3D video-based motion capture system without markers or sensors

[Margrit-R. Meier]

Dear All,

I am interested in a 3D video-based motion capture system without markers or sensors appropriate for rehabilitation settings. It will be used for posture, gait and upper-body analyses, thus having variable sample frequencies would be of benefit.

Requirements:

• Area to cover (without loosing accuracy): 10m x 4m [L x W, more is fine too ]
• Integration with force plates (Bertec)
• Sorftware that provides common outputs such as 3D kinematics, kinetics, force plate data including COP trajectories, temporal-spatial parameters
• Possibility to compare different trials/conditions and create outputs as described above
• Possibility to extract data for statistical purposes
• Portable
• Fair cost

Nice-to-have:

• Overlay of force vectors on video
• Possibility to extract self-running videos of particular motion sequences for teaching purposes

Your assistance is greatly appreciated.

Kind regards,
Margrit Meier, PhD
m.meier@conreha.com

[Boris Doynov]

Re: 3D video-based motion capture system without markers or sensors

Hi Dr. Meier,

Appreciate the email exchange - continuing our discussion here to include the community on use cases.

We provide a markerless motion capture and analysis system intended for physical therapy clinics. We use a single 3D sensor + computer vision algorithms. We track depth without the need for wearables or markers. You can see more at https://www.eumotus.com.

In the context of biomechanical research, while we are not as accurate as gold-standard accuracy traditional mocap systems (e.g. multi-camera, marker-based systems), we track closely in a variety of measures. See https://www.eumotus.com/evidence.

The major benefit in the context of a clinical setting is time savings (order of magnitude +) in eliminating the need for complex calibration and setup of markers, as well as post-processing of data.

How might this help in the world of research-grade biomechanics? I'd love to hear the Biomch-L community's thoughts.

One idea is to extend the range of/access to the mocap lab outside the physical lab. Our system is lightweight and portable, and can be used to gather quantitative data outside the lab, complementing high fidelity mocap systems. Perhaps there is a use case to do movement screens outside the lab. Upon meeting certain criteria/finding athletes/research subjects of interested in the field, referral to the biomechanics lab for in-depth research.

[Jaswandi Pitale]

Re: 3D video-based motion capture system without markers or sensors

Hi,

Am curious, what solution did you go with?

-Jaswandi

[Andrea Castelli]

Re: 3D video-based motion capture system without markers or sensors

Actually the most affordable and accurate solution for markerless motion capture in large volumes and with consistent reliability is:

https://thecaptury.com/

Hope this helps!
Best regards