Movatterモバイル変換

[0]ホーム
URL:

CN115604614B - System and method for local sound amplification and remote interaction by using hoisting microphone - Google Patents

System and method for local sound amplification and remote interaction by using hoisting microphoneDownload PDF

Info

Publication number
CN115604614B
CN115604614BCN202211611396.4ACN202211611396ACN115604614BCN 115604614 BCN115604614 BCN 115604614BCN 202211611396 ACN202211611396 ACN 202211611396ACN 115604614 BCN115604614 BCN 115604614B
Authority
CN
China
Prior art keywords
filter
signal
microphone
processing
sub
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202211611396.4A
Other languages
Chinese (zh)
Other versions
CN115604614A (en
Inventor
刘开文
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chengdu Haipudi Technology Co ltd
Original Assignee
Chengdu Haipudi Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Chengdu Haipudi Technology Co ltdfiledCriticalChengdu Haipudi Technology Co ltd
Priority to CN202211611396.4ApriorityCriticalpatent/CN115604614B/en
Publication of CN115604614ApublicationCriticalpatent/CN115604614A/en
Application grantedgrantedCritical
Publication of CN115604614BpublicationCriticalpatent/CN115604614B/en
Activelegal-statusCriticalCurrent
Anticipated expirationlegal-statusCritical

Links

Images

Classifications

Landscapes

Abstract

The invention relates to the field of audio signal processing, and aims to provide a lifting deviceSystem and method for local amplification and remote interaction with a microphone, wherein the method comprises processing the audio signal
Figure 124710DEST_PATH_IMAGE001
Dividing a number of sub-band signals
Figure 733546DEST_PATH_IMAGE002
Pre-filtering: first pass whitening filter
Figure 907039DEST_PATH_IMAGE003
Carrying out smooth half-wave rectification treatment; step-length-variable adaptive signal processing: for whitening filter in each sub-band
Figure 448878DEST_PATH_IMAGE003
The processed signals are respectively subjected to variable step length adaptive filtering processing to obtain
Figure 416834DEST_PATH_IMAGE004
(ii) a And (3) nonlinear processing: by non-linear processing filters
Figure 512966DEST_PATH_IMAGE005
For the signal after the adaptive signal processing with variable step length
Figure 224571DEST_PATH_IMAGE004
Performing nonlinear filtering to obtain processed signal
Figure 620917DEST_PATH_IMAGE006
(ii) a For each sub-band signal after nonlinear processing
Figure 759774DEST_PATH_IMAGE006
Performing comprehensive filtering to obtain time domain signal with feedback eliminated
Figure 343202DEST_PATH_IMAGE007
And the acoustic feedback elimination based on a pre-filtering method, variable step size self-adaptive signal processing and nonlinear processing is effectively realized in real time.

