Improved wavelet threshold function denoising method based on Teager energy operator

Abstract

The invention claims for protecting an improved wavelet threshold function denoising method based on a Teager energy operator, relating to the field of voice signal denoising. According to the method,for a wavelet coefficient after wavelet decomposition, the calculation for the Teager energy operator is carried out, the differences between noises and the signal wavelet coefficient are increased,and thus the self-adaption selection for the threshold is facilitated; by improving a soft and hard threshold denoising function, a novel threshold function is provided, the function can overcome thePseudo-Gibbs effect caused by the discontinuity of the hard threshold function, and also can overcome the constant deviation problem occurring after soft threshold function denoising. With the methodprovided by the invention, the signal to noise ratio and the mean square error are greatly improved.

Description

technical field [0001] The invention belongs to the field of voice signal denoising, in particular to an improved wavelet threshold function denoising method based on a Teager energy operator. Background technique [0002] Speech signals usually contain noise in the process of acquisition or transmission, which seriously affects the follow-up work of signal processing and analysis. Therefore, the denoising of speech signals has become the most basic and important work in the field of signal processing. [0003] Traditional denoising methods mainly include linear filtering and nonlinear filtering, such as median filtering, Wiener filtering and Kalman filtering. The disadvantage of these methods is that it is difficult to reflect the non-stationary characteristics and correlation of the signal while removing the noise. In recent years, wavelet transform has been widely used in signal denoising processing due to its multi-resolution characteristics. It can show good signal lo...

AI Technical Summary

Problems solved by technology

The disadvantage of these methods is that it is difficult to reflect the non-stationary characteristics and correlation of the signal while removing the noise.

The hard threshold and soft threshold denoising methods proposed by Donoho and Johnstone have been widely used in practice and have achieved good results, but the discontinuity of the hard threshold function makes the reconstructed signal prone to false Gibbs effects; Although the soft threshold function has better continuity, there is always a constant deviation between the estimated value and the actual value

If the threshold is too small, there will be insufficient denoising and some noise will be retained; if the threshold is too large, excessive denoising will occur, and the weak feature components in the signal will be mistaken for noise and eliminated, resulting in the loss of some useful information in the signal

Images