Study On Algorithm Of Image Denoising Based On Wavelet Transform And Partial Differential Equations

Image processing and analysis is a major field of information science and engineering.The image restoration is an item basic moreover the important technology in the image processing domain.The image denoising is the earliest and most research in the image restoration domain, it is the essential step and the key link in the process of image processing. Now,Wavelet and Partial Differential Equation(PDE) are main denoising methods in image processing aspect.Wavelet can analyse locally time-frequency domain and extract characteristics of the signal singularity flexibilitily in the same time,so it can filter out noise and keep high-frequency information of the signal.Its disadvantage is that there will be some dumbbell effect.But, image denoising technology based on PDE is an adaptive denoising method,which deal with the whole of an image. The denoising results can balance between smoothing image noises and preserving edges well. The disadvantage is that the texture information of image will be blurred.Combining with these two methods,it will overcome drawbacks with each other and result in better effect.A new image denoising method based on the theory of wavelet thresholding denoising is presented.We use this algorithm to denoise initially.Base on it,this paper researches the relations between wavelet and Partial Differential Equation in the image denoising.A new thresholding denoising method is then obtained.It can receive higher SNR than other methods.At last,according to denoising characteristic of wavelet and PDEs,a mixed method of combining wavelet with PDE for image denoising is proposed.It is indicated from the experimental results that new method can attain the purpose of preserving edge and smoothing noise at the same time.