Data-driven image completion by image patch subspaces.

Hossein Mobahi,Shankar R. Rao,Yi Ma
DOI: https://doi.org/10.1109/PCS.2009.5167452
2009-01-01
Abstract:We develop a new method for image completion on images with large missing regions. We assume that similar patches form low dimensional clusters in the image space where each cluster can be approximated by a (degenerate) Gaussian. We use sparse representation for subspace detection and then compute the most probable completion. Our results show almost no blurring or blocking effects. In addition, both the texture and structure of the missing regions look realistic to the human eye.
What problem does this paper attempt to address?