I have a little confusion about collision detection and hope to get your reply.
When you modify the poses of collision objects which are used to wrap the robot links, I found the code snippet here.
What I understand is that you only use the vector direction between the joints and use it to reconstruct the poses of the collision objects.
So, my questions are
1) Is this because the default base frame of the fcl collision object is different from the robot base frame?
2) How did you check if the geometry of the fcl wrapps the robot links correctly? Because the collision detection results are important to construct the training set.
Hi!
Thanks for your works and generous sharing.
I have a little confusion about collision detection and hope to get your reply.
When you modify the poses of collision objects which are used to wrap the robot links, I found the code snippet here.
What I understand is that you only use the vector direction between the joints and use it to reconstruct the poses of the collision objects.
So, my questions are
1) Is this because the default base frame of the
fcl collision object
is different from the robot base frame?2) How did you check if the geometry of the
fcl
wrapps the robot links correctly? Because the collision detection results are important to construct the training set.Thanks!