Research On Image Denoising Based On Wavelet Theory And Particle Swarm Optimization

Digital signal filtering has always been a hot topic in information processing,Especially in the past 50 years,a variety of new theories,new methods put forward,making information processing one of the fastest growing subjects in the twentieth Century.Wavelet analysis theory is the most effective signal processing tool to deal with signals at the present stage.This paper focuses on nonlinear threshold function denoising in wavelet domain.The image denoising method based on wavelet analysis is one of the most commonly used methods in image denoising.The wavelet transform is used to decompose the image,and the singularities of the effective information and the interference information in the wavelet domain have been known,the effective information is medium low frequency,while the interference information is high frequency.According to this characteristic,the threshold is often set at the disturbance information to judge the wavelet coefficients,Information that is greater than the threshold is retained,and less than the threshold cleared or other ways to deal with.A new multilayered threshold function is constructed,the fixed estimation threshold is chosen as the threshold criterion,and the optimization is improved,the problem that the fixed estimation threshold is too large is improved.An adaptive image denoising algorithm is realized by combining the particle swarm optimization algorithm with the new multilayered threshold function.By choosing the different international standard test images,different types and the variance of the noise as well as more than two kinds of mixed noise simulation of subjective and objective data by comparing the threshold function with the method proposed by other researchers,the presented method is effective in noise on the image.