Double-microphone voice noise reduction device and noise reduction method

Abstract

The invention discloses a double-microphone voice noise reduction device. The device comprises a first microphone and a second microphone, wherein the first microphone and the second microphone are located on the same horizontal line, the first microphone is set to be in the 0-degree direction of the mouth, the second microphone is far away from the 180-degree direction of the mouth, and the distance between the first microphone and the second microphone ranges from 1.1cm to 3cm. The invention also discloses a double-microphone voice noise reduction method, which comprises the steps of firstly, receiving and acquiring two paths of voice signals s0 and s1, wherein the sampling rate is integral multiple of the sampling rate of actually processed data; then down-sampling to an actual samplingrate needing to be processed; performing spectral analysis on the down-sampled data, and iteratively calculating a filter coefficient; multiplying the filter by one path of input signal in a frequency domain, and performing inverse Fourier transform and overlapping addition to obtain an output voice signal s2. According to the invention, stable and non-stable voices in a non-target direction canbe effectively filtered to obtain processed clean voices; not only can an efficient noise reduction effect be achieved, but also the noise reduction cost is reduced.

Description

technical field [0001] The invention relates to the technical field of speech noise reduction, in particular to a dual-microphone speech noise reduction device and a noise reduction method thereof. Background technique [0002] Dual microphone noise reduction technology refers to the use of two microphones as the receiving unit, through a series of signal processing, to complete the signal enhancement in a given direction and the attenuation of speech signals in non-receiving directions and environmental noise, so as to enhance the reception of speech in a given direction The purpose of the signal-to-noise ratio. [0003] In the prior art, one technique is to use the main microphone to receive the desired voice and background noise, and the secondary microphone to receive the background noise, and assume that the background noise received by the two microphones is consistent, and use the noise signal received by the secondary microphone as a reference Signal, carry out digi...

AI Technical Summary

Problems solved by technology

However, the beamforming composed of dual mics in this scheme has a main lobe that is too wide, and there are certain requirements for the distance between the dual mics.

Images