Page 1 of 1

Generating MarkerSet for publicly available .cd3 databases

Posted: Sun Aug 29, 2021 3:00 pm
by vigneshrk
Hello,

Sorry for the inconvenience but I am new to OpenSim and I am trying to use its scaling and inverse kinematic pipelines for my project. I would like to load marker based gait data from publicly available datasets and analyse the effects of subject specific scaling and IK.

I do not have the markerset for such databases but have an idea of where the markers are placed on the body. How do I go about creating a custom Markerset for such datasets?

Thanks

Re: Generating MarkerSet for publicly available .cd3 databases

Posted: Wed Sep 01, 2021 4:39 pm
by aymanh
Hello Vignesh,

Assuming the data was collected using the same protocol, the names and approximate placement of markers on the model would be similar. You can setup a script that takes the data files one at a time, convert it to the formats accepted by OpenSim (trc) and then run scaling and IK. However, before doing any scripting I'd advise that you try scaling/IK on a couple subjects in the GUI, so you can assess the scaling quality/issues, save the setup files and when all is working satisfactorily, switch to the scripting mode for automation.

Hope this helps,
-Ayman

Re: Generating MarkerSet for publicly available .cd3 databases

Posted: Thu Sep 02, 2021 7:08 am
by vigneshrk
Hello Ayman,

Thanks for your response. Two additional questions sorry. Would I need to generate a new Scale marker file based on the experiemental markers on the database to allow virtual markers and experimental markers to match during the IK operation?

Secondly when generating trc files usimg c3dExport provided by OpensIm, do we need to have force plate data? Would it work on c3d files with just marker data?

Thanks a lot

Re: Generating MarkerSet for publicly available .cd3 databases

Posted: Sun Sep 12, 2021 9:35 am
by vigneshrk
Any help with the previous questions?

How to automatically create Scale Marker xml files based on .c3d data? I have not collected the data but using a publicly available data

Thanks,
Vignesh