CN105845152A - Method for detecting audio signal echoes - Google Patents
Method for detecting audio signal echoes Download PDFInfo
- Publication number
- CN105845152A CN105845152A CN201610169131.1A CN201610169131A CN105845152A CN 105845152 A CN105845152 A CN 105845152A CN 201610169131 A CN201610169131 A CN 201610169131A CN 105845152 A CN105845152 A CN 105845152A
- Authority
- CN
- China
- Prior art keywords
- audio signal
- frame
- echo
- power
- value
- 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.)
- Pending
Links
- 230000005236 sound signal Effects 0.000 title claims abstract description 64
- 238000000034 method Methods 0.000 title abstract description 8
- 238000002592 echocardiography Methods 0.000 title abstract description 5
- 238000001514 detection method Methods 0.000 claims abstract description 24
- 238000001228 spectrum Methods 0.000 claims description 17
- 230000011218 segmentation Effects 0.000 claims description 9
- 230000009466 transformation Effects 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 abstract description 8
- 238000004891 communication Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000009432 framing Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000000205 computational method Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
- G10L25/24—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters the extracted parameters being the cepstrum
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/78—Detection of presence or absence of voice signals
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)
Abstract
The invention discloses a method for detecting audio signal echoes. The method comprises following steps: sectional calculation is performed on power cepstrum of the audio signal to obtain kurtosis values and the kurtosis values are compared with a preset value to determine whether there are echoes in the audio signal according to the comparison results; thereby, whether there are echoes in the audio signal transmission process is detected in real time without reference signals and the detection accuracy is high; the problems that traditional detection methods are subject to reference resources and therefore the detection accuracy is difficult to guaranteed are solved.
Description
Technical field
The present invention relates to audio transmission system technical field, be specifically related to a kind of audio signal echo detection method.
Background technology
Along with the fast development of communication technology, people are more and more higher to the requirement of communication quality, and at audio frequency
During transmission, the existence of echo is it is difficult to ensure that communication quality.In order to eliminate echo, and then improve communication
Quality, lifting Consumer's Experience, detect whether to there is echo most important in sound transmission course in real time.Pass
System echo checking system be use anechoic audio signal as reference signal, with the audio frequency of current transmission
Signal compares, thus judges whether the audio signal of current transmission exists echo, but this dependence is with reference to letter
Number detection method be limited by reference source, accuracy it is difficult to ensure that.
Summary of the invention
The technical problem to be solved is on the basis of without necessarily referring to signal, it is achieved believe at audio frequency
Number transmitting procedure detects whether to there is echo in real time, it is therefore intended that provide a kind of audio signal detection of echoes side
Method, solves current detection method and relies on reference source, there is detection inaccurate in the case of reference source instability
Problem.
The present invention is achieved through the following technical solutions:
A kind of audio signal echo detection method, comprises the steps:
A, continuous print audio signal is carried out sub-frame processing；
B, calculate the power cepstrum of every frame audio signal；
C, to power cepstrum segmentation, calculate the kurtosis numerical value of every section；
D, comparison peak number of degrees value and setting threshold size, it is judged that whether audio signal exists echo.
Especially, described step B specifically includes:
B1, the audio signal that every frame is to be measured is carried out Fourier transformation, it is thus achieved that the power spectrum of every frame audio signal；
B2, power spectrum is carried out logarithm operation；
B3, the power spectrum through logarithm operation is carried out inverse Fourier transform, it is thus achieved that power cepstrum.
Especially, described step C specifically includes:
C1, power cepstrum to every frame audio signal carry out segmentation；
C2, calculate the kurtosis numerical value of every section of power cepstrum；
C3, obtain every frame audio signal maximum peak number of degrees value.
Especially, described step D specifically includes:
D1, comparison every frame audio signal maximum peak number of degrees value and setting threshold size；
D2, when the maximum peak number of degrees value of N continuous frame audio signal is all higher than setting threshold value, then judge this sound
Frequently there is echo in signal, and N is positive integer.
