Echo noise reduction method through multi-dimension technology

Abstract

The invention discloses an echo noise reduction method through the multi-dimension technology. The realization process of the echo noise reduction method comprises the steps that the sound source azimuth is located according to the number of audio acquisition equipment, wherein the audio acquisition equipment refers to at least two acquisition devices; the transmission process of sound waves is simulated according to the geometrical characteristics of a multidimensional space, and echo data transmitted in the delayed way are processed by using the multi-dimension technology; clutter signals are intelligently filtered through the geometrical characteristics, and synthetic processing of audio signals is accurately controlled; and the synthesized audio data are exported. Compared with the methods in the prior art, the special requirements for the acquisition equipment are not required in the whole noise reduction process so that the echo noise reduction method through the multi-dimension technology has the advantages of being low in cost, easy and convenient to arrange and low in the requirement for the professional level of users in comparison with professional audio acquisition equipment; and the method is high in practicality and wide in range of application so that the technology can be widely applied to VoIP products of various platforms and plays an important role in enhancing conversation quality and service quality and thus the method is easy to popularize.

Description

technical field [0001] The invention relates to the technical field of noise processing, in particular to a highly practical echo noise reduction method using multi-dimensional technology. Background technique [0002] In the traditional audio noise reduction technology, it is usually necessary to use program algorithms to compare and remove noise data in the audio. However, similar methods have an obvious defect, that is, the removal of noise is often not complete, or Will reduce the quality of the session audio. [0003] This is mainly caused by the basic principle of audio noise reduction processing, as shown in the attached figure 1 As shown, due to the inherent characteristics of traditional audio files, the source of the sound and the voiceprint characteristics of the speaker cannot be well distinguished. The collected sound can only be analyzed by simultaneously collecting and recording the audio data of the remote / local speaker. For comparison, the audio data of th...

AI Technical Summary

Problems solved by technology

[0002] In the traditional audio noise reduction technology, it is usually necessary to use program algorithms to compare and remove noise data in the audio. However, similar methods have an obvious defect, that is, the removal of noise is often not complete, or will reduce the quality of the session audio

[0003] This is mainly caused by the basic principle of audio noise reduction processing, as shown in the attached figure 1 As shown, due to the inherent characteristics of traditional audio files, the source of the sound and the voiceprint characteristics of the speaker cannot be well distinguished. The collected sound can only be analyzed by simultaneously collecting and recording the audio data of the remote / local speaker. For comparison, the audio data of the far-end speaker is eliminated from the returned audio data, so that the far-end speaker can not hear his own voice

However, in such a process, due to the complexity of the actual environment, the processing of audio data is often asynchronous, which causes a delay in the processing of video data and makes the returned audio signal have a certain degree of echo drag. tail phenomenon

Moreover, since the audio waveform data cannot be separated accurately, only a simple clipping process can be performed on the waveform, which causes the integrity of the audio data to be damaged to a certain extent, which will inevitably cause the loss of sound or call decline in quality

Images