Image Denoising Algorithm Based On TV Model

Image processing technology is widely used in computer science,statistics, engineering, informatics, biology, physics, chemistry and social science and technology and other fields. Image denoising problem is a basic problem of image processing and a hot topic in the field of image processing. In this paper, we mainly study the image denoising algorithm based on the Total Variation model. The main work of this paper is as follows.First of all, the basic model and concept of image denoising problem are introduced, and two forms of TV model in the image can be different and non- different cases are given, and two kinds of adaptation conditions are briefly analyzed. At the same time, two effective denoising algorithms Primitive dual algorithm (PDHG algorithm) and Alternating direction multiplier method (ADMM algorithm) are introduced, and the basic idea and iterative scheme of the two algorithms are given.Secondly, a fast fixed point iterative algorithm is proposed to reduce the time of image denoising by using the Nesterov acceleration formula to accelerate the fixed point iterative algorithm. In this paper, a fast iterative algorithm is proposed to solve the problem of image denoising.Experiments verify the effectiveness of the algorithm.Thirdly,a high-order model for removing Poisson noise is proposed by applying the high-order TV model to the Le model with the low denoising rate and the existence of the ladder effect. The Alternating direction multiplier method (ADMM Algorithm) to solve the model. The comparison between the numerical experiments and the Le model shows that the model can better remove the noise in the image and better solve the ladder effect in the image.At last, summarizes and prospects the main contents of the article,and puts forward the further research on the existing problems of some effective algorithms.