OpenSim Joint Angles

[Aaron Henry]

Hello,
I have a few questions in regards to how joint angles are calculated in OpenSim. I have looked through the SimTK theory manual but still have some confusion.

1. What method is used to calculate joint angles? Is it quaternions, Euler Angles, or another method?
2. If Euler angles are used how can I determine what sequence was utilized?
2. Are calculations of joint angles model specific, i.e. could one model use Euler angles and another quaternions?
3. How can I change the method of angle calculation if at all?

I may be misunderstanding the solution to the generalized coordinated problem described in the theory manual, so if that is the case please let me know.

Thanks,
Aaron

[Thomas Uchida]

Perhaps you are referring to inverse kinematics- the process of calculating joint angles from marker data? If so, please see the documentation on the Inverse Kinematics Tool: https://simtk-confluence.stanford.edu:8 ... Kinematics. The "How Inverse Kinematics Works" page describes the weighted least squares problem that is solved. If you wish to use a different method to calculate joint angles from marker data, you could write a script using the OpenSim API via Matlab or Python (see the "Scripting" section of the documentation: https://simtk-confluence.stanford.edu:8 ... /Scripting).

[Aaron Henry]

Hi Thomas,
Yes I am referring to inverse kinematics. To give some context, I am attempting to compare OpenSim joint angles to those calculated using the typical method of anatomical coordinate frames. I believe my confusion is coming from determining what would cause differences between the weighted least squares method and the anatomical coordinate frame method. If the marker errors were all 0 in the least squares method (i.e. the model markers match the experimental markers exactly) would all joint angles be exactly the same as those calculated with the anatomical frame method?

[Radhakrishnan Vignesh]

Hi,

@Aaron, not sure if it is too late. From my understanding, if the marker error is 0, it should match the angles calculated using anatomical coordinate frames. That is only if you used a model without constraints. I think adding constraints causes some marker error as some DoF are restricted. Did you manage to solve this? Which anatomical reference method were you using?

I also have a question on how marker positions calculated after least squares is used to determine joint angles?

thanks
Vignesh