A Vector Based Fuzzy Filter for Colour Image Sequences

Tom Mélange, Mike Nachtegael and Etienne E. Kerre

This paper introduces a new vector based filter for the removal of additive gaussian noise in colour image sequences. The proposed method consists of two subfilters of which the first is a colour extension of our previously presented fuzzy logic based motion and detail adaptive filter (FMDAF) for greyscale image sequences. The additional second subfilter is a 3D adaptation of the colour restorating subfilter from a recent fuzzy noise reduction method for colour images. Experimental results show that in terms of average PSNR and NCD, the proposed colour extension outperforms the usual colour extension in which the Y component in the YUV transform is filtered with the original greyscale method. However, better objective measures need to be found for correspondence to visual observations.

PDF full paper