Shearlet-based image denoising using adaptive thresholding and non-local means filter

This paper presents an efficient image denoising algorithm by incorporating shearlet-based adaptive thresholding with non-local means filter. Firstly, an adaptive Bayesian maximum a posteriori estimator, where the normal inverse Gaussian distribution is used as the prior model of shearlet coefficients, is introduced for removing the Gaussian noise from corrupted image. Secondly, the nonlocal means filter is used to suppress unwanted nonsmooth artifacts caused by the shearlet transform and shrinkage. Numerical experiments demonstrate that the proposed method achieves state-of-art performance than several published methods in terms of subjective and objective evaluations.