Research On Multi-channel Color Image Denoising Algorithm Under Regular Constraints

With the advancement of digital technology,the acquisition of information becomes easier and easier.Color images with R,G,and B channels are widely used.However,due to various factors such as illumination and transmission,the acquired color images often have a large amount of redundant noise.Therefore,effective processing of large-scale noisy color digital images and fully excavating the essence of digital images have become a hotspot in the research of intelligent information processing.For color image denoising,color image denoising based on the same ratio of R,G,B three-channel matrix has been extensively studied.Due to the self-similarity between each channel of the color image,the noise intensity of the channel is different.Simple splitting in the same proportion will destroy the internal structure of the image.To solve the above problems,a dual-weighted color image denoising model is proposed.Firstly,the R,G,and B channels are divided into blocks,and connects the block matrixes to take advantage of channel redundancy.Then,according to the different noise statistics in each channel,the weighting matrix is introduced to balance the data fidelity.Using the weighted Schatten-p norm as a lowrank penalty term to construct an optimization problem with equality constraints,and use the alternating multiplier direction method to solve the problem.Each iteration update step has a closed-form solution to ensure the convergence of the final result.The experimental results show that compared with the latest denoising algorithm,the proposed algorithm has better performance under the same conditions.The organization structure of this paper is as follows: firstly,it explains the research background,significance and research status of machine learning,comprehensively explains the definitions,theorems and lemmas involved in this paper,and leads to the low-rank matrix approximation model and its improved models.For the existing problems of the current algorithms,a multi-channel denoising model with the minimization of the double-weighted Schatten-p norm is proposed,and the global optimal solution of the model is solved.Secondly,the synthetic data and image data were selected to do experiment on MATLAB,the denoising effect of the proposed model was evaluated through subjective and objective standards,and the experimental results were analyzed and summarized.Finally,summarize the full text.