WO2013097192A1

WO2013097192A1 - Method for detecting disturbed audio signal, method for correcting same, and device therefor

Info

Publication number: WO2013097192A1
Application number: PCT/CN2011/085056
Authority: WO
Inventors: 梁锦泉; 徐仁德; 黄炳原; 楚斐韬
Original assignee: 宝添管理有限公司
Priority date: 2011-12-30
Filing date: 2011-12-30
Publication date: 2013-07-04
Also published as: TWI496140B; TW201327549A

Abstract

The present invention relates to a method for detecting a disturbed audio signal, comprising the following steps: providing an audio signal to be tested; determining whether the audio data of the audio signal to be tested is disturbed at each time point respectively. In particular, IX₀-X_avI-K*X_STD is calculated, wherein X₀ is the audio data at the time point to be determined, X_av is the average of the audio data at a time points before the time point to be determined, a is a natural number, K is a constant greater than zero, and X_STD is the standard deviation of the audio data at the a time points. If IX₀-X_avI-K*X_STD is greater than zero, then it is determined that the audio data at the time point to be determined is disturbed. The present invention also relates to a detecting device for a disturbed audio signal, a method for correcting the disturbed audio signal, and a device therefor. The present invention is low in complexity and consumes less computational resources.

Description

Method, method and device for detecting disturbed audio signal

[Technical Field]

The present invention relates to the field of acoustics, and in particular, to a method and apparatus for detecting an interfered audio signal, and to a method and apparatus for correcting an interfered audio signal.

【Background technique】

With the development of society and the advancement of science and technology, the use of wireless communication technology has become more and more frequent. However, during the transmission of the communication signal, it is susceptible to interference, which causes the audio signal to generate an error. The error signal that is more easily felt by the human ear is usually spike noise, as shown in Figure 1.

[Summary of the Invention]

Based on this, it is necessary to provide a method of detecting an interfered audio signal for correcting the detected interfered audio signal.

A method for detecting an interfered audio signal includes the following steps: providing an audio signal to be tested; respectively determining whether the audio data of the audio signal to be tested is interfered at each time point; specifically calculating IXo-J - K*X _STO , where Xo is the audio data of the time point to be judged, _v is the average value of the audio data at a time point before the time point to be judged, a is a natural number, and K is a constant greater than zero X _STO is the standard deviation of the audio data of the a time point. If IXo-J - K*X _{STO is} greater than zero, it is determined that the audio data of the time point to be determined is interfered.

Preferably, a is 64.

Preferably, the method further includes the following steps: performing cyclic redundancy check on the audio signal to be tested, and if the verification result is an error, and IXQ U - * ^ is greater than zero, determining the time to be determined Preferably, the audio data is a frequency value.

The above-mentioned method for detecting an interfered audio signal has a low complexity and consumes less computational resources.

It is also necessary to provide a method of correcting the interfered audio signal.

A method for correcting an interfered audio signal includes the following steps: Step 1: providing an audio signal to be tested; Step 2, respectively determining whether the audio data of the audio signal to be tested is interfered at each time point; Specifically, IXQ U - K*X _STO is calculated, where Χ ο is the audio data of the time point to be judged, _ν is the average value of the audio data at a time point before the time point to be determined, and a is a natural number, K Is a constant greater than zero, X _STO is the standard deviation of the audio data of the a time point, and if the IXQ U - K*X _{STO is} greater than zero, it is determined that the audio data of the time point to be determined is interfered; The compensated audio data is compensated based on the audio data that has not been interfered before the time point at which the interfered audio data is located.

Preferably, the third step is to compensate according to the following formula:

X^bX i + l— b) X ₂

Where Xo is the compensated interfered audio data, and b is a weight constant, which is the undisturbed audio data of the first two time points of the time point at which the interfered audio data is located.

Preferably, the third step is to perform mirror compensation.

Preferably, the method comprises the following steps: determining whether the interfered audio data is isolated or continuous; and the isolated audio data in the third step is compensated according to the following formula: =13⁄4 - _{1 +} (lb X− ₂ , where is the compensated audio data that is interfered, b is a weight constant, and is the undisturbed audio data of the first two time points of the time point at which the interfered audio data is located; In step 3, a plurality of consecutive interfered audio data are mirror compensated.

Preferably, the step of performing mirror compensation further comprises the step of smoothing the mirror-compensated audio data. Preferably, the method further comprises the step of performing volume reduction processing on the compensated audio data. Preferably, a is 64.

