Inverse Kinematics

Provide easy-to-use, extensible software for modeling, simulating, controlling, and analyzing the neuromusculoskeletal system.
POST REPLY
User avatar
Yuya Miyazaki
Posts: 16
Joined: Tue May 11, 2021 8:45 pm

Inverse Kinematics

Post by Yuya Miyazaki » Thu Jul 29, 2021 9:49 pm

Hello, I am university student.

I would like to create a motion that uses a bicycle.
However, IK does not work, the limbs and body become shaky.
What should I do in such a case?
https://www.youtube.com/watch?v=V4h10GzsSG0
I am using a translator.

Tags:

User avatar
Najoua Assila
Posts: 34
Joined: Thu Aug 30, 2018 9:24 am

Re: Inverse Kinematics

Post by Najoua Assila » Thu Jul 29, 2021 11:22 pm

Hi,

You can start by verifying your experimental marker data.
Also, while running your inverse kinematics, you obtain error values at each time frame with the name of the marker with max error between brackets, which should give you an idea why the model is extra-shaky at some time frames.
Finally, you can play with the marker weights to minimize your tracking error.
You should find additional information here: https://simtk-confluence.stanford.edu/d ... leshooting

Regards,

Najoua

User avatar
Ayman Habib
Posts: 2252
Joined: Fri Apr 01, 2005 12:24 pm

Re: Inverse Kinematics

Post by Ayman Habib » Fri Jul 30, 2021 10:21 am

Thanks Najoua,

It appears that some frames are solved correctly while others are not. Some reasons to explain:
1. Model has coordinate bounds that prevents IK solver from matching markers well.
2. Missing or mislabeled Markers for the specific frames where solution fails.

If you preview experimental data and show/hide specific markers that are in question you'll be able to tell which is the case.

Hope this helps,
-Ayman

User avatar
Yuya Miyazaki
Posts: 16
Joined: Tue May 11, 2021 8:45 pm

Re: Inverse Kinematics

Post by Yuya Miyazaki » Tue Aug 03, 2021 2:22 am

Dear Najoua Assila, Ayman Habib.

Thank you for reply.
I have confirmed that this error is caused by the camera losing sight of the marker.
This problem was solved by interpolating the missing markers.

Thank you very much.

POST REPLY