Explicit \textit{a posteriori} error representation for variational problems and application to TV-minimization (Q6845993)

For convex minimization problems using basic convex duality relations. The convex duality relations are transferred to a discrete level, making the explicit a posteriori error representation, practicable from a numerical point of view. Furthermore a generalized Marini formula is provided that determines a discrete primal solution in terms of a given discrete dual solution. All these concepts are benchmarked via the Rudin-Osher-Fatemi model, leading to an adaptive algorithm with a (quasi-optimal) linear convergence rate. The theoretical findings are verified on several numerical test problems: with and without Lipschitz continuous dual solution, with Lipschitz continuous primal and dual solutions, without Dirichlet boundary condition and without exact solution. Numerical experiments with application to image processing are also performed for the Cameraman and Merle images.