Toward an affordable and user-friendly visual motion capture system.

Area of Science:

• Biomechanics
• Robotics
• Computer Vision

Background:

• Accurate estimation of joint angles is crucial for understanding human motion and developing advanced robotic systems.
• Existing motion capture systems can be expensive and complex, limiting their accessibility.
• Low-cost, portable solutions are needed for widespread application in research and clinical settings.

Purpose of the Study:

• To design and evaluate an affordable, user-friendly, and portable system for estimating arm joint angles during grasping motions.
• To utilize a single RGB-D camera and customized markers for non-invasive motion tracking.
• To validate the system's accuracy against a gold-standard stereophotogrammetric approach.

Main Methods:

• Development of a system using a single RGB-D camera and three custom markers.
• Automatic detection and tracking of marker positions.
• Application of an offline inverse kinematics process incorporating biomechanical constraints to refine marker data.
• Handling of noise and marker occlusion through the inverse kinematics approach.
• Validation on four subjects performing various grasping-like movements.

Main Results:

• The proposed system demonstrated good accuracy in estimating joint angles.
• High correlation coefficient (r=0.92) was achieved when comparing the system's estimates to a stereophotogrammetric system.
• Low average Root Mean Square (RMS) error of 3.8 degrees was recorded, indicating precise measurements.

Conclusions:

• The developed low-cost system provides a viable and accurate solution for arm joint angle estimation.
• The integration of biomechanical constraints effectively improves the robustness and accuracy of the marker-based tracking.
• This system offers a practical and accessible alternative for motion analysis in diverse applications.