Study On Algorithms Of Image Denoising And Compression Based On Wavelet Transforms

Multimedia communication is a new synthetical technology, which is rapidly developing rencently.Image processing is an important technique in multimedia communication because of its strong ability of reducing the bit-rates of image signals and improving the quality of image communication.Wavelet analysis is a novel research field in the world. It has a perfect performance on both local time and frequency domain, so it is an effective analytical tool .Digital image processing by the aid of wavelet analysis is one of the hot issues of wavelet research and application. After a brief description of the basic theory of wavelet, this paper makes a deep researeh on image denoising and compression.The existence of image noise severely affects the effect of image processing, so image denoising is beneficial to subsequent image processing. The paper researchs and analysises the method of image denoising using wavelet transform.A new method for thresholding estimation is presented in this paper .The new thresholding that increases a variable on the basis of BayesShrink can effectively use the relation of wavelet coefficients and get high PSNR comparing to BayesShrink.An improved shrink function on the basis of soft shrink function and hard shrink function is presented in this paper. It gets better denoising effect.Image compression benefits to the transmission and storage of image. A deep researeh on image denoising and compression using wavelet transform is presented in this Paper. This paper analysises the merits and flaws of the EZW and SPIHT. An improved algorithm is presented on the basis of SPIHT. It revises the SPIHT's sorting strategy, enhances the code speed, induces the memory and improves the recovery image quality. This paper presents a code algorithm using many threshold for image with noise.Comparing to EZW algorithm, the new algorithm has better effect on denoising and compression.