Block Adaptive Bayesian Wavelet Shrinkage for 2D Signal De-noising

A block adaptive Bayesian wavelet shrinkage is proposed in this paper to accommodate the reduction of a kind of two-dimensionally, signal amplitude-related contamination, which is taken place in some military and biomedical applications. Wavelet shrinkage conventionally works under the assumption of signal-independent, additive Gaussian noise. For the temporally Gaussian but spatially signal amplitude-related noise, its denoising efficiency is depressed. To make use of the merit of wavelet denoising, the signal space is split to blocks and then wavelet shrinkage is conducted in each block, in which the signal amplitude is assumed to vary little. Results from simulation show that the proposed method outperforms the traditional one in signal-to-noise ratio improvement.