Obtaining Camera Pose via TF

edit

This isn't really an answer tbh. Seeing linked ROS tutorial pages is not why I asked the question. If you have a concrete example showing slightly how this unestablished link between the camera and the world could be created, be my guest. Otherwise, anyone can Google these pages and source files.

Maybe I haven't understand your question but you said that your inputs are a link between world and board and a link between board and camera. So your tf tree is world -> board -> camera right ? The tf listener will make the calculation between each link...

... eventhough they are not directly connected. If you run :

rosrun tf tf_echo world camera

You will get the camera pose in relation to world.

You understood it well, that's exactly what I am after. Without having a direct connection, I want to get the camera pose with respect to world. In order words world -> camera.

That's the answer yes, we can add that to the answer's body.

For completeness, the line you were looking for in the tutorial was listener.lookupTransform("/camera", "/world", ros::Time(0), transform);, tf_echo just outputs the result to the command line.