Preferably, the method further includes the following steps: performing cyclic redundancy check on the audio signal to be tested, and if the verification result is an error, and IXQ U - * ^ is greater than zero, determining the audio at the time point to be determined The data is indeed interfered. The third step is to compensate the audio data that is indeed interfered with based on the audio data that was not interfered before the time point at which the audio data was actually interfered.

Preferably, the audio data is a frequency value.

It is also necessary to provide a detection device for an interfered audio signal.

An apparatus for detecting an interfered audio signal, comprising: an audio signal acquisition module and a first interference determination module; the audio signal acquisition module is configured to acquire an audio signal to be tested; and the first interference determination module is configured to separately determine Whether the audio data of the audio signal to be tested is interfered at each time point; specifically, calculating IXQ U - K*X _STO , where Χ ο is audio data at a time point to be determined, and _ν is the time to be determined The average value of the audio data at a time point before the point, a is a natural number, K is a constant greater than zero, and X _STO is the standard deviation of the audio data of the a time points, if IXQ U - K*X _{STO is} greater than Zero determines that the audio data at the time point to be determined is interfered.

Preferably, the formatter and the second interference determination module are further configured to: perform a cyclic redundancy check on the audio signal to be tested acquired by the audio signal acquisition module; Determining a verification result of the cyclic redundancy check and a determination result of the first interference determination module, and if the verification result is an error, and ι ς _ν ι - K*X _{sto is} greater than zero, determining the to-be-determined The audio data at the point in time of the judgment is indeed disturbed.

It is also necessary to provide a correction device for an interfered audio signal

An apparatus for correcting an interfered audio signal, comprising: an audio signal acquisition module, a first interference determination module, and a compensation module; the audio signal acquisition module is configured to acquire an audio signal to be tested; Determining the audio signal to be tested separately at each time Whether the audio data of the inter-point is interfered; specifically, the IXQ U - K*X _STO is calculated, where Xo is the audio data of the time point to be judged, and _v is the audio of the a time point before the time point to be judged The average value of the data, a is a natural number, K is a constant greater than zero, X _STO is the standard deviation of the audio data at the a time points, and the time to be judged is determined if IXQ U - K*X _{STO is} greater than zero The audio data of the point is interfered with; the compensation module is configured to compensate the interfered audio data according to audio data that is not interfered before the time point at which the interfered audio data is located.

Preferably, the compensation module is a first compensation sub-module for compensating according to the following formula:

X^bX i + l— b) X ₂

Where Xo is the compensated interfered audio data, b is a weight constant, and X ₂ is the undisturbed audio data of the first two time points of the time point at which the interfered audio data is located.

Preferably, the compensation module is a second compensation sub-module for performing mirror compensation.

Preferably, the data type determining module is further configured to determine whether the interfered audio data is isolated or continuous; the compensation module includes a first compensation submodule and a second compensation submodule, and the first compensation The submodule is used to compensate for isolated, disturbed audio data according to the following formula: X. =bX- _{1 +} (lb)X- ₂ , where is the compensated interfered audio data, b is a weight constant, which is the first two time points of the time point at which the interfered audio data is located Disturbed audio data; the second compensation sub-module is configured to perform mirror compensation on a plurality of consecutive interfered audio data.

Preferably, a smoothing processing module is further included for smoothing the mirror-compensated audio data.

Preferably, the method further includes a volume processing module, configured to perform volume reduction processing on the compensated audio data.

Preferably, the method further includes a formatter and a second interference determination module; The audio signal to be tested acquired by the audio signal acquiring module performs cyclic redundancy check; the second interference determining module determines according to the verification result of the cyclic redundancy check and the determination result of the first interference determination module If the verification result is an error, and ι _{ν ν} ι - K*X _{sto is} greater than zero, it is determined that the audio data of the time point to be determined is indeed interfered; the compensation module is actually interfered with The audio data is compensated.

The method and apparatus for detecting the above-mentioned interfered audio signal, and the method and apparatus for correcting the interfered audio signal have low complexity and consume less computational resources.

