Gaussian Noise And Salt And Pepper Noise Image Denoising Algorithms

Image denoising is a hot research topic in the field of digital image processing, and a lot of models and algorithms of denoising have been proposed. In human history, a testimony to their era image denoising has such a great influence on modern society, science and technology. With the development of computer technology, image denoising problem has been attracted more and more attention. Denoising methods are divided into the following three categories:methods based on probability and statistics model, methods based on variational and partial differential equations and methods based on the Fourier and wavelet transform. According to the relationship between the image and the noise, the noise model is mainly divided into additive noise and multiplicative noise. In this paper, methods based on probability and statistics model and the variational were used to remove the additive noise in image.The first chapter introduces the background and classification of noise model for research in the field of image denoising. At the same time, the subjective and objective criteria of evaluation of the denoising is given. Brief introduction to the field of denoising methods and theories was given, as well as the noise classification categories based on different standards. In this paper.research we just pay attention to Gaussian noise and salt and pepper noise.The second chapter describes the sparse coding image denoising algorithm based on polynomial fitting. Sparse coding theory and its application was given. A brief review of the sparse coding algorithm for image denoising was given, and we pointed out its shortcomings. The simulation results show the effectiveness of the proposed algorithm.The third chapter describes the denoising algorithm based on partial differential equations and the variational method. The method uses a continuous domain of thinking to deal with discrete digital image. First, establish the energy function to denoising the image. Second, achieve the purpose of denoising by minimizing the energy function. Simulation results show that the method has good performance in the removal of low-density Salt and Pepper Noise. However when the density of image noise increase, the denoising performance decrease significantly.The fourth chapter describes the improved variational method based on adaptive median filtering algorithm. In this chapter, a new method of two-phase removal of salt and pepper noise was proposed based on the improved adaptive median filter and variational filtering. In the first phase, an improved adaptive median filter is used to identify pixels which are likely to be contaminated by noise (noise candidates). In the second phase, the image is restored using a variational method that applies only to those selected noise candidates. This two-stage approach can avoid the shortcomings of the improved adaptive filtering and variational filtering, and give full play to their strengths. Simulation results show that this algorithm is effective and better than traditional variational method based on adaptive median filter.The fifth chapter summarizes the denoising algorithm described above and the future direction of research to put forward.