Video-Based Water Reconstruction
We have developed a method for reconstructing water from real video footage. Previous systems of image based water reconstruction normally required carefully designed capturing set-ups and only reconstructed the geometry of a water surface. In contrast, our method requires a single input video captured in ordinary outdoor conditions and provides a more informative reconstruction that includes 3D velocities. The key is the combination of vision algorithms and physics laws. Shape from shading is used to capture the change of the water’s surface, from which a vertical velocity gradient field is calculated. Such a gradient field is used to constrain the tracking of horizontal velocities by minimizing an energy function as a weighted combination of mass-conservation and intensity-conservation. Hence the final reconstruction contains a dense velocity field that is incompressible in 3D. Our method is efficient and performs consistently well across water of different types.
Researchers
David Pickup, Chuan Li, Darren Cosker, Peter Hall and Phil Willis.
Results
Left: Original video. Middle: Reconstructed surface geometry. Right: Details of the reconstructed surface geometry and 3D velocities within the yellow box.
Papers
Posters
David Pickup, Chuan Li, Darren Cosker, Peter Hall and Phil Willis, Reconstructing Mass-Conserved Water using Shape from Shading and Optical Flow, Bristol-Cardiff Young Vision Researchers' Colloquium (2010).