Voice noise reduction method based on spectral subtraction and wavelet transform

Abstract

The invention provides a voice noise reduction method based on spectral subtraction and wavelet transform. The method comprises the following steps of removing a direct current component of an input pure voice signal; performing normalization processing on an amplitude value; superimposing white gaussian noise on the processed voice signal so as to obtain the voice signal with noise; setting a leading no-word section; calculating the frame number of the leading no-word section; processing the voice signal with noise by spectral subtraction; performing wavelet transform on the obtained voice signal subjected to spectral subtraction to obtain a wavelet signal with a discrete telescopic factor and a discrete translation factor; according to a low-frequency component in the wavelet signal anda high-frequency component of each decomposition layer, obtaining a low-frequency coefficient and a high-frequency coefficient of each decomposition layer; respectively processing the high-frequency coefficient of each decomposition layer by a zero setting processing method and a soft threshold processing method so as to obtain the processed high-frequency coefficient of each decomposition layer;and performing inverse wavelet transform to obtain a final noise reduction voice signal. The music noise can be obviously reduced; in addition, only little high-frequency distortion exists; and the noise reduction capability is further improved.

Description

technical field [0001] The invention relates to a voice noise reduction method. In particular, it relates to a speech noise reduction method based on spectral subtraction and wavelet transform. Background technique [0002] Voice communication is often affected by various noises, resulting in a decline in communication quality. The main purpose of speech noise reduction is to extract the original speech as pure as possible from the noisy speech and improve the recognition rate. Speech enhancement has always been a research hotspot, and new methods are constantly emerging. At present, the common speech noise reduction methods are speech noise reduction based on autocorrelation method, speech noise reduction based on spectral subtraction, speech noise reduction based on wavelet transform, speech noise reduction based on adaptive filtering and Wiener filtering method. Voice noise reduction, etc. Among them, although the speech noise reduction based on the autocorrelation me...

AI Technical Summary

Problems solved by technology

Among them, although the speech noise reduction based on the autocorrelation method can concentrate the energy of the speech signal in a small number of formants, and maximize the signal-to-noise ratio in the frequency domain, it is more beneficial to the noise removal at the same frequency as the speech, but Its denoising threshold is difficult to determine, which will cause damage to speech

Although the speech signal after noise reduction based on spectral subtraction has almost no distortion, there is obvious "music noise" after noise reduction, and the signal after noise reduction based on wavelet transform has serious high-frequency distortion

Images