Stationary wavelet denoising based on wavelet coefficients obeying prior distribution in subbands
Fengjun Zhang, Xie Chengjun, Jianhui Yin, Zhou Zhiqiang · 2011
Edge distortion in image denoising can be improved effectively based on Stationary Wavelet Transformation (SWT) instead of Orthogonal Wavelet Transformation (OWT). The statistics of wavelet coefficients in subbands can be supposed obeying a Prior Distribution Model (PDM), through which the threshold function with ideal shrinkage can be deduced and improve image denoising in the smooth regions effectively. The method based on SWT combining with PDM proposed in this paper can improve both Peak Signal to Noise Ratio (PSNR) and visual effect of the image denoised through Matlab simulation experiments.