摘要
图像去噪过程中由于噪声的影响,无法学习到准确的先验知识,因此难以获取较优的稀疏系数。针对该问题,本文提出一种基于非凸加权 lp范数稀疏误差约束的图像去噪算法。该算法将系数求解过程分解为两个子问题,采用广义软阈值算法求解 lp范数中的稀疏系数,再利用代理算法求解稀疏误差约束中的稀疏系数,根据二者的均值来获取更具鲁棒性的稀疏系数。与当前几种典型的算法进行对比分析,实验结果表明:本文算法不仅具有更高的峰值信噪比(PSNR),而且在运行时间上具有更高的效率,同时在视觉角度上产生了更好的视觉感受。
Due to noise during image denoising,it is difficult to learn accurate prior knowledge.Therefore,obtaining a desirable sparse coefficient proves to be difficult.To solve this problem,this paper proposes an image denoising meth-od based on the non-convex weighted-lp-norm sparse-error constraint.This algorithm decomposes the coefficient-solv-ing process into two sub-problems.First,the algorithm solves the sparse coefficient in the lp norm by the generalized soft threshold value algorithm and then uses the surrogate algorithm to solve the sparse coefficient in the sparse-error constraint.Finally,the algorithm obtains a robust sparse coefficient according to its average value.The experimental res-ults show that the proposed algorithm features a high peak signal-to-noise ratio and high efficiency in terms of the run-ning time.Simultaneously,a desirable visual perception is obtained.
作者
徐久成
王楠
王煜尧
徐战威
XU Jiucheng;WANG Nan;WANG Yuyao;XU Zhanwei(College of Computer and Information Engineering,He’nan Normal University,Xinxiang 453007,China;Engineering TechnologyResearch Center for Computing Intelligence and Data Mining in Colleges and University of He’nan Province,Xinxiang 453007,China)
出处
《智能系统学报》
CSCD
北大核心
2019年第3期500-507,共8页
CAAI Transactions on Intelligent Systems
基金
国家自然科学基金项目(61370169,61402153)
河南省科技攻关重点项目(142102210056,162102210261)
河南省高等学校重点科研项目(16A520057)
关键词
图像去噪
稀疏表示
稀疏系数
先验知识
L1范数
非凸加权
LP范数
稀疏误差约束
峰值信噪比
image denoising
sparse representation
sparse coefficient
prior knowledge
l1 norm
non-convex weightedlp norm
sparse error constraint
peak signal-to-noise ratio
