Applying a parametric approach for the task of nonstationary noise removal with missing information
Luis González-Jaime, Mike Nachtegeal, Etienne E. Kerre, Gonzalo Vegas‐Sánchez‐Ferrero, Santiago Aja‐Fernández · 2013
The image capturing process still today introduces degradations that are unavoidable. The research community is compromised with this issue developing algorithms for the noise removal task. Most of the existing approaches in the literature are parametric, i.e. some information from the underlying model is required. However, there are situations in which this information cannot be captured accurately and the use of these approaches is dismissed. Therefore, we propose an approach where averaging functions are applied over different realizations of a parametric filter. Then, the required information for the parametric filter is extracted and combined from the different parameter configurations used. So, we give the possibility to use parametric approaches in situations where some information is missing. Results show that the averaging functions present promising outcomes for the nonstationary noise removal task.