02-05-2011, 10:27 AM
A Fuzzy Noise Reduction Method for Color Images - IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 16, NO. 5, - java
Abstract—A new fuzzy filter is presented for the reduction of additive
noise for digital color images. The filter consists of two subfilters.
The first subfilter computes fuzzy distances between the color
components of the central pixel and its neighborhood. These distances
determine in what degree each component should be corrected.
All performed corrections preserve the color component
distances. The goal of the second subfilter is to correct the pixels
where the color components differences are corrupted so much that
they appear as outliers in comparison to their environment. Experimental
results show the feasibility of the proposed approach. We
compare with other noise reduction methods by numerical measures
and visual observations. We also illustrate the performance
of the proposed method as preprocessing step for edge detection.