Description

System and method for local sound amplification and remote interaction by using hoisting microphone
Technical Field
The invention relates to an acoustic feedback suppression algorithm in the field of audio signal processing, in particular to a system and a method for carrying out local sound amplification and remote interaction by adopting a hoisting microphone.
Background
In the process of simultaneously carrying out local sound amplification and remote interaction by adopting a hoisting microphone, the local sound amplification can generate howling, the remote interaction can generate Echo, and the howling and the Echo are the two most difficult problems in audio processing, so that the Feedback inhibition (AFC) and the Echo Cancellation (AEC) are hot points of audio research, the hoisting microphone belongs to remote sound pickup, the sensitivity of the microphone is very high, the Echo generated by the local sound amplification and the remote interaction is more serious, and no audio system equipment can really realize the local sound amplification and the remote interaction at the same time.
The source of the generation of howling and echo is that sound emitted by a sound box is picked up by a microphone, a hoisting microphone is adopted to pick up sound in a long distance, the microphone has high sensitivity and more serious howling and echo, and to realize the simultaneous local sound amplification and remote interaction of the hoisting microphone, the most effective method is to remove the sound fed back to the microphone by the sound box through an algorithm, only near-end audio is reserved, the technology of removing the sound fed back to the microphone by the sound box is changed into an acoustic feedback technology, but the removal of the sound fed back to the microphone by the sound box faces the following technical problems:
(1) Correlation in feedback suppression algorithms
In the local amplification process, because the voice coming out of the loudspeaker box and the local speaker speaking are the voice of the same person, the two voices have strong correlation, and the correlation can cause the convergence mismatch of the adaptive filter.
(2) Nonlinear effects of microphones, sound boxes and sound transmission channels in practical applications
In the process of local sound amplification and remote interaction, due to the fact that nonlinearity exists in a sound field, the self-adaptive filter is difficult to really match with an actual sound field environment.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, provides a system and a method for carrying out local sound amplification and remote interaction by adopting a hoisting microphone, and effectively realizes the acoustic feedback elimination based on a pre-filtering method, variable step length self-adaptive signal processing and nonlinear processing in real time.
The method is realized by the following technical scheme: on the one hand, a hoisting microphone is adoptedMethod for local sound amplification and remote interaction, sound signal collected by microphone
Figure DEST_PATH_IMAGE002
Mix the near-end audio>
Figure DEST_PATH_IMAGE004
And the audio frequency which is fed back to the microphone by the sound box>
Figure DEST_PATH_IMAGE006
The method is characterized by comprising the following steps: for the sound signal>
Figure 537212DEST_PATH_IMAGE002
Performing analysis filtering processing to divide the signal into several sub-band signals->
Figure DEST_PATH_IMAGE008
；
Pre-filtering: first pass whitening filter
Figure DEST_PATH_IMAGE010
Carrying out smooth half-wave rectification treatment;
step-length-variable adaptive signal processing: for each sub-band to whiten the filter
Figure 623986DEST_PATH_IMAGE010
The processed signals are respectively processed by variable step length adaptive filtering to obtain->
Figure DEST_PATH_IMAGE012
；
And (3) nonlinear processing: by non-linear processing filters
Figure DEST_PATH_IMAGE014
For the signal after the adaptive signal processing with variable step length
Figure 127779DEST_PATH_IMAGE012
Performing nonlinear filtering to obtain each sub-band signal->
Figure DEST_PATH_IMAGE016
；/>
For each sub-band signal to be actually output
Figure 163869DEST_PATH_IMAGE016
Comprehensively filtering to obtain the near-end audio frequency after feedback elimination>
Figure 968184DEST_PATH_IMAGE004
。
Further, the whitening filter is
Figure DEST_PATH_IMAGE018
In the formula (I), the compound is shown in the specification,
Figure DEST_PATH_IMAGE020