[Description of the Drawings]

Figure 1 is a schematic diagram showing the occurrence of spike noise when the original audio signal is interfered;

2 is a flow chart of a method for detecting an interfered audio signal in an embodiment; FIG. 3 is a flow chart of a method for correcting an interfered audio signal in an embodiment; FIG. 4 is a diagram in which an audio signal detects spike noise and is removed. Schematic diagram

Figure 5 is a schematic diagram of compensation of a pair of interfered audio data by a method; Figure 6 is a schematic diagram of mirror compensation for continuous interfered audio data; Figure 7 is a schematic diagram of an audio signal after smoothing ;

Figure 8 is a flow chart showing a method of correcting an interfered audio signal in another embodiment; Figure 9 is a block diagram showing an apparatus for detecting an interfered audio signal in an embodiment; Figure 10 is an audio interfered with in an embodiment. FIG. 11 is a schematic structural diagram of a correction apparatus for an interfered audio signal in an embodiment; and FIG. 12 is a schematic structural diagram of a correction apparatus for an interfered audio signal in another embodiment.

【detailed description】

The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

In order to correct the interfered audio signal, it is first necessary to detect whether the audio signal is affected. To interference. 2 is a flow chart of a method for detecting an interfered audio signal in an embodiment, including the following steps:

511. Provide an audio signal to be tested.

512: Determine whether the audio data of the audio signal to be tested is interfered at each time point. In a preferred embodiment, the audio data is the frequency value of the audio signal. Here, it is mainly intended to detect the spike noise in the audio signal to be tested, which can be calculated by calculation.

IXo U - K*X _STO (1) to judge. Where is the audio data of the time point to be judged, _v is the average value of the audio data at a time point before the time point to be determined, and a is an empirical value, which is 64 in the present embodiment, then

X _av = (X_, + X— ₂ + ... + X— ₆₃ + X— _M )/64 (2) K is a constant greater than zero and is an empirical value that can be adjusted. X _STO is the standard deviation of the audio data at a time point. In the present embodiment, x _sro = j ∑ (x ₁ - Xav) ² 64 (3) If the equation (1) is greater than zero, the judgment is interfered, otherwise It was judged that there was no interference.

The detection method of the above-mentioned interfered audio signal determines that there is two cases of the interfered audio data, one is that interference is actually received, and the other is that there is a mutation in the original audio signal. In order to improve the accuracy of the detection, in a preferred embodiment, the method of detecting the interfered audio signal further comprises the step of performing a cyclic redundancy check (CRC) on the audio signal to be measured. If the formula (1) is greater than zero, and the verification result of the cyclic redundancy check is erroneous, it is determined that the audio data at the time point to be judged is indeed interfered. If equation (1) is greater than zero and the result of the cyclic redundancy check is correct, step S12 is considered to be a false test, and Xo is not required to be compensated in the subsequent correction step. Refer to the verification result of the cyclic redundancy check to determine whether the audio data is actually received. Interference can reduce the chance of false detections (falsely misidentified audio data as disturbed audio data). Indeed, the audio signal is considered undisturbed, so step S12 is no longer required. Step S12 is executed only if the verification result is an error.

3 is a flow chart of a method of correcting an interfered audio signal in an embodiment. After detecting that the audio signal has been disturbed, it is necessary to compensate the interfered audio signal to remove spike noise, as shown in Figure 4. The method of correcting the interfered audio signal includes the following steps:

521. Provide an audio signal to be tested.

522: Determine whether the audio data of the audio signal to be tested is interfered at each time point. Practice and steps S12—we will not repeat them here.

In an embodiment, the step S21 further includes a step of performing a cyclic redundancy check (CRC) on the audio signal to be tested. If the check result is an error, and the formula (1) is greater than zero, determining the time point to be determined. The audio data is indeed interfered; if equation (1) is greater than zero and the result of the cyclic redundancy check is correct, step S22 is considered to be a false test, and Xo is not required to be compensated in the subsequent correction step.

It is also possible to set the step of the cyclic redundancy check before step S22, and continue to execute step S22 only if the result of the cyclic redundancy check is an error.

523. Compensate the interfered audio data according to the audio data that is not interfered before the time point when the interfered audio data is located. Two compensation methods are provided below.

method one:

The audio data that has been interfered is recorded as ', and the audio data of the undisturbed time points at several time points before the time point is multiplied by the respective weights to compensate for '. In one embodiment, compensation is based on undisturbed audio data at the first two time points:

Χο Ί— ^(1— b)X— ₂ (4) b is the weight constant and is an adjustable risk value. In this embodiment, it is 1/8. In other embodiments, it is also possible to compensate for the undisturbed audio data of the first three, four or other number of time points of the time point at which they are respectively multiplied by the respective weight pair '. Generally, the closer the audio data is on the time axis, the smaller the weight value, and the sum of the weights of all the multiplications is 1.

