CN106816156A - A kind of enhanced method and device of audio quality - Google Patents
A kind of enhanced method and device of audio quality Download PDFInfo
- Publication number
- CN106816156A CN106816156A CN201710064271.7A CN201710064271A CN106816156A CN 106816156 A CN106816156 A CN 106816156A CN 201710064271 A CN201710064271 A CN 201710064271A CN 106816156 A CN106816156 A CN 106816156A
- Authority
- CN
- China
- Prior art keywords
- audio signal
- signal
- audio
- treatment
- carried out
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 70
- 230000005236 sound signal Effects 0.000 claims abstract description 210
- 238000001228 spectrum Methods 0.000 claims description 37
- 238000009499 grossing Methods 0.000 claims description 19
- 108090000623 proteins and genes Proteins 0.000 claims description 13
- 230000001629 suppression Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 238000010586 diagram Methods 0.000 description 8
- 230000002708 enhancing effect Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 230000007704 transition Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 230000001965 increasing effect Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000005728 strengthening Methods 0.000 description 2
- 241000209140 Triticum Species 0.000 description 1
- 235000021307 Triticum Nutrition 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000002401 inhibitory effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000007493 shaping process Methods 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
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
-
- 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
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
- G10L21/0216—Noise filtering characterised by the method used for estimating noise
-
- 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
- G10L25/21—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 power information
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Computational Linguistics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Quality & Reliability (AREA)
- Circuit For Audible Band Transducer (AREA)
- Soundproofing, Sound Blocking, And Sound Damping (AREA)
- Stereophonic System (AREA)
Abstract
The application is related to a kind of enhanced method and device of audio quality, wherein, methods described includes：Obtain the audio signal of preset format；Pre-processed for the audio signal, the pretreatment includes calculating the average signal of the audio signal Zhong Ge roads audio signal and/or carries out beam forming treatment to the audio signal；Based on the signal that pretreatment is obtained, noise suppressed treatment is carried out to the audio signal, obtained by the enhanced audio signal of tonequality.The enhanced method and device of audio quality that the application is provided, can effectively lift the audio quality of stereo microphone array.
Description
Technical field
The application is related to audio signal processing technique field, the enhanced method and device of more particularly to a kind of audio quality.
Background technology
With the development of science and technology, every field for audio quality pursuit more and more higher, the object of audio research
By initial single channel (mono), stereo (stereo), surround sound (surround) and 3D (3- are gradually transitions
Dimensional) audio.Different from SCVF single channel voice frequency, MCVF multichannel voice frequency is obtained typically by microphone array.For 3D sounds
Frequently, in order to pick up the audio of all directions, usually stereo microphone array, the array can obtain the level orientation of signal
The three-dimensional information at angle, Vertical Square parallactic angle harmony source and microphone array reference point distance.
In the prior art, the audio enhancing technology of linear microphone array and plane microphone array can be had
The effect of effect.But for stereo microphone array, prior art can't reach effective audio enhancing effect.
The content of the invention
The purpose of the application is to provide a kind of audio quality enhanced method and device, can effectively lift three-dimensional wheat
The audio quality of gram wind array.
To achieve the above object, on the one hand the application provides a kind of enhanced method of audio quality, and methods described includes：
Obtain the audio signal of preset format；Pre-processed for the audio signal, the pretreatment includes calculating the audio
The average signal of signal Zhong Ge roads audio signal and/or beam forming treatment is carried out to the audio signal；Based on pre-processing
The signal for arriving, noise suppressed treatment is carried out to the audio signal, is obtained by the enhanced audio signal of tonequality.
Further, when the pretreatment is the average signal of the calculating audio signal Zhong Ge roads audio signal, it is based on
The signal that pretreatment is obtained, carries out the step of noise suppressed is processed and specifically includes to the audio signal：According to the average letter
Number, determine the corresponding noise energy spectrum of the audio signal and signal energy spectrum；According to noise energy spectrum and signal energy
Spectrum, noise suppressed treatment is carried out to the audio signal, is obtained by the enhanced audio signal of tonequality.