in order to whiten the filtered sub-band signal,smooth() Is a smooth rectification function.
Further, the sub-band signal processing specifically includes the following steps,
s31: sound signal collected by microphone
Figure DEST_PATH_IMAGE021
Performing processing sub-band analysis filtering processing to obtain sub-band input of microphone as ^ or ^>
Figure DEST_PATH_IMAGE022
The sound box with sub-band feeds back sound signal of->
Figure DEST_PATH_IMAGE024
And the near end audio of a sub-band +>
Figure DEST_PATH_IMAGE026
，
Figure DEST_PATH_IMAGE028
In, is greater than or equal to>
Figure DEST_PATH_IMAGE030
Is a subband signal;
S32：
Figure 479936DEST_PATH_IMAGE008
and &>
Figure DEST_PATH_IMAGE031
Passes through a whitening filter->
Figure 521842DEST_PATH_IMAGE010
After processing, get>
Figure DEST_PATH_IMAGE033
，
Figure DEST_PATH_IMAGE035
；
S33: to pair
Figure DEST_PATH_IMAGE037
And &>
Figure DEST_PATH_IMAGE039
Send as input to adaptive filter to background filter +>
Figure DEST_PATH_IMAGE041
；
S34: background filter
Figure 802650DEST_PATH_IMAGE041
Assign a value to the foreground filter>
Figure DEST_PATH_IMAGE043
Get->
Figure DEST_PATH_IMAGE045
，
Figure 786656DEST_PATH_IMAGE045
Is sent to a non-linear filter->
Figure 395492DEST_PATH_IMAGE014
Get->
Figure 506667DEST_PATH_IMAGE016
；
S35: by passing
Figure 517348DEST_PATH_IMAGE016
Judging whether the near-end audio is mute, if so, executing S36, and if not, executing S31;
s36: the sub-bands are subjected to comprehensive filtering to obtain the near-end audio frequency after feedback elimination
Figure DEST_PATH_IMAGE046
。
Further, in said S31, for
Figure 485304DEST_PATH_IMAGE021
Pre-emphasis processing, framing processing, low-pass filtering processing and analysis filtering processing are sequentially carried out.
Further, the S31 includes 256 subbands.
On the other hand, in some embodiments, the system further comprises a system for local sound amplification and remote interaction by using a ceiling microphone, wherein the sound signal collected by the microphone comprises a plurality of sub-band signals
Figure DEST_PATH_IMAGE047
Wherein each sub-band signal->
Figure 771317DEST_PATH_IMAGE047
In which a near-end audio is mixed>
Figure DEST_PATH_IMAGE048
And the signal fed back to the microphone by the sound box>
Figure DEST_PATH_IMAGE049
Sub-band signal->
Figure DEST_PATH_IMAGE050
Has an actual output of->
Figure DEST_PATH_IMAGE051
The system comprises a first whitening filter->
Figure 341975DEST_PATH_IMAGE010
The second whitening filter->
Figure 472742DEST_PATH_IMAGE010
The foreground filter->
Figure 611600DEST_PATH_IMAGE043
Background filter
Figure 132711DEST_PATH_IMAGE041
Non-linear processing filter->
Figure 382427DEST_PATH_IMAGE014
And a feedback acoustic analog filter->
Figure DEST_PATH_IMAGE053
；
The first whitening filter
Figure DEST_PATH_IMAGE055
For each sub-band signal->
Figure DEST_PATH_IMAGE056
The output is->
Figure DEST_PATH_IMAGE057
；
The second whitening filter
Figure DEST_PATH_IMAGE058
Is the actual output->
Figure 492334DEST_PATH_IMAGE051
And the output is->
Figure DEST_PATH_IMAGE059
；
The background filter
Figure DEST_PATH_IMAGE061
Is the output of the second whitening filter;
the foreground filter
Figure 926726DEST_PATH_IMAGE043
Is inputted as->
Figure 997451DEST_PATH_IMAGE016
；
The foreground filter
Figure 240738DEST_PATH_IMAGE043
In or on the output>
Figure DEST_PATH_IMAGE063
And sub-band signal>
Figure DEST_PATH_IMAGE065
Combining as the non-linear processing filter>
Figure 18201DEST_PATH_IMAGE014
The input of (a) is performed,
the nonlinear processing filter
Figure DEST_PATH_IMAGE067
The output being the actual output of each subband>
Figure 685811DEST_PATH_IMAGE051
，/>
By means of said feedback acoustic analogue filter
Figure 978252DEST_PATH_IMAGE053
Signal fed back to microphone by analog loudspeaker box>
Figure DEST_PATH_IMAGE068
，
Computing actual outputs of subbands
Figure 773033DEST_PATH_IMAGE016
Comprehensively filtering to obtain a near-end time domain signal after feedback elimination>
Figure DEST_PATH_IMAGE070
。
Further, the foreground filter
Figure DEST_PATH_IMAGE072
Coefficient passing background filter>
Figure DEST_PATH_IMAGE073
Is obtained by the update of (c).
Further, the second whitening filter
Figure DEST_PATH_IMAGE074
For attenuating the actual output>
Figure DEST_PATH_IMAGE075
The sound box feeds back the sound signal->
Figure DEST_PATH_IMAGE077
。
Further, the first whitening filter
Figure 778904DEST_PATH_IMAGE074
And said second whitening filter>
Figure 368148DEST_PATH_IMAGE010
Smooth half-wave rectification is adopted.
