Section: New Results

Image Reconstruction from Highly Corrupted Gradients

Participants : Hicham Badri, Hussein Yahia, Driss Aboutajdine.

Surface-from-Gradients (SfG) is an important step in many imaging applications. It consists in reconstructing an image/surface from corrupted gradient fields, which results in an ill-posed problem. We propose to use sparsity to regularize the problem. The first approach uses sparsity in the gradient field together with a robust norm on the data-fitting term and was presented at CVPR 2014. The new approach uses a non-local regularization that manipulates non-local similar patches of the corrupted gradient and forcing them to be low-rank. The two approaches significantly outperform previous optimization-based SfG methods on both synthetic and real data.

One submitted publication.