Denoising Based On Wavelet Transform Image Of The Partial Differential Equations

Noise is always present in image acquisition and transmission which affected visual effects of the image and follow-up image processing seriously. Wavelet Transform and Partial Differential Equations(PDE) are two of the most popular denoising methods in recent years. In this paper, window improvements for Wiener filtering of the wavelet domain have been made, proposed a PDE integrated model, and then combined them by using a algorithm hybrid.This paper introduces assessment criteria of image quality and several traditional image denoising methods.Then the author introduces the basic theory and basic algorithm of wavelet transform to analyze and compare advantages and disadvantages of the wavelet threshold denoising and Wiener filtering. Based on the merits of this paper, the windows improved wiener filtering were proposed, from which a good noise reduction can be found.Then, the basic concept and image processing algorithms of PDE were introduced and formed the comprehensive denoising model by analyzing a variety of PDE's denoising model. Based on the characteristics of the P-M equation, the directed diffusion equation and LLT high-order equation, two weight functions can be used to build the combined denoising model. This model combines advantages of the second and fourth order of PDE model, inhibits disadvantages of each model, enhances edge detail preservation effectively, and a better denoising effect was got after the experiment.Finally, the author combines a hybrid algorithm to window improved Wiener filtering and the PDE combined model which was implicated in the details of gray-scale images. the mixed denoising method not only can overcome the drawbacks of three PDE models and wiener filtering, but also can obtain a better visual effect and enhance efficiency of the algorithm.