feat: accept frame transform from non-base frame #145
Conversation
…ology/control-libraries into feature/frame-transform Gitbot updated in remote
| * @return the desired poses | ||
| */ | ||
| std::vector<state_representation::CartesianPose> forward_kinematics(const state_representation::JointPositions& joint_positions, | ||
| const std::vector<unsigned int>& frame_ids); | ||
| const std::vector<unsigned int>& frame_ids, | ||
| const unsigned int ref_frame_id = 0); |
There was a problem hiding this comment.
I'm temped to say that the reference frame has to be a vector too @eeberhard what do you think?
Or should we provide a vector of pairs?
There was a problem hiding this comment.
Do you mean like multiple reference frames so we get back multiple viewpoints of the same frame, or rather we will get back num_of_frame_ids * num_of_ref_frame_ids back
There was a problem hiding this comment.
More in the sense that if you provide a vector of frame ids, you would also need to provide a reference frame id for each frame id that you request.
There was a problem hiding this comment.
IMO a common reference frame is sufficient. From an API side, it's simpler this way. I don't know how common it would be to request each frame in a separate reference frame, and even then you can just use state representation after the fact to transform a frame secondarily if it's super necessary (i.e., get elbow and wrist in base frame, then find wrist in elbow frame afterwards)
|
Maybe we can pick this up another time. Thanks for the initial effort! |
Description
This PR solves the issue by adding a parameter, and doing the frame transform using AICA's state representation
Review guidelines
Estimated Time of Review: 5 minutes
Checklist before merging: