Improvement Of P-M Model Diffusion Function

Image is the objective reflection of the natural scenery and it plays an important role in the production and life of human society. Image denoising technology generally refers to the processing technology of the computer in the image noise. As an important part of image processing, image denoising technology is the basic and necessary module of many image post-processing techniques.This paper firstly introduces the image noise in the introduction part. We also introduce the history and research status of the image processing of partial differential equations.In the second chapter, the digital image processing technology is divided into three parts. In the first part, we introduce the concept and the classification of image and the mathematical expression of the operator related to image in details. In the second part, we introduce the concept and the classification of noise and some of the classical denoising methods, such as median filtering, mean filtering and so on. And the corresponding images are given in this part. In the third part, the paper gives some important quantitative indexes of evaluation standard of image denoising.In the third chapter, we mainly introduce PDE image processing technology. The chapter is divided into two parts. In the first part, we introduce three important concepts on denoising model related to PDE, namely gradient, divergence, edge detection. In the second part, basing on PDE, we introduce a classical denoising algorithm ROF model. Some numerical experiments of ROF model are presented in this part.The fourth chapter mainly introduces P-M denoising model and its improvement, which is the core part of this paper. This chapter is divided into two parts. The first part introduces the P-M denoising model in details. The second part mainly introduces the improvement on P-M denoising model.In the fifth chapter, we mainly use some numerical experiments to compare PSNR and running time of the two methods. From the analysis of the data, we can see when PSNR is almost the same, the running time of the improved P-M model is shorter. It shows that the improved P-M model not only holds good denoising ability, but also has a fast convergence speed.