Digital watermark based echo inhibition method and system

Abstract

The invention discloses a digit watermark based echo inhibition method and system. The method comprises that first audio information is calculated according to obtained far-end audios and pre-generated watermark information, and played via an audio play unit; an audio collection unit collects audios to obtain second audio information; a characteristic sequence is extracted from the second audio information; the correlation degree between temperature audio watermark information and the characteristic sequence is calculated; whether the second audio information includes an echo is determined according to the correlation degree; and if the second audio information is determined to be the echo, the second audio information is zeroed. Echo inhibition is carried out on the watermark information on the basis of frequency domain and discrete logarithm transformation, the audio quality is not decreased, high robustness is provided for audio loss, a watermark can be detected normally under relatively low embedding intensity, and the echo can be detected more accurately.

Description

technical field [0001] The invention relates to the field of voice communication, in particular to a digital watermark-based echo suppression method and system. Background technique [0002] Since the appearance of voice communication, acoustic echo cancellation technology has emerged at the historic moment and has received extensive attention. Most of the current acoustic echo cancellation technologies are based on an adaptive filter, which is a process of simulating acoustic echo parameters through an LMS algorithm (Least mean square, minimum mean square error algorithm). However, the filter method has the following defects. First, in order to achieve a better echo cancellation effect, people often use longer parameter sequences to simulate the real echo environment, resulting in an increase in computational complexity and a high demand for storage space. Considering that the process of acoustic echo cancellation needs to be carried out with very low hardware computing an...

AI Technical Summary

Problems solved by technology

However, the filter method has the following defects. First, in order to achieve a better echo cancellation effect, people often use longer parameter sequences to simulate the real echo environment, resulting in an increase in computational complexity and a high demand for storage space. Considering that the process of acoustic echo cancellation needs to be carried out with very low hardware computing and storage capacity in many cases, for example, in a conference call, the sound of the speaker is repeatedly fed back to the microphone, causing the sound of the speaker to be superimposed , resulting in echoes in the acoustics, leading to a decrease in communication quality, so it is particularly important to simplify the process of echo cancellation

However, the application of watermark embedding in echo cancellation has not been found so far in the prior art

Images