Question

Using OpenCV.

I have an RGB image, each value is a float.
I also have a standard color correction matrix 3×4.

What would be the fastest way to ‘apply’ this matrix over the image?

If you don’t know color correction… This is a simple matrix operation.

If the Image looks like this (each pixel is 3 floats):

R G B
R G B
R G B
R G B
R G B
.
.
.

Then I would like to perform the following:

1 R G B     [ A1, A2, A3 ] 
1 R G B     [ R1, R2, R3 ]
1 R G B  *  [ G1, G2, G3 ]
1 R G B     [ B1, B2, B3 ]
1 R G B
.
.
.

All the values in the 3×4 matrix are constants.

Thanks.

Answer 1

Multiply by the rgb parts of the color correction matrix:

transform(imageIn,imageOut,M3x3);

Then add in the A channel correction:

add(imageOut,Scalar(A1,A2,A3),imageOut);

Read about transform at opencv2refman
where it implies you could use

transfrom(imageIn,imageOut,M4X3);

to get the same result in one step (it does dst(I) = mtx · [src(I);1], so helpful), avoiding having to add in the A components.
Your pardon, if that should be an M3X4. I am pretty dyslexic when it comes to matrix math, rows vs cols and which comes first.