Use of Idempotent Functions in the Aggregation of Different Filters for Noise Removal

The majority of existing denoising algorithms obtain good results for a specific noise model, and when it is known previously. Nonetheless, there is a lack in denoising algorithms that can deal with any unknown noisy images. Therefore, in this paper, we study the use of aggregation functions for denoising purposes, where the noise model is not necessarily known in advance; and how these functions affect the visual and quantitative results of the resultant images.