A SP3A Image Restoration Algorithm for Removing Salt-and-pepper Noise
ZHANG Pei
FANG Zhuang
Abstract：To address the issue that high-density salt-and-pepper noise severely damaged image details,a novel image restoration algorithm for removing salt-and-pepper noise with the plug-and-play alternating direction method of multipliers(SP3A)was proposed.Firstly,the noise locations were identified by a noise detection filter,and the image denoising problem was transformed into an image inpainting problem.Secondly,a mathematical model based on variational energy was established.Using the plug-and-play alternating direction method of multipliers(PnP-ADMM)as the solution framework,the regularization subproblem was replaced by an existing denoiser,and the SP3A algorithm was derived.Finally,the convergence of the SP3A algorithm was demonstrated under the condition that the denoiser was bounded.Numerical experiments showed that the SP3A algorithm was improved by 0.74 dB in peak signal-to-noise ratio(PSNR)and 0.12 in structural similarity index measure(SSIM),compared with L0-overlapping group sparse total variation(L0-OGSTV)algorithm under high noise levels.This study confirmed that the detailed texture parts of images were effectively restored by the SP3A algorithm,and a new solution was provided for the removal of high-density salt-and-pepper noise.
Keywords：salt-and-pepper noiseimage restorationplug-and-play priorADMM algorithmnoise levelpeak signal to noise ratio
Publication Date：2025-12-30
Online Publishing Date：2025-12-10(First online date of this platform, not the publication date of the document)
Pages：8( 558-565 )