ofxAruco: translating between opencv camera space and opengl camera space

I am able to detect a AR board using ofxAruco. I am able to draw a 3D box that appears to rest on the board.

I would like to be able to shoot two different angles of the same AR board and calculate the vector between the two cameras. Additionally, I’d like to calculate the scale and rotations I’d have to apply to the first camera to orient it to look in the same direction as the second camera.

I am able to get a modelview matrix of the position of the cube in each camera setup. However, I am having difficulty using these modelview matrices to figure out the relationship between the cameras.

Thank you in advance!