Denoising Algorithm Of Digital Image Mixing

In the field of digital image processing, due to various disturbances, the image will inevitably suffer a variety of noise pollution. Digital image denoising is necessary in the stage of preprocess because noise destroyed the original information of image, it can bring a great impact in the following image process.Image noise is mixed noise which basically make up by salt and pepper noise and gaussian noise. Most of the existing denoising algorithm for both noise.In this paper, based on the analysis of the histogram and mathematical model of this two kind of noise For the salt and pepper noise. Based on the idea of traditional denoising model that first judge then filter. In the side of noise judgement. Analyse some classical noise judgement, Such as, Switching threshold, extreme value judgement and bipolar threshold detection.In the side of noise fliter. Analyse the advantages and disadvantage of some algorithms, take median filter, removing extreme value median and filter direction median filter for example. direction median filter have a good ability to judge the direction. But also have shortcomings of lack of ability to judge the details. Finally, a fuzzy reasoning-based direction median filter is proposed.For the gaussian noise. Study several models of partial differential equations. Such as heat conduction equation, P-M equation and direction of diffusion equation. Analyse the advantages and disadvantage of this model. Direction of diffusion equation with edge stopping function have good edge protection. But also have some shortcomings. For these shortcomings, this paper improve and propose a new model of the direction of diffusion.In this last, based on fuzzy reasoning-based direction median filter of salt and pepper filter algorithm and improved direction of diffusion partial differential algorithm, a new hybrid noise filtering algorithm is proposed.This paper compare the simulation experiments and data by Matlab of this new algorithm, experimental results show the improved algorithm is better than the previous algorithm.