Blind Image Deblurring using the l0 Gradient Prior
Jérémy Anger, Gabriele Facciolo, Mauricio Delbracio
⚠ This is a preprint. It may change before it is accepted for publication.


Many blind image deblurring methods rely on unnatural image priors that are explicitly designed to restore salient image structures, necessary to estimate the blur kernel. In this article, we describe the blur kernel estimation method introduced by Pan, Hu, Su and Yang in 2014 that uses an l0 prior on the gradient image. We analyze the method after removing unnecessary steps, leading to a fast and elegant blur kernel estimator. We present deconvolution results using the estimated blur kernels. Our experiments show the effectiveness of the method as well as some of its shortcomings.