Further, the foreground filter
Figure 413465DEST_PATH_IMAGE072
Is greater or less than>
Figure DEST_PATH_IMAGE078
For feeding back signals from loudspeaker to microphone
Figure DEST_PATH_IMAGE080
Is estimated.
The invention has the beneficial effects that:
(1) In order to effectively remove the audio frequency fed back to the microphone by the sound box
Figure 998554DEST_PATH_IMAGE006
In the method, a signal collected by a lifting microphone is firstly collected>
Figure 547347DEST_PATH_IMAGE021
Performing subband (subband) analysis filtering processing, dividing the full-band audio signal into a plurality of subbands with equal bandwidth, and performing acoustic feedback elimination on each subband by using the method shown in FIG. 1;
(2) The method adopts a foreground and background operation mode, i.e. a foreground filter
Figure 635389DEST_PATH_IMAGE043
Adaptive filter with coefficients passing through the background->
Figure 902422DEST_PATH_IMAGE041
Is updated to obtain and is sent to the background adaptive filter>
Figure 773427DEST_PATH_IMAGE041
Is passed through a first whitening filter->
Figure DEST_PATH_IMAGE081
To weaken the signal correlation and thereby effectively weaken the near-end audio->
Figure 442305DEST_PATH_IMAGE046
And the signal fed back to the microphone by the sound box>
Figure 888199DEST_PATH_IMAGE006
The strong correlation of the adaptive filter increases the robustness of the adaptive filter, and signals running in the foreground are not subjected to corresponding whitening processing, and the output tone quality of the system is not influenced.
Drawings
FIG. 1 is a schematic diagram of subband acoustic feedback cancellation in one embodiment of the present invention;
FIG. 2 is a flow chart of a subband acoustic feedback cancellation process in one embodiment of the present invention;
fig. 3 is a working schematic diagram of a ceiling-mounted microphone in one embodiment of the invention.
Description of the reference numerals: 100. an echo signal; 101. a howling signal; 102. near-end audio.
Detailed Description
The technical solutions in the embodiments of the present invention are clearly and completely described below with reference to fig. 1, fig. 2 and fig. 3 of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other implementations made by those of ordinary skill in the art based on the embodiments of the present invention are obtained without inventive efforts.
In the description of the present invention, it is to be understood that the terms "counterclockwise", "clockwise", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate orientations or positional relationships based on those shown in the drawings, and are used for convenience of description only, and do not indicate or imply that the referenced devices or elements must have a particular orientation, be constructed and operated in a particular orientation, and thus, are not to be considered as limiting.
Example 1:
method for local sound amplification and remote interaction by using hoisting microphone, signal d (n) collected by hoisting microphone (sound signal collected by microphone)
Figure DEST_PATH_IMAGE083
) Including near-end tonesFrequently v (n) (proximal tone @)>
Figure DEST_PATH_IMAGE085
) And the signal y (n) fed back to the microphone by the sound box (the audio frequency fed back to the microphone by the sound box &)>
Figure DEST_PATH_IMAGE087
) Namely: d (n) = v (n) + y (n), the main purpose of the invention is to remove the signal y (n) fed back to the microphone by the sound box in d (n), and only keep the near-end audio v (n), thereby realizing the purpose of no howling in local sound amplification and no echo in remote interaction. In order to effectively remove y (n), sub-band (subband) analysis filtering processing is firstly carried out on a signal d (n) collected by a hoisting microphone, a full-band audio signal is divided into a plurality of sub-bands, each sub-band is subjected to acoustic feedback elimination by adopting the method shown in the figure 1, the core technology of the method mainly comprises a sub-band pre-filtering method, variable step length self-adaptive signal processing and nonlinear processing, and a good application effect is achieved in practical application.
A pre-filtering method
The pre-filtering method is similar to the whitening process in signal processing, in that the signal is passed through a fixed whitening filter
Figure 376949DEST_PATH_IMAGE010
