Markerless 3D kinematics and force estimation in cheetahs.

Area of Science:

• Biomechanics
• Zoology
• Robotics

Background:

• Studying animal locomotion in the wild is difficult due to the need for non-invasive, remote methods.
• Cheetahs (Acinonyx jubatus) are of particular interest due to their exceptional speed and agility, but obtaining detailed motion data remotely has been a challenge.

Purpose of the Study:

• To present a novel trajectory optimization approach for remote estimation of 3D kinematics and joint torques in wild cheetahs.
• To enable in-depth analysis of animal locomotion in natural environments for biological and robotic applications.

Main Methods:

• Developed the kinetic full trajectory estimation (K-FTE) method using data from wild cheetahs.
• Validated the approach using synchronized video and force plate data.
• Reconstructed 3D kinematics and estimated joint torques remotely.

Main Results:

• Achieved an average reprojection error of 17.69 pixels for 3D kinematics reconstruction.
• Estimated ground reaction forces showed an average root-mean-square error of 171.3N compared to force plate data.
• Estimated joint torques align with findings from previous studies on quadrupeds in controlled environments.

Conclusions:

• The kinetic full trajectory estimation (K-FTE) method provides a viable approach for remote analysis of wild animal movement.
• This technique facilitates a deeper understanding of animal locomotion dynamics in natural settings.
• The findings are valuable for both biologists studying animal movement and roboticists designing legged robots.