On optimal multiwavelet shrinkage
Tai-Chiu Hsung, Daniel Pak-Kong Lun, K. C. Ho · 2005
Recent studies show that multiwavelet approaches give consistently better performance over traditional wavelet approaches in application such as denoising. It is because multiwavelet transform, with appropriate initialization, provides better representation of signals so that their difference from noise can be clearly identified. Traditional multiwavelet denoising only considers multiwavelets of multiplicity 2. In this paper, we first suggest a simple 2nd order orthogonal prefilter design method for multiwavelets of higher multiplicity. Then, we study how to select optimal threshold for each resolution level. Simulation results show that higher multiplicity wavelets usually give better denoising results and the proposed threshold estimator suggests good indication for optimal thresholds.