•  
  •  
 

Applied Mathematics & Information Sciences

Author Country (or Countries)

Jordan

Abstract

Digital images of various types, gray and color, are used in many vital applications, necessitating the need to rid them of the noise that can infect them during the messaging process. One of the most important types of noise that negatively affects the characteristics of the digital image is salt and pepper noise, which leads to changing some of the pixels in the digital image to values of 0 or 255. The negative effect of this noise increases with the increase in the noise ratio (the number of affected pixels). This paper will discuss a new method to reduce the adverse effects of salt and pepper noise. This research aims to provide an effective way to deal with the noise of salt and pepper, especially if the noise ratio is higher than 50%, which the rest of the filters cannot deal with. This method will be used to treat the affected pixels only by using six matrices with specific dimensions divided into two types: the examination (checking) matrix (WC) and the execution (processing) matrix (WP); where these two types of matrices are used to process the noise-forming pixels. The six generated matrices are used in the first round (3 of each type), while the first two matrices (one of each type) are used again in the second round to eliminate the adverse effects of noise. The proposed method will be implemented, and the obtained experimental results will be compared with median and average filters to show how the proposed method will enhance the quality of the processed noisy image; a visual and statistical analysis will be performed to prove the quality provided by the proposed method. The proposed method will be compared with other existing methods, such as MDBUT MF, MDBPT GMF, AWM F, and AAMF. MSE, PSNR, SSIM, and CC parameters will be used for comparison purposes.

Digital Object Identifier (DOI)

http://dx.doi.org/10.18576/amis/170119

Share

COinS