No announcement yet.

Gait/Motion MoCap Biomechanics - Optical Marker Shape

  • Filter
  • Time
  • Show
Clear All
new posts

  • Gait/Motion MoCap Biomechanics - Optical Marker Shape

    Hi all,

    I was wondering if anyone has looked at the difference in using spherical vs. hemispherical markers for MoCap gait analysis biomechanics.

    I believe the hemispherical markers will remain adhered to the skin better than the spherical markers, and will also 'jiggle' less.

    However, the centroid calculated from a hemisphere by the MoCap system will be different than that of a sphere. But is this a problem? Presumably the centroid would move closer to the skin surface with a hemispherical marker. And, in instances where medial/lateral pairs of markers are used to determine a 'virtual' midpoint between them, both marker centroids will have shifted closer to the skin surface, so we shouldn't see any resulting offsets, correct?

    If we go this way, I'll be looking at trialing both markers and comparing outputs. I just wanted to see if there was consensus or opinion from the MoCap biomechanics community.



  • #2
    I think this is a good idea.

    My first thought of a minor problem would be that because the marker is so close to the skin that it slightly reduces the line of sight to that marker from the camera. Seems like it would be minimal depending on the circumference of the segment though.


    • #3
      For general Gait Analysis, hemi-spherical markers are probably not a good idea. When Cameras are looking at the marker from an oblique angle, from the shape of the capture image, it will be hard for the software to compute the cetntroid with resonable accuracy. Hemipsherical markers are genraly used for facial expression capture, with cameras lookig dircetly from the front, and little head movement.
      EDIT: There will likely be more marker labelling errors with hemi-spherical markers.
      Last edited by Manvendra Singh; October 28th, 2020, 12:56 AM.


      • #4
        Thanks both for your input!

        If I'm able to perform a comparison trial, I'll report back to this thread!


        • #5
          Performing a recording with a motion capture system that supports calculating and recording the marker location residuals will give you an indication of the effects. If the cameras are observing a spherical marker from different angles as the marker moves through the data collection volume then each camera will contribute to an estimation of the same centroid when the marker is visible. However if the marker appears to be a sphere when seen head on, and a only semi circle when seen from the side, then the marker centroid estimation will move, reducing the marker location accuracy - it's the same effect that you see when a spherical marker, mounted on a limb, rotates out of sight of the cameras.

          People occasionally say that this is a "problem" with optical marker systems but in reality measurement noise is universal. The advantage of an optical marker system that records the marker residuals is that the measurement noise is measurable and can be seen, it's not hidden by software.