Further, when the pretreatment is when carrying out beam forming to the audio signal to process, based on pre-processing
The signal for arriving, carries out the step of noise suppressed is processed and specifically includes to the audio signal：Be utilized respectively the first steering vector with
And second steering vector in opposite direction with first steering vector carries out inner product treatment to the audio signal, obtains inner product
First via audio signal and first via audio signal after treatment；Wherein, the sound can obtain according to first steering vector
The audio signal of the pre-configured orientation in frequency signal；According to the first via audio signal after inner product treatment and the second tunnel audio letter
Number, determine the corresponding noise energy spectrum of the audio signal and signal energy spectrum；According to noise energy spectrum and signal energy
Spectrum, noise suppressed treatment is carried out to the first via audio signal after inner product treatment, obtains believing by the enhanced audio of tonequality
Number.
Further, when the pretreatment is for the average signal of the calculating audio signal Zhong Ge roads audio signal and to institute
When stating audio signal and carrying out beam forming and process, based on the signal that obtains of pretreatment, noise suppressed is carried out to the audio signal
The step for the treatment of, specifically includes：Inner product treatment is carried out to the audio signal using the first steering vector, after obtaining inner product treatment
Audio signal；Wherein, the audio signal of the pre-configured orientation in the audio signal is can obtain according to first steering vector；
According to the average signal, the corresponding noise energy spectrum of the audio signal and signal energy spectrum are determined；According to the noise energy
Amount spectrum and signal energy spectrum, noise suppressed treatment is carried out to the audio signal after inner product treatment, obtains strengthening by tonequality
Audio signal.
Further, when the pretreatment is for the average signal of the calculating audio signal Zhong Ge roads audio signal and to institute
When stating audio signal and carrying out beam forming and process, based on the signal that obtains of pretreatment, noise suppressed is carried out to the audio signal
The step for the treatment of, specifically includes：Using the first steering vector and the second guiding arrow in opposite direction with first steering vector
Amount carries out inner product treatment to the audio signal, obtains first via audio signal and the second tunnel audio signal after inner product treatment；
Wherein, the audio signal of the pre-configured orientation in the audio signal is can obtain according to first steering vector；According to described flat
First via audio signal and the second tunnel audio signal after equal signal and inner product treatment, determine the corresponding noise of the audio signal
Inhibiting factor；According to the noise suppression factor, the first via audio signal after inner product treatment is carried out at noise suppressed
Reason, obtains by the enhanced audio signal of tonequality.
Further, before being pre-processed for the audio signal, methods described also includes：Obtain the audio
The sound field parameters of signal, the sound field parameters include at least one in sound bearing, sound source power and sound source divergence.
Further, estimate that the corresponding noise energy spectrum of the audio signal is specifically included：Judge Z in the sound field parameters
Size between the sound source power and first threshold of signal, when the sound source power of Z signals in the sound field parameters is more than described the
During one threshold value, estimate that the corresponding noise energy of the audio signal is composed less than the smoothing factor of Second Threshold using numerical value；Work as institute
When the sound source power for stating Z signals in sound field parameters is less than or equal to the first threshold, institute is more than or equal to using numerical value
The smoothing factor for stating Second Threshold estimates the corresponding noise energy spectrum of the audio signal.
Further, beam forming treatment is carried out to the audio signal to specifically include：According in the sound field parameters
Sound bearing determines goal orientation vector；Inner product treatment is carried out using the goal orientation vector and the audio signal, with
To the audio signal of beam forming.
Further, the signal for being obtained based on pretreatment, noise suppressed treatment is carried out to the audio signal and is specifically included：
Sound source divergence in the sound field parameters, it is determined that the Dynamic gene for carrying out noise suppressed treatment；According to what is determined
The Dynamic gene, noise suppressed treatment is carried out to the audio signal.
