Extract rotations about all three axes from 2d projective transformation

Hi

I have estimated a 2-D projective transformation matrix https://se.mathworks.com/help/images/ref/projective2d.html using https://se.mathworks.com/help/images/ref/fitgeotrans.html:

tform = fitgeotrans(imagePoints1,worldPoints,'projective');

Where imagePoints1 (u,v) and worldPoints (x,y) are two sets of points that are related by:

The matrix tform

[x y 1] = [u v 1] * tform

where tform has the form:

[a b c;...

d e f;...

g h i];

the points (u, v) comes from detecting corners of a (plane) checkerboard in an image I have aquireded. The points (x,y) are the points location in coordinates lying in the checkerboard plane. I would like to estimate the orientation of the camera relative to the checkerboard (rotations about the x, y, and z axes). Can this be done using tform and can someone suggest how to do this?

Best regards,

Brian

[Brian]

I found the methods in opencv that can do this. I have not used it but it is described with examples here: https://docs.opencv.org/master/d9/dab/tutorial_homography.html and can used with matlab like described here: https://kyamagu.github.io/mexopencv/matlab/decomposeHomographyMat.html