Page 1 of 1
Cameras freeze
Posted: Wed Aug 24, 2022 10:08 am
by vienvu408
Once I get to the data collection portion, the cameras freeze. I am able to click record and new session and all, but the skeleton and camera views all freeze and display the person just standing in the calibration position. Any ideas on what may be going wrong? I have a two camera set up 45 degrees from the participant.
Thank you,
Vien
Re: Cameras freeze
Posted: Thu Aug 25, 2022 10:10 pm
by antoinefalisse
When you reach the data collection screen, you will first see the result of the previous step when the participant took a neutral standing pose. This is just a couple of frames, so the skeleton should be static and the videos might look like they are frozen. The videos do not live stream what the phones are filming, they display what has been recorded.
If you check out this example:
https://app.opencap.ai/session/7272a71a ... e11cb7542c, the neutral trial should look like what you see when you reach the data collection step. You can then move on and collect additional trials. When processed, you will see the skeleton moving and the corresponding videos next to it.
Re: Cameras freeze
Posted: Thu Aug 25, 2022 10:24 pm
by vienvu408
Thank you! Resolved. That was my misunderstanding on it.
Do you all plan to have knee abduction values available in the future, or how would you suggest we calculate dynamic knee valgus? As a clinician, I'm not tech savvy, so I was going to use Arundale et al's proposed value for easy comparison between limbs which is below.
At time of peak knee flexion during landing:
Valgus Collapse Value = hip add + hip IR + (knee abd * -1)
More positive value means more valgus LE pattern.
https://onlinelibrary.wiley.com/doi/ful ... /sms.13278
Thank you!
Re: Cameras freeze
Posted: Mon Aug 29, 2022 9:19 pm
by suhlrich
The knee joint in the model we are currently using only moves in flexion and extension. Adding the additional abduction degree of freedom is possible with some fairly easy offline processing in OpenSim. You could add the degree of freedom to the model, and re-run inverse kinematics using this new model and the marker data that comes from the OpenCap data download.
We may offer multiple model options in the online OpenCap pipeline at some point, but this isn't a top priority at the moment.