Cloth Motion Capture from Optical Flow

Volker Scholz and Marcus Magnor


Cloth simulation is a well-known subject in computer graphics. The physical models used in simulation are rather complex and we doubt that absolute realism will ever be achieved. The acquisition of real cloth motion is a way to circumvent the simulation problem.

We have developed an algorithm for capturing the motion of deformable surfaces, in particular textured cloth. In a calibrated multi-view camera setup, we determine the optical flow between consecutive video frames and derive 3D vertex motion from optical flow. The resulting velocity field can cause distortions of the triangulated surface. We use a deformable surface model with constraints for vertex distances and curvature for smoothing and regularisation. Tracking errors in long video sequences are corrected by a silhouette matching procedure. The results presented here are based on synthetic cloth animations.

V. Scholz and Marcus A. Magnor, Cloth Motion from Optical Flow, Proceedings of 9th International Fall Workshop on Vision, Modeling and Visualization (VMV 2004), Stanford (California), USA.

 Videos

Input

021

  Left camera (DivX, 2.2 MB)    Middle camera (DivX, 2.2 MB)  Right camera (DivX, 2.2 MB)

Reconstruction


Rotating viewpoint (DivX, 4.6 MB)  Wire frame (DivX, 9.1 MB)    Shaded view (DivX, 0.9 MB)  




Home
Last Update: April 25th, 2005. Copyright © 2004 Volker Scholz.