Joint Reconstruction from Upper-Limb Motion-Capture Markers Data
Upper-limb motion is mechanically complex, yet the nervous system performs these movements with effortless coordination. To understand this behavior and enable human-like control in robotics or neurorehabilitation, upper-limb movement datasets are essential. One of the interesting dataset available with extensive modalities is the U-Limb dataset. It consists of a database collected across six European universities, including healthy and post-stroke participants. It contains detailed upper-limb motion-capture marker trajectories, EMG recordings, EEG and additional sensor modalities. However, the raw motion-capture markers alone do not directly provide anatomical joint positions or orientations. A processing pipeline is required to interpret the marker cluster positions and reconstruct meaningful joint kinematics enabling further analysis.
This is an ongoing project, no applications are accepted for it at the moment.