What's the best way to describe the different co-ordinate frames?

I'm working on computing the rotation to get from the camera to the world frame given the camera's rotation matrix, and the camera's forward vector (assumption is that the camera is always "upright"). The extra complication, is that the camera co-ordinates need to be in OpenGL co-ordinates.

[The identity rotation points in the +x direction in each frame.](https://preview.redd.it/321ph5du7vg91.png?width=560&format=png&auto=webp&s=30c4b73841614c04d9dc4e2de8933a1a5c3809c7)

This is not a coding question, as I have code that does the right thing. But while figuring this out, I realised …

computervision