Robust Human Motion Reconstruction in the Presence of Missing Markers and the Absence of Markers for Some Body Segments

Optical motion capture systems are one of the most frequently used technologies for motion data collection. However, they are affected by the missing marker problem which is almost unavoidable for this technique. Another problem with motion capture systems is the impossibility of completely tracking the motion of some bones (e.g. vertebrae or scapula) using external markers fixed on the skin because they move under it.

We propose three strategies for handling the missing marker problem and the lack of sufficient number of markers for determining the motion of some body segments. These strategies have been developed in the frame of a new reconstruction method called Optimal Tracking Method, which requires the solution of a nonlinear constrained optimization problem. The three strategies have been tested on fourteen movements collected by INRETS in the frame of the REALMAN project. The benefits, drawbacks and limitations of the three strategies are discussed.