Single threshold denoising based on wavelet transform is effective in certain cases as fixed models are used for threshold estimation

but it is often unsatisfactory in some specific projects.This paper presents a wavelet transform-based mixed threshold denoising method

in which weight coefficients are added to different thresholds and fairly ideal threshold are obtained by optimizing the weight for the object function.The simulation results show that compared with the conventional methods

this method better in denoising effect and maintains the detailed features.