Simon Masnou

Centre de Recherche en Mathématiques de la Décision, Sorbonne Université

Papers

Total Citations

714

H-Index

About

Simon Masnou is a leading figure in mathematical image processing, best known for his pioneering work on disocclusion—the computational recovery of hidden or missing parts of objects in digital images. His research sits at the intersection of variational methods, partial differential equations, and geometric analysis, with profound applications in object recognition, robotic vision, and image restoration. Masnou’s most influential contribution is his 2002 paper *"Level lines based disocclusion,"* which has garnered over 528 citations. In this work, he introduced a novel framework that leverages the topological structure of level lines to intelligently interpolate occluded areas from their visible surroundings. He further refined this approach in his 2002 companion paper (186 citations), presenting a rigorous variational formulation that elegantly bridges the gap between theoretical mathematics and practical image reconstruction. By formalizing disocclusion as an energy minimization problem, Masnou provided a robust, principled method that has become a cornerstone for subsequent research in inpainting and image completion. His work is celebrated for its mathematical elegance and direct impact on technologies ranging from digital restoration to autonomous vision systems, cementing his reputation as a key innovator in the field.