Observations on adaptive vector filters for noise reduction in color images

Abstract

In a series of papers, Plataniotis et al. proposed a number of filters for noise reduction in color images where the noise type is unknown. In this letter, those filters with a unified notation are summarized, and it is shown that they are essentially variants of the same filtering procedure. It is also shown that the class of adaptive vector filters can be considered as interpolants between the arithmetic mean filter and the vector median filter. Results are presented of numerical computations with the filters on test images corrupted with noise. It is found that the adaptive vector filters perform well with general applicability.