Computer Graphics
TU Braunschweig

High Resolution Image Correspondences for Video Post-Production

High Resolution Image Correspondences for Video Post-Production

We present an algorithm for estimating dense image correspondences. Our versatile approach lends itself to various tasks typical for video post-processing, including image morphing, optical flow estimation, stereo rectification, disparity/depth reconstruction, and baseline adjustment. We incorporate recent advances in feature matching, energy minimization, stereo vision, and data clustering into our approach. At the core of our correspondence estimation we use Efficient Belief Propagation for energy minimization. While state-of-the-art algorithms only work on thumbnail-sized images, our novel feature downsampling scheme in combination with a simple, yet efficient data term compression, can cope with high-resolution data. The incorporation of SIFT (Scale-Invariant Feature Transform) features into data term computation further resolves matching ambiguities, making long-range correspondence estimation possible. We detect occluded areas by evaluating the correspondence symmetry, we further apply Geodesic matting to automatically determine plausible values in these regions.

Author(s):Christian Lipski, Christian Linz, Thomas Neumann, Markus Wacker, Marcus Magnor
Published:December 2012
Journal:Journal of Virtual Reality and Broadcasting (JVRB) Vol. 9.2012
Project(s): Virtual Video Camera  Reality CG 

  title = {High Resolution Image Correspondences for Video Post-Production},
  author = {Lipski, Christian and Linz, Christian and Neumann, Thomas and Wacker, Markus and Magnor, Marcus},
  journal = {Journal of Virtual Reality and Broadcasting ({JVRB})},
  volume = {9.2012},
  number = {8},
  pages = {1--12},
  month = {Dec},
  year = {2012}