Alternating minimization for a single step TV-Stokes model for image denoising
arXiv:2009.11973v2
Originality Synthesis-oriented
AI Analysis
This is an incremental improvement for image processing researchers, offering a generalized approach to denoising models.
The paper tackled image denoising by proposing a fully coupled TV-Stokes model and an alternating minimization algorithm, with the first iteration matching a prior modified TV-Stokes model, and provided a convergence analysis.
The paper presents a fully coupled TV-Stokes model, and propose an algorithm based on alternating minimization of the objective functional whose first iteration is exactly the modified TV-Stokes model proposed earlier. The model is a generalization of the second order Total Generalized Variation model. A convergence analysis is given.