Using stationary wavelet transformation for signal denoising

Abstract

Because singular points were existed in the signal, the Pesudo-Gibbs phenomenon would produce in the singular points when the traditional wavelet threshold value algorithm was used for signal denoising. The threshold denoising algorithm based on the stationary wavelet transformation may be possible to suppress the Pesudo-Gibbs phenomenon effectively, because the staionary wavelet transformation is proposed on the foundation of orthogonal wavelet transformation, which possess the properties of rotation, shift and scale invariance. Denoising method based on the stationary wavelet transformation need to carry on the multi-layered wavelet decomposition to the signal firstly, then carries on thresholding processing to the high frequency coefficients, finally realizes wavelet reconstruction to achieve the denoising goal. The threshold value function uses half soft threshold value which is combination of soft and hard threshold value. The simulation experiment results indicated: denoising method based on the stationary wavelet transformation can enhance the signal-to-noise ratio obviously, its denoising effects is better than soft and hard threshold value method, has higher use value.