The present invention compared with prior art, has such advantages as and beneficial effect:
A kind of audio signal echo detection method of the present invention, it is possible on the basis of without necessarily referring to signal,
Realize detecting whether to exist echo during audio signal transmission in real time, detection accuracy is high, and customer service passes
The detection method of system is limited by reference source, the accuracy of detection it is difficult to ensure that problem.
Accompanying drawing explanation
Accompanying drawing described herein is used for providing being further appreciated by the embodiment of the present invention, constitutes the application's
A part, is not intended that the restriction to the embodiment of the present invention.In the accompanying drawings:
The audio signal echo detection method flow chart that Fig. 1 provides for the embodiment of the present invention 1.
Detailed description of the invention
For making the object, technical solutions and advantages of the present invention clearer, below in conjunction with embodiment and attached
Figure, the present invention is described in further detail, and exemplary embodiment and the explanation thereof of the present invention are only used for
Explain the present invention, not as a limitation of the invention.
Embodiment 1
As it is shown in figure 1, the audio signal echo detection method flow chart that Fig. 1 provides for the embodiment of the present invention 1.
In the present embodiment, audio signal echo detection method specifically includes:
S101, continuous print audio signal is carried out sub-frame processing.
Audio signal can not directly carry out signal processing as quasi-steady state signal, in the present embodiment, is carrying out letter
Number process before to continuous print signal framing, every frame length about 20-30ms, the audio signal after framing is as stable state
Signal is further carried out processing.
S102, the audio signal that every frame is to be measured is carried out Fourier transformation, it is thus achieved that the power of every frame audio signal
Spectrum.
Power spectrum is the abbreviation of power spectral density function, and it is defined as the signal power in per unit band.Its table
Show that signal power is along with the situation of change of frequency, i.e. signal power are in the distribution situation of frequency domain.The present embodiment
In audio signal x (n) that certain frame is to be measured is carried out Fourier transformation, it is thus achieved that the power spectrum of audio signal is F (x (n)).
S103, power spectrum is carried out logarithm operation.
The power spectrum of audio signal x (n) is taken natural logrithm is ln (| F (x (n)) |2)。
S104, the power spectrum through logarithm operation is carried out inverse Fourier transform, it is thus achieved that power cepstrum.
Power cepstrum refers to the Fourier inversion that the Fourier transform spectrum of audio signal is carried out after logarithm operation again,
When there is echo in audio signal, can form more sharp-pointed signal in power cepstrum, the present embodiment uses
This characteristic is as the standard judging echo.For audio signal x (n), its power cepstrum PCx (n)=
F-1 [ln (| F (x (n)) | 2)], wherein, F (.) represents Fourier transformation, and F-1 (.) represents Fourier inversion, ln (.)
Represent and take natural logrithm.
S105, power cepstrum to every frame audio signal carry out segmentation.
Power cepstrum segmentation to every frame audio signal calculates kurtosis numerical value, if certain section exists echo, then should
The kurtosis numerical value of echo section can be relatively big, in this, as the standard judging echo.
S106, calculate the kurtosis numerical value of every section of power cepstrum.
Kurtosis characterizes probability distribution density curve characteristic number of peak value height at meansigma methods.Coefficient of kurtosis is the biggest,
The distribution of sample data just has more extremum, then its residual value must be more concentrated on around mode,
Its distribution is inevitable the most precipitous.
The kurtosis computational methods of the stochastic audio signal are defined as: the fourth central square of stochastic variable and the ratio of variance
Value.Wherein, central moment defines: for positive integer k, if E (X) exists, and E [| X-E (X) |k)] < ∞,
Then claim E{ [X-E (X)]kIt it is the k rank central moment of stochastic variable X.If the variance of X is the second-order moment around mean of X,
I.e. D (X)=E{ [X-E (X)]2}；Variance defines: variance is to weigh stochastic variable at theory of probability and statistical variance
Or the tolerance of dispersion degree during one group of data.In theory of probability, variance is used for measuring stochastic variable and its mathematic expectaion
Departure degree between (i.e. average).Variance (sample variance) in statistics be each data respectively with its
The difference of average square the average of sum.
