Image reconstruction with multiresolution edge-continuation and non-quadratic data fidelity
Résumé
The standard approach to image reconstruction, which consists in minimizing a cost function combining data- fidelity with edge-preserving regularization, tends to pro- duce noisy object boundaries and to create a staircase ef- fect. We propose to incorporate the smoothness of the edge field via an additional penalty term defined in a mul- tiresolution domain. We provide an efficient half-quadratic algorithm to solve the resulting optimization problem, in- cluding the case when the data-fidelity is non-quadratic and the cost function non-convex.