The frequency spectrum distribution of the signal sent into the adaptive filter, namely the correlation characteristic of the signal itself, is changed, and the purpose of weakening the signal correlation is achieved. Fig. 1 shows a basic schematic diagram thereof. The method adopts a foreground and background operation mode, i.e. a filter of the foreground
Figure DEST_PATH_IMAGE088
Coefficient passing background filter>
Figure 317223DEST_PATH_IMAGE041
Is updated to obtain and is sent to the background filter>
Figure 840609DEST_PATH_IMAGE041
By means of a whitening filter>
Figure DEST_PATH_IMAGE089
To attenuate the signal correlation and thereby effectively attenuate the near-end audio &>
Figure 191824DEST_PATH_IMAGE046
And the signal fed back to the microphone by the sound box>
Figure DEST_PATH_IMAGE090
The strong correlation of the method increases the robustness of the adaptive filter, the signals operated by the foreground do not carry out corresponding whitening processing, so the output tone quality of the system cannot be influenced, extra calculation and storage space are inevitably added to meet the operation of the foreground and the background, but along with the continuous development of chip technology, the CPU calculation capability is continuously enhanced, and the influence of the increased calculation and storage space on the product is very small, and the whitening filter provided by the invention has the advantages that the signal operated by the foreground does not carry out corresponding whitening processing, the output tone quality of the system is not influenced, and the method has the advantages that the CPU calculation capability is continuously enhanced, the calculation capability and the storage space are greatly influenced>
Figure 433450DEST_PATH_IMAGE010
By means of smooth half-wave rectification,
Figure 974153DEST_PATH_IMAGE018
，
in the formula (I), the compound is shown in the specification,
Figure 289727DEST_PATH_IMAGE020
are variables of the function in the whitening filter,smooth() To smooth the rectification function, the rectification function is, in this embodiment,
Figure 890473DEST_PATH_IMAGE020
for a sub-band signal>
Figure DEST_PATH_IMAGE091
。
Step-length-variable adaptive signal processing
The linear processing part is to whiten the filter in each sub-band
Figure 353815DEST_PATH_IMAGE010
The processed signals are respectively processed in variable step length and nonlinear processing, and sub-band signal processing is mainly carried out according to the following formula:
Figure DEST_PATH_IMAGE093
Figure DEST_PATH_IMAGE095
in the formula, T is a matrix transpose symbol,
Figure DEST_PATH_IMAGE096
is the subband signal (the input variable value in the function).
Figure DEST_PATH_IMAGE098
Figure DEST_PATH_IMAGE100
Figure DEST_PATH_IMAGE102
Figure DEST_PATH_IMAGE104
Figure DEST_PATH_IMAGE106
Wherein m is the iteration times, m is a natural integer, m is more than or equal to 0,
Figure DEST_PATH_IMAGE108
for a step factor (step constant), ->
Figure DEST_PATH_IMAGE110
Adaptive filter for background->
Figure DEST_PATH_IMAGE112
Linear part of the processed results.
Figure DEST_PATH_IMAGE114
Figure DEST_PATH_IMAGE116
Figure DEST_PATH_IMAGE118
Figure DEST_PATH_IMAGE120
Where L is the length (number of iterations) of the background adaptive filter for each subband,
Figure DEST_PATH_IMAGE122
for smoothing coefficients>
Figure 961576DEST_PATH_IMAGE112
For background adaptive filter, <' > based on the background adaptive filter>
Figure DEST_PATH_IMAGE124
For foreground adaptive filter,>
Figure DEST_PATH_IMAGE126
for the result processed by the linear part of the background adaptive filter, then>
Figure DEST_PATH_IMAGE128
Is an input signal of a microphone>
Figure DEST_PATH_IMAGE130
For the input signal of a microphoneResults processed by an over-whitening filter>
Figure 397236DEST_PATH_IMAGE063
Is the signal which is fed back to the microphone by the sound box>
Figure 903304DEST_PATH_IMAGE080
In combination with a predetermined number of previous evaluations>
Figure DEST_PATH_IMAGE132
Is a variable-step coefficient whose magnitude is input +>
Figure 306472DEST_PATH_IMAGE037
And processing result>
Figure DEST_PATH_IMAGE134
Is determined.
Three, non-linear processing
Due to the non-linearity of the system, a non-linear filtering process is required to be carried out on each sub-band, and a non-linear processing filter
Figure DEST_PATH_IMAGE136
The calculation method is as follows:
Figure DEST_PATH_IMAGE138
Figure DEST_PATH_IMAGE140
Figure DEST_PATH_IMAGE142
Figure DEST_PATH_IMAGE144
wherein
Figure 313612DEST_PATH_IMAGE122
Is smoothed out>
Figure DEST_PATH_IMAGE146
