Adaptive image denoising by exploiting wavelet intra-scale dependency
Chongzhao Han · Optical Technique · 2007
The method of wavelet thresholding for image denoising has been researched extensively due to its effectiveness and simplicity.Spatially adaptivity can improve the wavelet thresholding performance because it makes threshold value adaptive to the spatially changing statistics of images.After analyzing the defects of current methods,a new spatially adaptive coefficient model is proposed.Each wavelet coefficient is modeled as a random variable of a generalized Gaussian distribution with an unknown parameter,and the modeling is used to estimate the parameter for each coefficient.The method provides a good indication of local variability.Experiments show that higher peak-signal-to-noise ratio and better visual effect can be obtained as compared to other methods.