Multi Threshold Image Denoising Method Based On Dyadic Wavelet

In practical applications,the existence of noise in the image is still inevitable and must be solved.We all hope that the important features of the original image can be well retained after denoising,and the visual effect can meet the practical needs,which is crucial to our daily production and life.For the above purposes,based on the threshold image denoising method proposed by Donoho et al.,a multi-threshold denoising algorithm based on dyadic wavelet is proposed in this paper.In view of the application defects of traditional soft and hard threshold functions,a new threshold function specifically for dyadic wavelet à trous algorithm is proposed.The threshold function innovatively introduces the trigonometric function,which makes the threshold function more flexible and brings greater convenience to the image denoising process.In some existing methods,the difference of coefficients at different scales is ignored and the same threshold is used for all coefficients.In view of this defect,an improved form of universal threshold is given in this paper.The threshold decreases with the increase of scale in the improved form.This change makes multiple thresholds used in the denoising process,which well caters to the feature that the larger the scale of the wavelet coefficient is,the smaller the coefficient is.In addition,in order to make the denoising effect better,two dyadic wavelet filters with high-order vanishing moments are constructed by using the dyadic lifting scheme.It is found that the filter with higher vanishing moment has better overall image processing effect.In order to analyze the denoising performance of this method,the orthogonal wavelet filter and the newly constructed dyadic wavelet filter with high-order vanishing moments are used to conduct comparative simulation experiments on different images superimposed with different levels of noise.From the subjective and objective aspects,it can be found that compared with the traditional threshold denoising algorithm,the proposed algorithm not only improves the discontinuity of the hard threshold function at the threshold.Moreover,it overcomes the shortcoming of constant deviation between the original coefficients of soft threshold model and the coefficients after threshold processing.From the results,it can be found that the threshold function corresponding to different algorithms can make the experimental effect better,and the same filter has different denoising effects on different images.Through the experimental simulation,it can be seen that the proposed method has higher signal-to-noise ratio and better visual effect after image denoising,which has better practicability and is an effective image denoising method.