For the autocorrelation of a subband transformation in the microphone input signal, a value is determined>
Figure DEST_PATH_IMAGE148
A cross-correlation of sub-band transforms for the microphone input signal.
Example 2:
referring to fig. 2, in this embodiment, the following steps are obtained by applying the above method to specific subbands:
the sub-band analysis filtering processing includes pre-emphasis processing, framing processing, low-pass filtering processing and analysis filtering processing on the signal d (n) collected by the near-end microphone, in this embodiment, 32KHz sampling is adopted to divide the signal into 256 sub-bands.
Microphone input for each sub-band
Figure 666095DEST_PATH_IMAGE065
And output->
Figure DEST_PATH_IMAGE149
And performing pre-filtering treatment.
Figure DEST_PATH_IMAGE151
Figure DEST_PATH_IMAGE153
Figure DEST_PATH_IMAGE155
Thirdly, carrying out variable step length self-adaptive signal processing and nonlinear processing on the pre-filtered signal of each sub-band
Figure DEST_PATH_IMAGE156
Figure DEST_PATH_IMAGE158
Figure 470628DEST_PATH_IMAGE098
Figure 908562DEST_PATH_IMAGE100
Figure 532442DEST_PATH_IMAGE102
Figure 739432DEST_PATH_IMAGE104
Figure DEST_PATH_IMAGE159
Figure DEST_PATH_IMAGE160
Figure 39832DEST_PATH_IMAGE138
Figure DEST_PATH_IMAGE161
Figure DEST_PATH_IMAGE162
Wherein
Figure 965063DEST_PATH_IMAGE122
For smoothing out the factor, this embodiment->
Figure 127054DEST_PATH_IMAGE122
=0.95. Again, the above parameters are illustrated, where L is the length of the background adaptive filter for each subband, and ` H `>
Figure 454130DEST_PATH_IMAGE122
For smoothing coefficients>
Figure 738481DEST_PATH_IMAGE112
For background adaptive filter, <' > based on the background adaptive filter>
Figure DEST_PATH_IMAGE163
For foreground adaptive filter,>
Figure 72379DEST_PATH_IMAGE126
for results processed in the linear part of the background adaptive filter>
Figure DEST_PATH_IMAGE164
Is an input signal of a microphone>
Figure 100378DEST_PATH_IMAGE130
For the result of the processing of the input signal of the microphone by means of the whitening filter, is->
Figure 219644DEST_PATH_IMAGE063
Is the signal that the sound box feeds back to the microphone->
Figure 674896DEST_PATH_IMAGE080
In combination with a predetermined number of previous evaluations>
Figure DEST_PATH_IMAGE165
Is a variable-step coefficient whose magnitude is input +>
Figure DEST_PATH_IMAGE166
And (b) aBased on results>
Figure DEST_PATH_IMAGE167
Is determined. m is the number of iterations, and m is a natural integer, m ≧ 0, < >>
Figure DEST_PATH_IMAGE168
Is a step size coefficient (step size constant). T is the matrix transposed symbol for the background adaptive filter for L iterations, for->
Figure DEST_PATH_IMAGE169
For a sub-band signal>
Figure DEST_PATH_IMAGE170
Adaptive filter for background->
Figure DEST_PATH_IMAGE171
Linear part of the processed results.
Obtaining foreground self-adaptive filter through background
Figure 699353DEST_PATH_IMAGE043
Calculating the actual output &' for each sub-band>
Figure DEST_PATH_IMAGE173
Figure DEST_PATH_IMAGE175
Figure 455344DEST_PATH_IMAGE118
Figure 491433DEST_PATH_IMAGE120
Figure DEST_PATH_IMAGE177
Fifthly, obtaining the processed near-end audio signal through sub-band comprehensive filtering
Figure DEST_PATH_IMAGE179
。
It should be noted that, referring to fig. 3, the near-end audio, i.e. the near-end audio 102 (near-end audio signal) in the embodiment, is included
Figure DEST_PATH_IMAGE180
) The audio signal fed back to the microphone by the sound box is greater or less than>
Figure DEST_PATH_IMAGE182
(howling signal 101) and also echo signal 100 generated by the far-end audio. The present embodiment focuses on eliminating howling signals generated by a ceiling microphone.
In summary, the acoustic feedback elimination based on the pre-filtering method, the variable-step-size adaptive signal processing and the nonlinear processing is realized, the frequency of the near-end audio is not changed while the howling noise generated by the hoisting microphone is removed, only the howling signal mixed in the near-end audio is removed, and the effective near-end audio signal is reserved.
Those skilled in the art will recognize that the functionality described in this disclosure may be implemented in a combination of hardware and software in one or more of the examples described above. When software is applied, the corresponding functionality may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a general purpose or special purpose computer.

