# Structure of tf tree.

Hello everyone.

My setup is the following, I have a vlp-16 lidar and a T265 tracking camera and I need an odometry guess to feed it into rtabmap. I have a bag that contains the nav_msgs/Odometry from the T265 and the pointclouds of the vlp-16. Anyway, I can't get my head around how the tf should be structured in order to get this odometry guess. My initial thoughts would be a tree like this: odom-->base_link, base_link---(static transformation)-->vlp-16, base_link--(nav_msgs/Odometry)----> T265.

But as I was looking more into it I saw that the positional part of the pose in the nav_msg starts around 0 and when the topic comes to end it is much further away so I guess that the pose refers to a "global" coordinate frame and as a result it should not be connected to the base_link as this coordinate frame is on the "robot" and it is not a global one. And this is where I got really confused and I do not know how to proceed.

I feel like I'm missing something and I have no clue what is it and it keeps me from figuring out how the structure of the tf tree should be. Does anyone have any idea of what I should do next?

edit retag close merge delete

Can you show TF tree from the bag? rosrun tf view_frames