Video + mocap data synced

[Bernard]

Dear all,

I am looking for one motion trial where marker trajectory data and video data (e.g. Vicon Vue) are time-synced. The motion can be of any kind such as walking, running, sit to stand. In addition, I was hoping for one static trial of mocap data so I can build a kinematic model of the movement.

I am hoping to benchmark some markerless mocap algorithms but am lacking the "gold-standard" time-synced motion capture data.

You can email me at bl19622@essex.ac.uk if you have data you might be willing to share.

Thank you all for your kind consideration.

Regards,
Bernard

[Edmund Cramp]

The BIOMCH-L discussion “Historical gait data accuracy” documents a test that worked back in the old days of 3D marker motion capture, simply dropping a reflective golf-ball onto a force plate and recording the ball Z-vector in comparison to the force plate Fz-vector timing.
The ball vector test is probably not something that would work in the markerless field these days but we made a small test device for a user back in the early days of video synchronization efforts, which is a foot-switch that lights a LED when the subject’s heel arrives on the ground – the idea was that the LED would indicate the heel strike timing for the video data, allowing it to be compared to the force plate data and the marker based heel-strike calculations.
Markerless vendors need to document how they verify that their markerless data is synchronized with force plates and other data sources so that users can collect reliable data in the biomechanics and clinical environments.