Motion Adaptive Video Denoising in the Wavelet Domain Based on Bivariate Shrinkage

This paper proposes a new wavelet-based video noise reduction algorithm. Each frame in the video sequence is transformed to the wavelet domain using the dual-tree complex wavelet transform. We model the interscale dependencies in subband representation of each frame using a non-Gaussian bivariate distribution. The parameters for this bivariate distribution are estimated adaptively using the local correlations that exist between neighboring coefficients within each subband. Based on this bivariate distribution a shrinkage function is developed using the maximum a posteriori (MAP) rule. To improve the performance of the filter, information from the adjacent frames is also incorporated in the shrinkage function. This is achieved by detecting the motion between the corresponding subband coefficients in successive frames. Experimental results show that the proposed scheme outperforms several state-of-the-art spatio-temporal filters in terms of the peak signal to noise ratio and the visual quality.