Open AccessA New Adaptive Filter for Eliminating Salt and Pepper Noise
Author(s) -
Fatimah Shamsulddin Abdulsattar
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/928/3/032001
Subject(s) - salt and pepper noise , pixel , filter (signal processing) , noise reduction , noise (video) , median filter , non local means , artificial intelligence , window (computing) , computer vision , mathematics , computer science , adaptive filter , image processing , image denoising , image (mathematics) , algorithm , operating system
Salt-and-pepper noise can substantially degrade the appearance of images and make further processing difficult. This paper develops a new filter for eliminating different levels of salt-and-pepper noise efficiently without degrading important image details. The filter uses a variable window size. After identifying distorted pixels, if there is at least one or more undistorted pixels are encountered in the processing window, the updated value of the distorted pixel is calculated by the weighted mean of undistorted pixels when a window size is 3×3 and replaced by the mean value of the undistorted pixels with the highest frequency distribution when a window size is larger. This filter is applied twice in order to sufficiently remove high noise levels. The comparison results with other denoising filters indicate that the developed filter has superior or comparable denoising capability in terms of visual appearance and objective measures.