Study On Denoising Algorithms Based On Wavelet And Fractal Coding

In the process of image storage and transmission,it is usually disturbed by different noises,which leads to image distortion and blurring.In order to obtain valuable information from noise-contaminated images and improve the quality of images,it is necessary to denoise the noisy images,and the denoising effect of the noisy images will directly affect the results of subsequent image processing.This paper mainly studies the image denoising algorithms of wavelet packet,partial differential equation and fractal code.Many methods are improved and combined,and a new denoising method is proposed.The results show that the new method retains the complete details and edge information and effectively removes noise.The main work includes the following aspects:1.The research status,research background and significance of image denoising at home and abroad are introduced in detail.The image denoising algorithm based on wavelet,partial differential equation and fractal image coding is deeply analyzed.2.The principle of wavelet packet and partial differential equation is studied and analyzed.Aiming at the problem that the edge texture information is easily polished after denoising by most partial differential equation denoising algorithms,especially the gray gradient area and the light edge of image,which can not effectively identify the damaged image,the excellent performance of gradient control by using the more precise signal decomposition ability of wavelet packet and partial differential equation is proposed.A new image denoising algorithm based on wavelet packet and partial differential equation is proposed to protect edge texture and other details,while taking into account the performance of denoising.3.The image denoising algorithm based on fractal code is deeply analyzed and studied,and an effective image denoising method based on fractal code is proposed.Hierarchical classification is used to improve the coding speed and avoid a large number of invalid mean square error(MSE)calculations.Quadtree-based image partitioning using dynamic range and domain sizes is used to increase the degree of noise removal.Further denoising is achieved using pyramidal decoding,using nonarbitrary seed image,and additional post processing.