Claims (9)

1. Method for carrying out local sound amplification and remote interaction by adopting hoisting microphone and sound signal collected by microphone
Figure QLYQS_1
Mix near-end audio->
Figure QLYQS_2
And the audio frequency fed back to the microphone by the sound box>
Figure QLYQS_3
The method is characterized by comprising the following steps: for the sound signal
Figure QLYQS_4
Analysis-filtering a signal into a number of subband signals->
Figure QLYQS_5
；
Pre-filtering: early pass whitening filter
Figure QLYQS_6
Carrying out smooth half-wave rectification treatment;
step-length-variable adaptive signal processing: for each sub-band to whiten the filter
Figure QLYQS_7
The processed signals are respectively processed by variable step length adaptive filtering to obtain->
Figure QLYQS_8
；
And (3) nonlinear processing: by non-linear processing filters
Figure QLYQS_9
For the signal after the variable-step adaptive signal processing>
Figure QLYQS_10
Performing nonlinear filtering to obtain actual output of each sub-band signal>
Figure QLYQS_11
；
For each sub-band signal to be actually output
Figure QLYQS_12
Comprehensively filtering to obtain the near-end audio frequency after feedback elimination>
Figure QLYQS_13
，
The sub-band signal processing comprises in particular the following steps,
s31: sound signal collected by microphone
Figure QLYQS_14
Performing subband analysis and filtering to obtain subband input->
Figure QLYQS_15
The signal fed back to the microphone by the sound box is->
Figure QLYQS_16
And the near end audio of a sub-band +>
Figure QLYQS_17
，
Figure QLYQS_18
In the formula (II)>
Figure QLYQS_19
Is a subband signal;
S32：
Figure QLYQS_20
and &>
Figure QLYQS_21
Passes through a whitening filter->
Figure QLYQS_22
After processing, get>
Figure QLYQS_23
，
Figure QLYQS_24
；
S33: for is to
Figure QLYQS_25
And &>
Figure QLYQS_26
Send as input to adaptive filter to background filter +>
Figure QLYQS_27
；
S34: background filter
Figure QLYQS_28
Assign a value to a foreground filter>
Figure QLYQS_29
Get>
Figure QLYQS_30
，
Figure QLYQS_31
Send to a non-linear filter>
Figure QLYQS_32
Get->
Figure QLYQS_33
；
S35: by passing
Figure QLYQS_34
Judging whether the near-end audio is mute, if so, executing S36, and if not, executing S31;
s36: the sub-bands are subjected to comprehensive filtering to obtain the near-end audio frequency after feedback elimination
Figure QLYQS_35
。
2. The method of claim 1, wherein the whitening filter is a ceiling microphone for local amplification and remote interaction
Figure QLYQS_36
In the formula (I), the compound is shown in the specification,
Figure QLYQS_37
for the variables of the input in the whitening filter,smooth() Is a smooth rectification function.
3. The method for local sound amplification and remote interaction by using ceiling-mounted microphone according to claim 2, wherein in S31, the pair
Figure QLYQS_38
Pre-emphasis processing, framing processing, low-pass filtering processing and analysis filtering processing are sequentially carried out.
4. The method for local sound amplification and remote interaction by using ceiling-mounted microphones as claimed in claim 3, wherein said S31 comprises 256 sub-bands.
5. A system for local sound amplification and remote interaction by using a hoisting microphone, which is applied to the method for local sound amplification and remote interaction by using a hoisting microphone as claimed in claim 4, wherein the sound signal collected by the microphone comprises a plurality of subband signals
Figure QLYQS_40
In which each sub-band signal->
Figure QLYQS_45
In which is mixed with near-end audio>
Figure QLYQS_48
And the signal fed back to the microphone by the sound box>
Figure QLYQS_41
Sub-band signal->
Figure QLYQS_44
Has an actual output of->
Figure QLYQS_47
The system includes a first whitening filter>
Figure QLYQS_50
A second whitening filter>
Figure QLYQS_39
Based on the foreground filter>
Figure QLYQS_43
Based on the background filter>
Figure QLYQS_46
Non-linear processing filter->
Figure QLYQS_49
And a feedback acoustic analog filter->
Figure QLYQS_42
；
The first whitening filter
Figure QLYQS_51
For each sub-band signal->
Figure QLYQS_52
And the output is->
Figure QLYQS_53
；
The second whitening filter
Figure QLYQS_54
Is actually output>
Figure QLYQS_55
And the output is->
Figure QLYQS_56
；
The background filter
Figure QLYQS_57
Is the output of the second whitening filter;
the foreground filter
Figure QLYQS_58
Is input as->
Figure QLYQS_59
；
The foreground filter
Figure QLYQS_60
Is greater or less than>
Figure QLYQS_61
And sub-band signal->
Figure QLYQS_62
Incorporating as a non-linear processing filter
Figure QLYQS_63
The input of (a) is performed,
the nonlinear processing filter
Figure QLYQS_64
The output being the actual output of each sub-band
Figure QLYQS_65
，
By means of said feedback acoustic analogue filter
Figure QLYQS_66
Signal fed back to microphone by analog loudspeaker box>
Figure QLYQS_67
，
Calculating the actual output of the subband signals
Figure QLYQS_68
Comprehensively filtering to obtain a near-end time domain signal after feedback elimination>
Figure QLYQS_69
。
6. The system of claim 5, wherein the front stage filter is configured to perform local sound amplification and remote interaction with a ceiling-mounted microphone
Figure QLYQS_70
Is passed through a background filter->
Figure QLYQS_71
The update of (2) is obtained.
7. The system of claim 6, wherein the second whitening filter is configured to provide local amplification and remote interaction using a ceiling-mounted microphone
Figure QLYQS_72
For attenuating the actual output>
Figure QLYQS_73
The sound box in (1) feeds back a signal to the microphone>
Figure QLYQS_74
。
8. The system of claim 7, wherein the first whitening filter is configured to locally amplify and remotely interact with a ceiling-mounted microphone
Figure QLYQS_75
And the second whitening filter->
Figure QLYQS_76
Smooth half-wave rectification is adopted.
9. The system of claim 8, wherein the front stage filter is configured to perform local sound amplification and remote interaction with a ceiling-mounted microphone
Figure QLYQS_77
Is greater or less than>
Figure QLYQS_78
For the signal fed back to the microphone by the loudspeaker box>
Figure QLYQS_79
Is estimated. />
CN202211611396.4A2022-12-152022-12-15System and method for local sound amplification and remote interaction by using hoisting microphoneActiveCN115604614B (en)

