Hello!
We are actively collecting opencap data on many subjects across several sport related movements. Is there a recommended data management approach or method that you've found that successfully identifies points of interest to set bounds for particular movements? For example, I have a person performing a drop vertical jump and I'd like an automated method of determining initial contact and toe off in order to time normalize stance (landing) phase and compare to others. We'd usually use force plates, but since there isn't a straight line for time syncing force plates and opencap. I would love your thoughts or others!
Thank you,
Jeff
Identifying initial contact and toe off
- Jeffrey Turner
- Posts: 10
- Joined: Wed Feb 08, 2023 6:59 am
- Scott Uhlrich
- Posts: 125
- Joined: Tue Jan 20, 2015 4:29 pm
Re: Identifying initial contact and toe off
We just use kinematic heuristics. For example, here's code for automatically segmenting squats using pelvis vertical position and velocity: https://github.com/stanfordnmbl/opencap ... ing.py#L39. For drop jumps, i could imagine using foot marker position and velocity for segmenting the landing phase. We'd like to build a library of kinematics-based segmentation algorithms for a variety of activities, so if you come up with some robust ones, open a PR in github.com/stanfordnmbl/opencap-processing and we'll look to incorporate them!