In the present embodiment, the concrete formula that kurtosis calculates is:
Wherein, N is power spectrum cepstrum data number, and i is power spectrum data index, YiEach for power spectrum
Numerical value,For the meansigma methods of power spectrum,S4It is four degree variance,
S107, obtain every frame audio signal maximum peak number of degrees value.
The kurtosis numerical value calculated for power cepstrum segmentation, it is thus achieved that every frame audio signal maximum peak number of degrees value, looks for
Go out the signal segment that kurtosis value is maximum.
S108, comparison every frame audio signal maximum peak number of degrees value and setting threshold size.
The threshold value of default suitable size is as thresholding, and the peak every frame audio signal power cepstrum segmentation calculated
Number of degrees value is compared with the threshold value of setting.
S109, when the maximum peak number of degrees value of N continuous frame audio signal is all higher than setting threshold value, then judging should
There is echo in audio signal.
If N continuous frame audio signal maximum peak number of degrees value all occurs more than the threshold value set, then judge this audio frequency
There is echo in signal, otherwise it is assumed that this audio signal does not has echo.
Technical scheme by the power cepstrum segmentation of audio signal being calculated kurtosis numerical value, and with in advance
If threshold value compare, and then judge whether audio signal exists echo according to comparative result, thus not
On the basis of needing to refer to signal, it is achieved detect whether in real time to there is echo during audio signal transmission,
Detection accuracy is high, and the traditional detection method of customer service is limited by reference source, the accuracy of detection it is difficult to ensure that
Problem.
Above-described detailed description of the invention, is carried out the purpose of the present invention, technical scheme and beneficial effect
Further describe, be it should be understood that the detailed description of the invention that the foregoing is only the present invention,
The protection domain being not intended to limit the present invention, all within the spirit and principles in the present invention, that is done is any
Amendment, equivalent, improvement etc., should be included within the scope of the present invention.
Claims (4)
1. an audio signal echo detection method, it is characterised in that comprise the steps:
A, continuous print audio signal is carried out sub-frame processing；
B, calculate the power cepstrum of every frame audio signal；
C, to power cepstrum segmentation, calculate the kurtosis numerical value of every section;
D, comparison peak number of degrees value and setting threshold size, it is judged that whether audio signal exists echo.
2. audio signal echo detection method as claimed in claim 1, it is characterised in that described step B specifically includes:
B1, the audio signal that every frame is to be measured is carried out Fourier transformation, it is thus achieved that the power spectrum of every frame audio signal；
B2, power spectrum is carried out logarithm operation；
B3, the power spectrum through logarithm operation is carried out inverse Fourier transform, it is thus achieved that power cepstrum.
3. audio signal echo detection method as claimed in claim 1, it is characterised in that described step C specifically includes:
C1, power cepstrum to every frame audio signal carry out segmentation；
C2, calculate the kurtosis numerical value of every section of power cepstrum；
C3, obtain every frame audio signal maximum peak number of degrees value.
