FPGA-based effective breathing sound removing method and system

Abstract

The invention relates to the technical field of signal processing. The invention discloses an FPGA-based effective breathing sound removing method and system, and the method comprises the steps: firstly calculating the feature values of a voice signal and a breath sound signal, then calculating the difference between the two groups of feature values, judging whether a current frame is a breath sound frame or not through two times of judgment, removing the current frame if the current frame is the breath sound frame, and otherwise, not carrying out the processing. The method is based on an FPGAhardware platform, is simple in process, easy to implement and high in reliability, and can remove most breathing sounds under the condition that the call quality is not affected. And meanwhile, thebreath sound in the speaking process can be effectively removed, and the voice quality is not influenced. According to the invention, the breathing sound can be effectively removed through the FPGA only by ensuring normal communication transmission without adding extra equipment.

Description

technical field [0001] The invention relates to the technical field of signal processing, in particular to an FPGA-based method and system for effectively removing breathing sounds. Background technique [0002] With the rapid development of modern communication transmission, the speed and content of communication transmission have been greatly improved and enriched. Voice signal transmission has always played an important role in communication transmission due to its directness and simplicity. Voice signal transmission transmits the sender's voice information directly to the receiver's earphone or other sound-producing equipment through a wired or wireless transmission channel. During the transmission process, there is no need to do other processing on the voice signal, just to ensure the normal and efficient communication transmission, and the content of the transmitted signal does not need to be known, which leads to the microphone sending out all the collected voice sig...

AI Technical Summary

Problems solved by technology

In other words, the conventional method of finding the eigenvalues ​​of breath sounds in the time domain and frequency domain cannot achieve a good removal effect

Due to the importance of the voice, the eigenvalues ​​of the breathing sound need to be clearly distinguished from the voice, and the voice cannot be affected while removing the breathing sound. If a part of the voice is accidentally removed during the call, it will seriously affect the quality of the entire call

Images