Revision history [back]

You should be able to find this information with a tangent function. With the pose of the camera and pose of the box in some global, identical frame (could even be camera frame) you can use the change in Y and change in X to pull out the azimuth angle if you're looking for 2D. You can use the change in Z as well to pull out azimuth and elevation angles.

It sounds like you already have distance, so that's a good start!