An enhanced decision based algorithm for the reduction of high density salt and pepper noise with reduced streaks

An enhanced decision based algorithm for the removal of high density salt and pepper noise with reduced streaking is proposed. A fixed 3×3 window is kept constant for the increasing noise densities. The processed pixel detects salt and pepper noise using the max-min pixel value of the current processing window. If the processed pixel lies between both maximum and minimum value then the processed pixel is termed as noisy. Under high noise densities the algorithm replaces the corrupted pixel with unsymmetrical trimmed median or midpoint or the neighborhood preprocessed pixel or trimmed global mean based on the decision tree formulated. if the processed pixel is termed as uncorrupted, it is left unaltered. The proposed algorithm (PA) is tested on different varying detail images. The proposed algorithm is compared with the standard and recently proposed algorithms and found to give good results both qualitative and quantitatively for increasing noise densities. The proposed algorithm eliminates salt and pepper noise at high noise densities and preserves fine details of an image. The proposed algorithm was compared with DBA, IDBA exclusively to show them that the algorithm eliminates salt and pepper noise with reduced streaking even at high noise densities.