Bilateral filtering as a tool for image smoothing with edge preserving properties

Abstract

In this paper, we give a review on Bilateral filtering as a tool for image smoothing with edge preserving properties. Bilateral filtering is a combination of domain and range filtering. Domain filtering measures the geometric closeness between the neighborhood center and nearby points. Therefore, domain filtering acts as a gaussian filtering. In range filtering, the similarity between two pixels are measured in which similar pixel values are weighted as a high influence while discarding dissimilar ones. We give a review on traditional bilateral filtering. We also describe a modified bilateral filtering based on fuzzy membership function, the fuzzy bilateral filtering. In fuzzy bilateral filtering, the skin is smoothed while maintaining the edges and non skin area, therefore, a more beautiful face image can be achieved. We will demonstrate some of its applications especially on skin smoothing.