Hi, I am learning the scaling tool. I have some mocap data with 25 experimental markers (upper body only). I am trying to scale the full body opensim model with 83 virtual markers. The two sets of markers have different names. How would I go about this?
I have checked out the tutorial. If I understand correctly, do I need to rename my 25 experimental markers to match 25 of the virtual markers, and discard the rest?
Or should I delete all 83 virtual markers, manually create 25 markers on the model and change their locations, then do the scaling?
Many thanks.
Scaling from mocap data (marker names don't match)
- Junwei Sun
- Posts: 12
- Joined: Mon Mar 14, 2022 11:05 am
- Ayman Habib
- Posts: 2238
- Joined: Fri Apr 01, 2005 12:24 pm
Re: Scaling from mocap data (marker names don't match)
Hello,
You can go either way (rename model markers or create new ones), it maybe more maintainable to do the latter and create a markers file that matches the names/placement of the experimental markers and save this XML file separately then add to the model during the scale step so you don't modify the model file (error prone) or the data files (also error prone since editors can modify formatting as a side-effect).
Hope this helps,
-Ayman
You can go either way (rename model markers or create new ones), it maybe more maintainable to do the latter and create a markers file that matches the names/placement of the experimental markers and save this XML file separately then add to the model during the scale step so you don't modify the model file (error prone) or the data files (also error prone since editors can modify formatting as a side-effect).
Hope this helps,
-Ayman
- Junwei Sun
- Posts: 12
- Joined: Mon Mar 14, 2022 11:05 am
Re: Scaling from mocap data (marker names don't match)
Thanks Ayman. I feel like there is quite a lot of work that has to be done manually. Once I added my experimental markers and carefully place them on the correct position on the model, I can go straight into the IK step and it successfully creates the motion result for me. Is it recommended to skip the scaling step?