Getting x, y , z axis data of segments of Body

[Muhammad Muaaz]

Hi,

We are working on project, where we are trying to find the influence of different human activities (like walking ) on the RF channel. I came across OpenSim, and realized that it contains muskloskeleton models which can be scaled closed to real person, and we can associate motion data to simulate the activity. I have tried simulate walking by using gait2359 model and available motion data. I am trying to find answers to following question.

How I can get X,Y,Z position, velocity, and acceleration of body segments like torso, thighs, shanks, feet, head, arms and hands, during the gait cycle. (kindly refer me to documents)?

Secondly, is it possible to lets say using 2392 Gait model, to simulate sit2stand or stand2sit or other activities like running, do I only need to associate motion data of these activities to the mdoel ?

Thanks.

[Thomas Uchida]

How I can get X,Y,Z position, velocity, and acceleration of body segments like torso, thighs, shanks, feet, head, arms and hands, during the gait cycle.

See the BodyKinematics Analysis: https://simtk-confluence.stanford.edu/d ... h+Analyses

Secondly, is it possible to lets say using 2392 Gait model, to simulate sit2stand or stand2sit or other activities like running, do I only need to associate motion data of these activities to the mdoel ?

It depends what you mean by "simulate". Please see the following pages in the Confluence documentation:
- "Overview of the OpenSim Workflow": https://simtk-confluence.stanford.edu/d ... m+Workflow
- "Examples and Tutorials": https://simtk-confluence.stanford.edu/d ... +Tutorials

[Muhammad Muaaz]

It depends what you mean by "simulate".

I mean using the same Gait2392 model, is it possible to make it run, perform stand 2 sit or sit to stand etc. ?

[Thomas Uchida]

I mean using the same Gait2392 model, is it possible to make it run, perform stand 2 sit or sit to stand etc. ?

You could track experimental data or design and optimize controllers (see, e.g., https://www.crowdai.org/challenges/nips ... ing-to-run and https://www.crowdai.org/challenges/neur ... -challenge). Depending on your application, research question, accuracy requirements, etc., you may need to adjust model parameters or use different models for different motions.