row for each measured ROI. The table always has these variables (columns):

Variable	Description
ROI	Index of the sampled ROI.
Measured_R	Mean value of red channel pixels in the ROI. Measured_R is a scalar of the same data type as chart.Image or im.
Measured_G	Mean value of green channel pixels in the ROI. Measured_G is a scalar of the same data type as chart.Image or im.
Measured_B	Mean value of blue channel pixels in the ROI. Measured_B is a scalar of the same data type as chart.Image or im.
Reference_L	Reference L* value of the ROI. Reference_L is a scalar of data type double.
Reference_a	Reference a* value of the ROI. Reference_a is a scalar of data type double.
Reference_b	Reference b* value of the ROI. Reference_b is a scalar of data type double.
Delta_E	Euclidean color distance between the measured and reference color values in the L*a*b* color space. Delta_E is a scalar of data type double. If you specify a test chart object chart, then the color distance is as outlined in CIE 1976. If you specify a test chart image im, then the color difference is based on the CIE DE2000 standard.

If you specify a test chart image im, then the table has this additional variable:

Variable	Description
ROIPositions	Position of the ROI, returned as a 4-element vector of the form [X Y Width Height]. X and Y are the coordinates of the top-left corner of the ROI. Width and Height are the width and height of the ROI, in pixels.

For more information on accessing the measurements within the table, see Access Color Measurements.

colorCorrectionMatrix — Color correction coefficients

4-by-3 matrix

Color correction coefficients, returned as a 4-by-3 matrix. colorCorrectionMatrix represents an affine transformation that you can use to color-correct images that are captured under similar lighting conditions as the test chart image. For an example, see Correct Colors Using Color Correction Matrix.

If you specify a test chart image im, then the measureColor function calculates the color correction matrix in linear color space.

Data Types: double

More About

Access Color Measurements

The order of the variables in the color table depends on whether you input a test chart image, im, or a chart object, chart. If you need to access variables of the color table, refer to variable names and not numeric indices.

For example, to access the Delta_E variable of the color table, use code such as this.

deltaE = colorValues.Delta_E;

For another example, to access the measured RGB values of the ROI with index 3, use code such as this. The first command returns the Measured_R, Measured_G, and Measured_B variables in a table for the ROI with index 3. The second command converts the table into a numeric row vector.

measuredRGB3 = colorValues(colorValues.ROI==3,["Measured_R" "Measured_G" "Measured_B"])
measuredRGB3vec = measuredRGB3{1,:};

Note that you can perform these two operations in a single command:

measuredRGB3vec = colorValues{colorValues.ROI==3,["Measured_R" "Measured_G" "Measured_B"]}

For more information, see Access Data in Tables and Access Data in Cell Arrays.

References

[1] Imatest. "Esfr". https://www.imatest.com/mathworks/esfr/.

[2] Calibrite. "ColorChecker Classic". https://calibrite.com/us/product/colorchecker-classic/.

Version History

Introduced in R2017b

R2024a: Support for additional test charts

You can now measure the color reproduction of color ROIs for test charts other than the Imatest Enhanced or Extended eSFR test chart and the Calibrite ColorChecker Classic test chart.

To measure the color of additional types of test charts, specify the test chart image, im, the position of the color patch ROIs, roiPositions, and the reference L*a*b* values, refLAB.