Sound processing device and method for intelligently suppressing noise, terminal equipment and readable medium

Abstract

The invention discloses a sound processing device capable of intelligently suppressing noise, and the device comprises a sound input module, a noise parameter set module, a noise processing module anda sound output module. The sound input module is used for receiving a total signal in a current environment, and the total signal comprises a useful signal, steady-state noise and transient noise. The noise parameter set module is used for storing specific types of noise parameters. The noise processing module is used for matching the transient noise with a specific type of noise and suppressingthe transient noise and the steady-state noise to obtain a suppressed useful signal; and the sound output module is used for outputting the suppressed useful signal. The transient noise is matched with the specific type of noise in real time through the noise processing module. The transient noise is suppressed; meanwhile, the suppressed useful signal can be obtained by suppressing the steady-state noise, the noise reduction function of the transient noise and the steady-state noise is achieved, the sound processing device for intelligently suppressing the noise is applied to different systems, and suppression of a certain kind of transient noise or suppression of several kinds of transient noise can be started or stopped through system setting.

Description

technical field [0001] The present invention relates to the technical field of audio signal processing, in particular to a sound processing device for intelligent noise suppression, a sound processing method for intelligent noise suppression, a sound processing terminal device for intelligent noise suppression, and a sound processing device for intelligent noise suppression computer readable media. Background technique [0002] Existing audio communication systems generally have a noise cancellation function, such as mobile phones or intercom terminals. When talking in a noisy environment, the ideal noise cancellation module will cancel the noise signal and retain the human voice signal, but the actual effect is that only It can suppress steady-state (stable, small-change) noise, but cannot suppress occasional, large-change or short-lived transient noise, and it cannot be set to eliminate certain noise while retaining some other noise signals. The general denoising method m...

AI Technical Summary

Problems solved by technology

The disadvantage of this type of method is that it is generally only effective for steady-state noise, and cannot perform noise reduction on specific noise, and the more noise reduction, the greater the impact on human voice, and the voice quality is not high.

[0003] In addition to the steady-state noise suppression method, there are also some transient suppression methods, such as webrtc's transient noise suppression method, which is only for this kind of noise similar to the keyboard sound, by judging the amplitude of different frequency bands of the sound signal in the corresponding time frames in a short period of time Whether there is a large change to judge whether it contains transient noise, so as to estimate the amplitude and achieve the purpose of suppressing transient noise. This noise feature is very single, so it can achieve certain effects, but the transient noise in the environment is often At the same time, in the actual application process, users cannot choose to suppress or not suppress certain types of noise. A sound processing device, method, terminal device, and readable medium for intelligently suppressing noise that can freely choose the transient noise function that needs to be suppressed or not suppressed

Images