Iterative methods for total variation image restoration (Q4938112)

We consider the use of iterative methods for total variation models for image restoration. Iterations are needed for solving the nonlinear problem, as well as for the linear algebra problems which arise at each step. Analogous to the situation for solving large discretized partial differential equations in several dimensions, the size of the problems is large enough that direct solution methods are too costly and iterative methods can be much more efficient. The goal of this paper is two-fold. First, we give a concise and self-contained overview of the total variation image restoration models and some existing numerical methods for solving them. Second, we review briefly some of our own research in this area. Specifically, we discuss two topics: the use of primal-dual Newton methods for solving a minimization problem involving the highly nonlinear total variation functional, and preconditioning techniques for differential-convolution type operators. For the entire collection see [Zbl 0921.00025].