Removal of Impulse Noise using Noise Adaptive Fuzzy Switching Median Filter
Sheena Govindan · 2011
This paper presents a novel method for the removal of random valued impulse noise. The proposed method consists of two stage detection where the noisy pixels are detected followed by filtering which replaces only the noisy pixels. The filtering process is a recursive process where the preprocessed pixels are also considered. Each and every pixel is processed and can restore images which are highly corrupted up to 60% noise density with improved PSNR value and low value of MSE. The image clarity is maintained with improved visual quality.