A Novel Image Correlated Noise Reduction Method Based on Multiwavelet Transformation
Xiaoguang Di, Fengqi Zhou, Yu Feng Yao, Shaowen Fu · 2006
A novel stationary multiwavelet transform (SMWT) method is proposed. The method possesses the good quality of both multiwavelet transform and stationary scalar wavelet transform on image denoising. Being redundant, the SMWT can decorrelate the image and noise. The mallat decomposition and reconstruction of 2D image based on SMWT was inferred. By level-dependent thresholding the SMWT coefficients of image with correlated noise, the image with high quality could be reconstructed. The simulation results show the effect of correlated noise reduction based on this method has obvious superiority compared with scalar wavelet and general multiwavelet transformation method. At the same time, the recovery image can preserve as many the characteristics of original image as possible.