【摘要】 传统的基于低秩矩阵恢复的图像去噪算法只对低秩部分进行约束,当高斯噪声过大时,会导致去噪不充分或细节严重丢失。针对此问题,提出了一种新的鲁棒的图像去噪模型。该模型在原有的低秩矩阵核范数约束的基础上引入高斯噪声约束项,此外为了提高低秩矩阵的低秩性和稀疏矩阵的稀疏性,引入了加权的方法。为了考察方法的去噪能力,选取了不同参数类型的混合噪声图像进行仿真,并结合峰值信噪比、结构相似度评价标准与传统的基于低秩矩阵恢复的图像去噪算法进行对比。实验结果表明,加权低秩矩阵恢复的混合噪声图像去噪算法能增加低秩矩阵的低秩性和稀疏矩阵的稀疏性,在保证去噪效果的同时,保留了图像的细节信息,具有更佳的视觉效果,同时,客观评价指标均有所提高。更多还原

【Abstract】 The traditional image denoising algorithm based on low-rank matrix recovery only has the low rank restraint,and when Gaussian noise is too large,it will lead to insufficient denoise or serious loss of detail.To overcome the disadvantages of the image denoising algorithm based on low-rank matrix recovery,a novel robust image denoising algorithm was proposed,which adds Gaussian restraint into the low rank restraint model.Inspired by reweighted L1 minimization for sparsity enhancement,reweighting singular values were used to enhance low rank of a matrix,and an efficient iterative reweighting scheme was proposed for enhancing low rank and sparsity simultaneously.Finally,to verify the denoising capability of the presented approach,images with different noise types and simulation parameters were generated using the presented method and the results were compared with the traditional image denoising algorithm based on lowrank matrix recovery.Performance analysis of peak signal to noise ratio and structural similarity index were carried on at the same time.The experimental results show that the mixture noise image denoising using reweighted low-rank matrix recovery algorithm can enhance low rank and sparsity of a matrix simultaneously,guarantee visual effect and keep the details,at the same time,the objective evaluation indexes are improved.更多还原