Research on adaptive image denoising based on wavelet transform

An effective method based on wavelet transform is introduced in This work for image denoising without blurring the useful edge information. Wavelet shrinkage at consecutive scales are utilized on the sub-images exerted wavelet decomposition, meanwhile a statistical model is referenced to determine the proper shrinkage functions and threshold for discriminate the edge information from that of noise. Finally, anisotropic diffusion equation is applied to the modified wavelet coefficients to preserve edges information that is not isolated. This method is of adaptability to different amounts of noise in the image, and robustness to larger noise contamination. Simulation results present a superior performance in the aspect of image denoising.