Image Denoising Using a Wavelet-Based VariationalBayesian Algorithm

Many good techniques have been discussed for image denoising that include improved adaptive wavelet denoising method based on neighboring coefficients (IAWDMBNC), improved wavelet shrinkage technique for image denoising (IWST), local adaptive wiener filter (LAWF) wavelet packet thresholding using median and wiener filters (WPTMWF), adaptive image denoising method based on thresholding (AIDMT), adaptive thresholding These techniques are based on local statistical description of the neighboring coefficients in a window. These methods however do not give good quality of the images since they cannot modify and remove too many small wavelet coefficients simultaneously due to the threshold. In this paper, the wavelet- based variational Bayesian estimation theory for image estimation theory for image signals such as images can signals such as images can their wavelet coefficients In this method, we suppose the mixture of normal matrices distribution over the noisy wavelet coefficients and the variational Bayesian Expectation Maximization VBEM algorithm is implemented on the wavelet coefficients distribution Our method overcomes these drawbacks and it has better performance than the NeighShrink IAWDMBNC IAWDMBNC hresholding methods.

Keywords:

image denarsing , waveter transform , mixtue of normal matrices distribution , variational baysesian inference