Priority Applications (1)

Application NumberPriority DateFiling DateTitle
CN202211611396.4ACN115604614B (en)2022-12-152022-12-15System and method for local sound amplification and remote interaction by using hoisting microphone

Applications Claiming Priority (1)

Application NumberPriority DateFiling DateTitle
CN202211611396.4ACN115604614B (en)2022-12-152022-12-15System and method for local sound amplification and remote interaction by using hoisting microphone

Publications (2)

Publication NumberPublication Date
CN115604614A CN115604614A (en)2023-01-13
CN115604614Btrue CN115604614B (en)2023-03-31

Family

ID=84854134

Family Applications (1)

Application NumberTitlePriority DateFiling Date
CN202211611396.4AActiveCN115604614B (en)2022-12-152022-12-15System and method for local sound amplification and remote interaction by using hoisting microphone

Country Status (1)

CountryLink
CN (1)CN115604614B (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN101438603A (en)*2006-04-012009-05-20唯听助听器公司Hearing aid, and a method for control of adaptation rate in anti-feedback systems for hearing aids
CN101939781A (en)*2008-01-042011-01-05杜比国际公司 Audio Encoders and Decoders

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
US5305307A (en)*1991-01-041994-04-19Picturetel CorporationAdaptive acoustic echo canceller having means for reducing or eliminating echo in a plurality of signal bandwidths
US5649012A (en)*1995-09-151997-07-15Hughes ElectronicsMethod for synthesizing an echo path in an echo canceller
SE505150C3 (en)*1995-10-181997-08-04Ericsson Telefon Ab L M Adaptive dual filter echo leakage procedure
WO2000072566A1 (en)*1999-05-202000-11-30Telefonaktiebolaget Lm Ericsson (Publ)Methods and apparatus for improving adaptive filter performance by signal equalization
EP2086250B1 (en)*2008-02-012020-05-13Oticon A/SA listening system with an improved feedback cancellation system, a method and use
CN101272161B (en)*2008-04-302012-12-05中山大学Self-adapting echo cancellation method
JP2012048134A (en)*2010-08-302012-03-08Nippon Telegr & Teleph Corp <Ntt>Reverberation removal method, reverberation removal device and program
CN106454642B (en)*2016-09-232019-01-08佛山科学技术学院Adaptive sub-band audio feedback suppression methods
CN109727604B (en)*2018-12-142023-11-10上海蔚来汽车有限公司Frequency domain echo cancellation method for speech recognition front end and computer storage medium
CN113824846A (en)*2021-09-182021-12-21合肥真讯信息科技有限公司Howling suppression method and equipment for trunking communication system
CN113838474B (en)*2021-11-252022-02-18全时云商务服务股份有限公司Communication system howling suppression method and device
CN115459742A (en)*2022-09-282022-12-09九江慧明电子科技有限公司Acoustic feedback suppression method introducing signal pre-whitening mechanism

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN101438603A (en)*2006-04-012009-05-20唯听助听器公司Hearing aid, and a method for control of adaptation rate in anti-feedback systems for hearing aids
CN101939781A (en)*2008-01-042011-01-05杜比国际公司 Audio Encoders and Decoders

Also Published As

Publication numberPublication date
CN115604614A (en)2023-01-13

Similar Documents

PublicationPublication DateTitle
CN109727604B (en)Frequency domain echo cancellation method for speech recognition front end and computer storage medium
CN110838300B (en)Echo cancellation processing method and processing system
CN110536215B (en)Method and apparatus for audio signal processing, calculation device, and storage medium
CN1689072B (en) Method and system for processing subband signals using adaptive filters
US8594320B2 (en)Hybrid echo and noise suppression method and device in a multi-channel audio signal
CN110611871B (en)Howling suppression method and system for digital hearing aid and special DSP
US9947337B1 (en)Echo cancellation system and method with reduced residual echo
US3784747A (en)Speech suppression by predictive filtering
CN102739886B (en)Stereo echo offset method based on echo spectrum estimation and speech existence probability
CN103730125A (en)Method and equipment for echo cancellation
US20030031315A1 (en)Echo canceller having nonlinear echo suppressor for harmonics calculations
US20070036344A1 (en)Method and system for eliminating noises and echo in voice signals
CN113870881A (en)Robust Hammerstein sub-band spline self-adaptive echo cancellation method
Rombouts et al.Robust and efficient implementation of the PEM-AFROW algorithm for acoustic feedback cancellation
CN111755020B (en)Stereo echo cancellation method
CN106340302A (en)De-reverberation method and device for speech data
CN113347536B (en)Acoustic feedback suppression algorithm based on linear prediction and sub-band adaptive filtering
CN115604614B (en)System and method for local sound amplification and remote interaction by using hoisting microphone
CN1988738A (en)Apparatus to remove a voice signal and method thereof
US8194850B2 (en)Method and apparatus for voice communication
CN115579016B (en)Method and system for eliminating acoustic echo
CN117912485A (en)Speech band extension method, noise reduction audio device, and storage medium
WO2021055513A1 (en)Mitigating acoustic feedback in hearing aids with frequency warping by all-pass networks
CN114584902B (en)Method and device for eliminating nonlinear echo of intercom equipment based on volume control
DE102018129525A1 (en) NONLINEAR ACOUSTIC ECHOCOMPENSATION

Legal Events

DateCodeTitleDescription
PB01Publication
PB01Publication
SE01Entry into force of request for substantive examination
SE01Entry into force of request for substantive examination
GR01Patent grant
GR01Patent grant
[8]ページ先頭
©2009-2025 Movatter.jp