Edge-Preserving Integration of a Normal Field: Weighted Least-Squares, TV and L^1 Approaches

Abstract. We introduce several new functionals, inspired from variational image denoising models, for recovering a piecewise-smooth surface from a dense estimation of its normal field1 . In the weighted least-squares approach, the non-differentiable elements of the surface are a priori detected so as to weight the least-squares model. To avoid this detection step, we introduce reweighted least-squares for minimising an isotropic TV-like functional, and split-Bregman iterations for L1 minimisation.

• Type: Journal
• Year: 2015
• Where: SCALESPACE
• Authors: Yvain Quéau, Jean-Denis Durou
• Keywords: Computer Vision, SCALESPACE 2015