Method Two:

If it is judged that the audio data is interfered with, the audio data of the previous time point (and the data is undisturbed audio data) ^^ is used for mirror compensation, that is, X is replaced by ^. Figure 6 is a schematic illustration of mirror compensation for successively interfering audio data, substituting substitutions, X ₂ substitutions, X ₃ substitutions X ₂ , X ₄ substitutions X ₃ , X _{5 ,} respectively. The compensated audio data is a mirror image of the audio data used in the compensation.

However, the audio data obtained by the mirror compensation is different from the original audio signal after all. Therefore, in one embodiment, after the step of mirror compensation, the smoothing of the mirror-compensated audio data is further included. A step of. Specifically, a mirror-compensated audio data is multiplied by the mirror compensation weight, and the corresponding interfered audio data (data before the mirror compensation) is multiplied by the interference data weight, and finally the two are added. The sum of the mirroring compensation weight and the interference data weight is 1. In one embodiment, for a plurality of consecutive interfered audio data, the interference data weights gradually increase over time. In this way, after the smoothing process, the compensated data can have a better connection with the subsequent data. For example, for the three smoothed audio data S1, S2, and S3 shown in FIG. 7, the audio data before the mirror compensation is XI, X2, and X3, respectively, and the audio data obtained by the mirror compensation are M1 and M2, respectively. M3, then:

Sl = 3/4* Μ1+ 1/4* Χ1 (5) S2= 1/2*M2+ 1/2* X2 (6) S3= l/4*M3+ 3/4* X3 (7) Three can be seen Interference data weights 1/4, 1/2, 3/4 are gradually increasing.

The method of mirror supplementation is better when the interfered audio data is continuous. The frequency is preserved, and it is simpler to implement, and does not require too high cost.

In one embodiment, the method of correcting the interfered audio signal further includes the step of reducing the volume of the compensated audio data (excluding normal audio data). Specifically, the volume can be reduced by 1/4, and progressive adjustment is used in the process of adjusting the volume. Properly reducing the volume can reduce the user's discomfort.

Figure 8 is a flow diagram of a method of correcting an interfered audio signal in another embodiment. The difference from the embodiment shown in FIG. 3 mainly includes: Step S222 is further included before step S23, and step S23 includes steps S231, S232, and S234, and step S24 is further included after step S23.

S222, that is, determining whether the audio data that is interfered is isolated or continuous. If it is isolated, the process goes to step S231, otherwise it goes to step S232.

5231, a pair of interfered audio data is compensated by the foregoing method.

5232. Perform mirror compensation on the interfered audio data by the foregoing method 2. S234, smoothing the audio data obtained by the mirror compensation. For the specific method of smoothing, please refer to the above.

S24. Perform volume reduction processing on the compensated audio data.

A detection device for an interfered audio signal is also provided.

Figure 9 is a block diagram showing the structure of a device for detecting an interfered audio signal in an embodiment, including an audio signal acquisition module 110 and a first interference determination module 120.

The audio signal acquisition module 110 is configured to acquire an audio signal to be tested.

The first interference determination module 120 is configured to determine whether the audio data of the audio signal to be tested is interfered at each time point. Specifically, it can be judged by calculating the formula (1). If the equation (1) is greater than zero, it is determined that the audio data is interfered, otherwise it is determined that it is not interfered.

Fig. 10 is a block diagram showing the construction of an apparatus for detecting an interfered audio signal in an embodiment, which is mainly different from the embodiment shown in Fig. 9 in that a formatter 111 and a second interference determination module 122 are added.

The formatter 111 is configured to perform the audio signal to be tested acquired by the audio signal acquiring module 110. Ring Redundancy Check (CRC).

The second interference decision module 122 is configured to determine whether the audio data is indeed interfered in conjunction with the check result of the cyclic redundancy check and the judgment result of the equation (1). If the result of the cyclic redundancy check is an error and the equation (1) is greater than zero, it is determined that the audio data at the point in time to be judged is indeed interfered.

11 is a schematic structural diagram of an apparatus for correcting an interfered audio signal in an embodiment. Adding a compensation module 330 to the apparatus for detecting an interfered audio signal shown in FIG. 9 obtains an adjustment of the audio signal shown in FIG. Device.