4. audio signal echo detection method as claimed in claim 1, it is characterised in that described step D specifically includes:
D1, comparison every frame audio signal maximum peak number of degrees value and setting threshold size；
D2, when the maximum peak number of degrees value of N continuous frame audio signal is all higher than setting threshold value, then judge that this audio signal exists echo, N is positive integer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610169131.1A CN105845152A (en) | 2016-03-23 | 2016-03-23 | Method for detecting audio signal echoes |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610169131.1A CN105845152A (en) | 2016-03-23 | 2016-03-23 | Method for detecting audio signal echoes |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105845152A true CN105845152A (en) | 2016-08-10 |
Family
ID=56583106
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610169131.1A Pending CN105845152A (en) | 2016-03-23 | 2016-03-23 | Method for detecting audio signal echoes |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105845152A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107360530A (en) * | 2017-07-03 | 2017-11-17 | 苏州科达科技股份有限公司 | The method of testing and device of a kind of echo cancellor |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2009029076A1 (en) * | 2007-08-31 | 2009-03-05 | Tellabs Operations, Inc. | Controlling echo in the coded domain |
| CN104052526A (en) * | 2013-03-17 | 2014-09-17 | 德克萨斯仪器股份有限公司 | Clipping Based On Cepstral Distance For Acoustic Echo Canceller |
-
2016
- 2016-03-23 CN CN201610169131.1A patent/CN105845152A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2009029076A1 (en) * | 2007-08-31 | 2009-03-05 | Tellabs Operations, Inc. | Controlling echo in the coded domain |
| CN104052526A (en) * | 2013-03-17 | 2014-09-17 | 德克萨斯仪器股份有限公司 | Clipping Based On Cepstral Distance For Acoustic Echo Canceller |
Non-Patent Citations (4)
| Title |
|---|
| WEI ZENG等: "《An Algorithm of Echo Steganalysis Based on Power Cepstrum and Pattern Classification》", 《2008 INTERNATIONAL CONFERENCE ON AUDIO,LANGUAGE AND IMAGE PROCESSING》 * |
| 周浩敏等: "《测试信号处理技术 第2版》", 31 May 2009, 北京航空航天大学出版社 * |
| 汪云路等: "《基于统计特征的语音回声隐藏分析》", 《数据采集与处理》 * |
| 管涛: "《数据分析基础及模型》", 31 July 2015, 合肥工业大学出版社 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107360530A (en) * | 2017-07-03 | 2017-11-17 | 苏州科达科技股份有限公司 | The method of testing and device of a kind of echo cancellor |
| CN107360530B (en) * | 2017-07-03 | 2020-12-01 | 苏州科达科技股份有限公司 | Echo cancellation testing method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10026418B2 (en) | Abnormal frame detection method and apparatus | |
| CN104464722B (en) | Voice activity detection method and apparatus based on time domain and frequency domain | |
| EP3696814A1 (en) | Speech enhancement method and apparatus, device and storage medium | |
| EP3493205B1 (en) | Method and apparatus for adaptively detecting a voice activity in an input audio signal | |
| US20150006162A1 (en) | Systems and methods for measuring speech signal quality | |
| CN103886865A (en) | Sound Processing Device, Sound Processing Method, And Program | |
| CN103067322A (en) | Method for evaluating voice quality of audio frame in single channel audio signal | |
| CN107070568A (en) | A kind of frequency spectrum sensing method based on Hilbert-Huang transform | |
| CN109410178A (en) | A kind of workpiece crack detection method and system | |
| CN108828570B (en) | A ranging method and ranging device based on dynamic estimation of path loss factor | |
| CN106663450A (en) | Method of and apparatus for evaluating quality of a degraded speech signal | |
| CN103578479A (en) | Speech intelligibility measuring method based on auditory masking effect | |
| CN104850752A (en) | Parameter search range determining method based on adaptive random resonance | |
| CN106685757A (en) | A method and device for evaluating network performance | |
| CN107800659A (en) | LFM signal modulation method for parameter estimation under Alpha Stable distritation noises | |
| EP3252756A1 (en) | Method and device for determining inter-channel time difference parameter | |
| CN110059294A (en) | Non-linear Granger Causality detection method based on core recurrence maximum cross-correlation entropy algorithm | |
| CN105845152A (en) | Method for detecting audio signal echoes | |
| CN110110341B (en) | Pulse detection method for automatic recommendation of decision threshold | |
| CN103970129A (en) | Control valve adhesion detecting method | |
| CN104730384A (en) | Power disturbance identification and localization method based on incomplete S transformation | |
| CN112216285A (en) | Multi-person session detection method, system, mobile terminal and storage medium | |
| CN105068973A (en) | Matrix decomposition singular value accepting or rejecting method used in frequency-response function calculation | |
| CN104243053B (en) | Input signal processing method and device | |
| CN115482835A (en) | Noise detection method, apparatus, electronic device, and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160810 |
|
| RJ01 | Rejection of invention patent application after publication |