Dual Block Coordinate Forward-Backward Algorithm with Application to Deconvolution and Deinterlacing of Video Sequences

Feriel Abboud (1, 2) , Emilie Chouzenoux (3, 4, 5) , Jean-Christophe Pesquet (3, 4, 5) , Jean-Hugues Chenot (2) , Louis Laborelli (2)

1 LIGM - Laboratoire d'Informatique Gaspard-Monge
2 INA - Institut National de l'Audiovisuel
3 GALEN - Organ Modeling through Extraction, Representation and Understanding of Medical Image Content
4 CVN - Centre de vision numérique
5 LIGM - Laboratoire d'Informatique Gaspard-Monge

Optimization methods play a central role in the solution of a wide array of problems encountered in various application fields, such as signal and image processing. Especially when the problems are highly dimensional, proximal methods have shown their efficiency through their capability to deal with composite, possibly nonsmooth objective functions. The cornerstone of these approaches is the proximity operator, which has become a quite popular tool in optimization. In this work, we propose new dual forward-backward formulations for computing the proximity operator of a sum of convex functions involving linear operators. The proposed algorithms are accelerated thanks to the introduction of a block-coordinate strategy combined with a preconditioning technique. Numerical simulations emphasize the good performance of our approach for the problem of jointly deconvoluting and deinterlacing video sequences.

Mots clés: Block-coordinate approach; Duality; Proximity operator; Deinterlacing; Video processing; Deconvolution

Domaines: Traitement du signal et de l'image [eess.SP]

Feriel Abboud, Emilie Chouzenoux, Jean-Christophe Pesquet, Jean-Hugues Chenot, Louis Laborelli. Dual Block Coordinate Forward-Backward Algorithm with Application to Deconvolution and Deinterlacing of Video Sequences. Journal of Mathematical Imaging and Vision, 2017, 59 (3), pp.415-431. ⟨10.1007/s10851-016-0696-y⟩. ⟨hal-01418393v2⟩