%0 Journal Article
%T 基于深度图像先验的椒盐噪声图像去噪<br>Salt-and-Pepper Noise Image Denoising Based on Deep Image Prior
%A 胡锦华
%A 赵泽华
%A 宋苏奇
%A 孙禹
%A 王毅
%A 许建楼
%J Advances in Applied Mathematics
%P 2734-2741
%@ 2324-8009
%D 2024
%I Hans Publishing
%R 10.12677/aam.2024.136262
%X 为了有效地去除图像中的椒盐噪声，本文利用低秩和深度图像先验，提出了一种基于加权核范数的优化模型。为了有效地求解优化模型，本文利用双线性分解，采用交替方向法将原问题分解成几个优化的子问题，对每个子问题给出相应的优化算法。数值实验表明，相比其它先进的方法，假设的新模型取得更好的去噪效果。<br />
To effectively remove salt-and-pepper noise from images, this paper proposes an optimization model based on weighted nuclear norm with the prior knowledge of low rank and depth image prior. To efficiently solve the optimization model, the paper utilizes bilinear decomposition and employs the Alternating Direction Method to decompose the original problem into several optimized sub-problems, for each of which corresponding optimization algorithms are provided. Numerical experiments demonstrate that compared to other advanced methods, the proposed new model achieves better denoising results.
%K 图像去噪，椒盐噪声，加权核范数，交替方向法，深度图像先验<br>Image Denoising
%K Salt-and-Pepper Noise
%K Weighted Nuclear Norm
%K Alternating Direction Method
%K Deep Image Prior
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=89755