After the first interference determination module 120 determines the audio data that has been interfered, the compensation module 330 compensates the interfered audio data based on the audio data that was not interfered before the time point at which the interfered audio data was received.

FIG. 12 is a schematic structural diagram of a correction apparatus for an interfered audio signal in another embodiment. The main difference from the embodiment shown in FIG. 11 is that a formatter 111, a second interference determination module 122, and a data type determination module 322 are added. And the volume processing module 240, and in this embodiment, the compensation module 330 specifically includes a first compensation sub-module 331, a second compensation sub-module 332, and a smoothing processing module 334.

The function of the second interference determination module 122 of the formatter 111 has been introduced in the embodiment shown in FIG. 10, and details are not described herein again.

The data type judging module 322 is configured to determine whether the audio data determined by the second interference judging module 122 that is actually interfered is isolated or continuous. In the embodiment where only the first interference determination module 120 is set and the second interference determination module 122 is not provided, it is determined whether the interfered audio data determined by the first interference determination module 120 is isolated or continuous.

If the audio data that is indeed interfered is isolated, the first compensation sub-module 331 compensates according to the first method described above.

If the audio data that is indeed interfered is continuous, the second compensation sub-module 332 performs mirror compensation according to the foregoing method 2. The smoothing processing module 334 is configured to perform smoothing processing on the mirror-compensated audio data. Specifically, a mirror-compensated audio data is multiplied by the mirror compensation weight, and the corresponding interfered audio data (data before the mirror compensation) is multiplied by the interference data weight, and finally the two are added. The sum of the mirroring compensation weight and the interference data weight is 1. In one embodiment, for a plurality of consecutive interfered audio data, the interference data weights gradually increase over time. In this way, after the smoothing process, the compensated data can have a better connection with the subsequent data.

The volume processing module 240 is configured to perform volume reduction processing on the compensated audio data. Specifically, the volume can be reduced by 1/4, and progressive adjustment is used in the process of adjusting the volume. Properly lowering the volume can reduce the user's discomfort.

However, it is not to be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the spirit and scope of the invention. Therefore, the scope of the invention should be determined by the appended claims.

Claims

Claim

1. A method for detecting an interfered audio signal, comprising the steps of:

Providing an audio signal to be tested;

Determining, respectively, whether the audio data of the audio signal to be tested is interfered at each time point; specifically, calculating IXQ-J - K*X _STO , where Xo is audio data at a time point to be determined, v is the The average value of the audio data at a time point before the point in time to be judged, a is a natural number, K is a constant greater than zero, and X _STO is the standard deviation of the audio data of the a time point, if

- K*X _{STO is} greater than zero, and it is determined that the audio data of the time point to be determined is interfered.

A method of detecting an interfered audio signal according to claim 1, wherein a is 64.

3. The method for detecting an interfered audio signal according to claim 1, further comprising the steps of:

Performing a cyclic redundancy check on the audio signal to be tested. If the verification result is an error, and -K*X _{STO is} greater than zero, it is determined that the audio data at the time point to be determined is indeed interfered.

4. The method of detecting an interfered audio signal according to claim 1, wherein the audio data is a frequency value.

5. A method for correcting an interfered audio signal, comprising the steps of:

Step one, providing an audio signal to be tested;

Step 2: respectively determining whether the audio data of the audio signal to be tested is interfered at each time point; specifically, calculating IXo-J - K*X _STO , where Xo is the audio data of the time point to be judged, _v Is the average value of the audio data at a time point before the time point to be judged, a is a natural number, Κ is a constant greater than zero, and X _STO is the standard deviation of the audio data of the a time points, if IXo - J - K * X _{STO is} greater than zero, and it is determined that the audio data of the time point to be determined is interfered; Step 3: Compensating the interfered audio data according to the audio data that is not interfered before the time point when the interfered audio data is located.

6. The method for correcting an interfered audio signal according to claim 5, wherein the step 3 is compensated according to the following formula:

X^bX i + l— b) X ₂

7. The method of correcting an interfered audio signal according to claim 5, wherein the step (3) is to perform mirror compensation.

8. The method of correcting an interfered audio signal according to claim 5, comprising the steps of:

Determining whether the interfered audio data is isolated or continuous;

In the third step, the isolated audio data that is interfered is compensated according to the following formula:

X. =bX- _{1 +} (lb)X- ₂ , where is the compensated interfered audio data, b is the weight constant,

X _I and X ₂ are respectively undisturbed audio data of the first two time points of the time point at which the interfered audio data is located; and in step 3, performing a plurality of consecutive interfered audio data Mirror compensation.

