I am planning to use OpenCap for my academic research and would like to learn more about the details of the system. I have a few specific questions that I hope you can assist me with:
1. I understand that OpenCap extracts 2D keypoints from images to calculate 3D keypoints. Could you please provide a definition and the locations of all the keypoints in both 2D and 3D? If possible, a diagram illustrating these keypoints would be greatly appreciated.
2. Regarding the kinematic data downloaded, could you clarify the meaning of the positive and negative values? Which directions do these values represent for the movements?
3. Will wearing shoes during data collection affect the accuracy of the data, specifically concerning the ankle joint?
Keypoint markers algorithm
-
Matt Petrucci
- Posts: 185
- Joined: Fri Feb 24, 2012 11:49 am
Re: Keypoint markers algorithm
Hello,
1. The keypoints are dependent on whether you use OpenPose or HRNet. The triangulation of those same keypoints between the cameras is described here.
2. It might be good to go through the OpenSim tutorials here, and you can move the model through different movements so you can get a sense for what is positive and negative.
3. No, but depending on the color of the shoes in contrast with the environment or the leg, it may have a harder time estimating the ankle joint center. See more helpful tips in the tutorial video here.
Hope this helps,
1. The keypoints are dependent on whether you use OpenPose or HRNet. The triangulation of those same keypoints between the cameras is described here.
2. It might be good to go through the OpenSim tutorials here, and you can move the model through different movements so you can get a sense for what is positive and negative.
3. No, but depending on the color of the shoes in contrast with the environment or the leg, it may have a harder time estimating the ankle joint center. See more helpful tips in the tutorial video here.
Hope this helps,