Adjust color balance of RGB image with chromatic adaptation
Color Balance Image by Specifying Gray Pixel
Read and display an image with a strong yellow color cast.
A = imread('hallway.jpg'); imshow(A) title('Original Image')
Pick a pixel in the image that should look white or gray, such as a point on a pillar. Do not pick a saturated pixel, such as a point on the ceiling light. Display the selected point in green.
x = 2800; y = 1000; gray_val = impixel(A,x,y); drawpoint('Position',[x y],'Color','g');
Use the selected color as reference for the scene illumination, and correct the white balance of the image.
B = chromadapt(A,gray_val);
Display the corrected image. The pillars now appear white as expected, and the rest of the image has no yellow tint.
imshow(B) title('White-Balanced Image')
Color Balance Image in Linear RGB Color Space
Open an image file containing minimally processed linear RGB intensities.
A = imread("foosballraw.tiff");
The image data is the raw sensor data after correcting the black level and scaling to 16 bits per pixel. Interpolate the intensities to reconstruct color. The color filter array pattern is RGGB.
A = demosaic(A,"rggb");
Display the image. Because the image is in linear RGB color space, apply gamma correction so the image appears correctly on the screen.
A_sRGB = lin2rgb(A); imshow(A_sRGB) title("Original Image")
The image has a ColorChecker® chart in the scene. To get the color of the ambient light, display the RGB values of a pixel in one of the neutral patches of the chart. The intensity of the red channel is lower than the intensity of the other two channels, which indicates that the light is bluish green.
x = 1510; y = 1250; light_color = [A(y,x,1) A(y,x,2) A(y,x,3)]
light_color = 1x3 uint16 row vector 7361 14968 10258
Balance the color channels of the image. Use the
ColorSpace name-value argument to specify that the image and the illuminant are expressed in linear RGB.
B = chromadapt(A,light_color,"ColorSpace","linear-rgb");
Display the color-balanced image with gamma correction.
B_sRGB = lin2rgb(B); imshow(B_sRGB) title("Color-Balanced Image")
Confirm that the gray patch has been color balanced. The three color channels in the color-balanced gray patch have similar intensities, as expected.
patch_color = [B(y,x,1) B(y,x,2) B(y,x,3)]
patch_color = 1x3 uint16 row vector 13010 13010 13010
A — RGB image
m-by-n-by-3 numeric array
RGB image, specified as an m-by-n-by-3 numeric array.
illuminant — Scene illuminant
3-element numeric vector
Scene illuminant, specified as a 3-element numeric vector. The illuminant
must be in the same color space as the input image,
Specify optional pairs of arguments as
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.
Before R2021a, use commas to separate each name and value, and enclose
Name in quotes.
I2 = chromadapt(I,uint8([22 97
118]),'ColorSpace','linear-rgb') adjusts the color balance of an
I, in linear RGB color space.
ColorSpace — Color space
'srgb' (default) |
Color space of the input image and illuminant, specified as the
comma-separated pair consisting of
'linear-rgb'. Use the
'linear-rgb' option to adjust the color balance
of an RGB image whose intensities are linear.
Method — Chromatic adaptation method
'bradford' (default) |
Chromatic adaptation method used to scale the RGB values in
A, specified as the comma-separated pair
'Method' and one of:
'bradford'—Scale using the Bradford cone response model
'vonkries'—Scale using the von Kries cone response model
'simple'—Scale using the illuminant
 Lindbloom, Bruce. Chromatic Adaptation. http://www.brucelindbloom.com/index.html?Eqn_ChromAdapt.html.