The method of correcting an interfered audio signal according to claim 7 or 8, wherein said step of performing mirror compensation further comprises the step of smoothing the mirror-compensated audio data.

10. The method of correcting an interfered audio signal according to claim 5, further comprising the step of performing volume reduction processing on the compensated audio data.

A method of correcting an interfered audio signal according to claim 5, wherein a is 64.

12. A method of correcting an interfered audio signal according to claim 5, characterized by It also includes the following steps:

Performing a cyclic redundancy check on the audio signal to be tested, if the verification result is an error, and - K*X _{STO is} greater than zero, determining that the audio data at the time point to be determined is indeed interfered, The third step is to compensate the audio data that is not interfered according to the audio data that is not interfered before the time point when the audio data that is actually interfered is present.

13. A method of correcting an interfered audio signal according to claim 5, wherein said audio data is a frequency value.

14. A device for detecting an interfered audio signal, comprising: an audio signal acquisition module and a first interference determination module;

The audio signal acquiring module is configured to acquire an audio signal to be tested;

The first interference determination module is configured to respectively determine whether the audio data of the audio signal to be tested is interfered at each time point; specifically, calculate I XQ U - K*X _sto , where Xo is the time to be determined The audio data of the point, _v is the average value of the audio data at a time point before the time point to be judged, a is a natural number, K is a constant greater than zero, and X _STO is the audio data of the a time point The standard deviation of the IXo-J - K*X _{STO is} greater than zero, and it is determined that the audio data of the time point to be determined is interfered.

15. The apparatus for detecting an interfered audio signal according to claim 14, further comprising a formatter and a second interference determination module;

The formatter is configured to perform cyclic redundancy check on the audio signal to be tested acquired by the audio signal acquiring module;

The second interference determination module determines according to the verification result of the cyclic redundancy check and the determination result of the first interference determination module, if the verification result is an error, and I XQ U - K*X _{If STO is} greater than zero, it is determined that the audio data at the time point to be determined is indeed interfered.

16. A correction device for an interfered audio signal, characterized by comprising an audio signal Number acquisition module, first interference determination module and compensation module;

The first interference determination module is configured to respectively determine whether the audio data of the audio signal to be tested is interfered at each time point; specifically, calculate IXQ U - K*X _sto , where Xo is a time point to be determined Audio data, _v is the average value of the audio data at a time point before the time point to be determined, a is a natural number, K is a constant greater than zero, and X _STO is the audio data of the a time point Standard deviation, if IXo-J - K*X _{STO is} greater than zero, it is determined that the audio data of the time point to be determined is interfered;

The compensation module is configured to compensate the interfered audio data according to audio data that is not interfered before the time point at which the interfered audio data is located.

17. The apparatus for correcting an interfered audio signal according to claim 16, wherein the compensation module is a first compensation sub-module for compensating according to the following formula:

X^bX i + l— b) X ₂

18. The apparatus for correcting an interfered audio signal according to claim 16, wherein the compensation module is a second compensation sub-module for performing mirror compensation.

The apparatus for correcting an interfered audio signal according to claim 16, further comprising: a data type determining module, configured to determine whether the interfered audio data is isolated or continuous;

The compensation module includes a first compensation sub-module and a second compensation sub-module, and the first compensation sub-module is configured to compensate the isolated audio data that is interfered according to the following formula: x. =bx- _{1 +} (lb)x- ₂ , where is the compensated interfered audio data, b is the weight constant, X _l , X ₂ are the first two of the time points of the interfered audio data Undisturbed audio data at a point in time; the second compensation sub-module is used to interfere with multiple consecutive The audio data is mirror compensated.

The apparatus for correcting an interfered audio signal according to claim 18, further comprising a smoothing processing module for smoothing the mirror-compensated audio data.

The apparatus for correcting an interfered audio signal according to claim 16, further comprising a volume processing module for performing volume reduction processing on the compensated audio data.

22. The apparatus for correcting an interfered audio signal according to claim 16, further comprising a formatter and a second interference determination module;

The second interference determination module determines according to the verification result of the cyclic redundancy check and the determination result of the first interference determination module, if the verification result is an error, and IXQ U - K*X _STO If it is greater than zero, it is determined that the audio data of the time point to be determined is indeed interfered; the compensation module compensates the audio data that is indeed interfered.