Image Denoising Method With Robust Weighted Kernel Norm

With the development of technology and data mining technology,human beings have entered a new era of big data.Huge amounts of data bring us rich information,and people demand more and more information about the accuracy and integrity of the information.Because of the disturbance and pollution of the outside uncertainty and random factors,the information is often mixed with a large number of noise,which reduces the accuracy of information.Therefore,it is becoming more and more important for people to denoise the damaged information.In this paper,the research background,significance and research status of image denoising are briefly introduced.Secondly,the definitions and theorems of the denoising algorithm are systematically studied.We comprehensively summarize the robust principal component analysis,the weighted robust principal component analysis,the l_2,1 norm robust principal component analysis model,and give algorithms for solving these models.Finally,in order to solve the problem of robust principal component analysis,a new model,robust weighted kernel norm principal component analysis model,is proposed in this paper.This model takes the l_2,1 norm substitute instead of the l₁ norm,and weighted kernel norm replace standard kernel norm.We build a robust weighted nuclear norm principal component analysis model,and use augmented Lagrange multiplier method to solve the proposed model.Experiments on matlab software show that the proposed model is better than the robust component analysis model for image denoising.