Further, the sound source divergence in the sound field parameters, it is determined that the tune for carrying out noise suppressed treatment
Integral divisor is specifically included：The size between the sound source divergence in the sound field parameters and the 3rd threshold value is judged, when the sound source
When divergence is more than three threshold value, Dynamic gene of the numerical value more than the 4th threshold value is determined；Sound in the sound field parameters
Source divergence be less than or equal to three threshold value when, determine numerical value less than or equal to the 4th threshold value adjustment because
Son.
To achieve the above object, on the other hand the application additionally provides a kind of enhanced method of audio quality, methods described
Including：Obtain the audio signal of preset format；Beam forming treatment is carried out for the audio signal, obtains strengthening by tonequality
Audio signal.
Further, the beam forming treatment is specifically included：Believe with the audio with reference to the steering vector of preset direction
Number inner product treatment is carried out, obtain enhanced audio signal on the preset direction.
Further, before being pre-processed for the audio signal, methods described also includes：Obtain the audio
The sound field parameters of signal, the sound field parameters include at least one in sound bearing, sound source power and sound source divergence.
Further, beam forming treatment is carried out to the audio signal to specifically include：According in the sound field parameters
Sound bearing determines goal orientation vector；Inner product treatment is carried out using the goal orientation vector and the audio signal, is obtained
Enhanced audio signal on target direction.
To achieve the above object, on the other hand the application also provides a kind of audio quality enhanced device, described device bag
Include：Audio signal acquiring unit, the audio signal for obtaining preset format；Pretreatment unit, for for audio letter
Number pre-processed, the pretreatment includes calculating the average signal of the audio signal Zhong Ge roads audio signal and/or to institute
Stating audio signal carries out beam forming treatment；Noise suppressed processing unit, for the signal obtained based on pretreatment, to the sound
Frequency signal carries out noise suppressed treatment, obtains by the enhanced audio signal of tonequality.
The enhanced method and device of a kind of audio quality that embodiment of the present invention is proposed, can be directed to the letter of preset format
Number audio enhancing treatment is carried out, can further entered with reference to sound field parameters (sound bearing, sound source power and sound source divergence)
The treatment of row noise suppressed and beam forming treatment, can effectively lift the quality of audio, reach Expected Results.
Brief description of the drawings
Fig. 1 is the enhanced method flow diagram of one implementation method sound intermediate frequency quality of the application；
Fig. 2 is the schematic diagram of four tunnel audio signals in one implementation method of the application；
Fig. 3 is another enhanced method flow diagram of implementation method sound intermediate frequency quality of the application；
Fig. 4 is another enhanced method flow diagram of implementation method sound intermediate frequency quality of the application；
Fig. 5 is another enhanced method flow diagram of implementation method sound intermediate frequency quality of the application；
Fig. 6 is another enhanced method flow diagram of implementation method sound intermediate frequency quality of the application；
Fig. 7 is another enhanced method flow diagram of implementation method sound intermediate frequency quality of the application；
Fig. 8 is the functional block diagram of the enhanced device of one implementation method sound intermediate frequency quality of the application.
Specific embodiment
In order that those skilled in the art more fully understand the technical scheme in the application, below in conjunction with the application reality
The accompanying drawing in mode is applied, the technical scheme in the application implementation method is clearly and completely described, it is clear that described
Implementation method is only a part of implementation method of the application, rather than whole implementation methods.Based on the embodiment party in the application
Formula, all other implementation method that those of ordinary skill in the art are obtained under the premise of creative work is not made all should
When the scope for belonging to the application protection.
Fig. 1 is referred to, the application implementation method provides a kind of audio quality enhanced method, and methods described includes following step
Suddenly.
S1：Obtain the audio signal of preset format.
In the present embodiment, the audio signal of the preset format can be the audio signal of Ambisonic A forms.
The audio signal of the Ambisonic A forms is four tunnel audio signals (LFU, RFD, LBD, RBU).Four tunnel audio signal
Can be as shown in Figure 2.
S2：Pre-processed for the audio signal, the pretreatment includes calculating the audio signal Zhong Ge roads sound
The average signal of frequency signal and/or beam forming treatment is carried out to the audio signal.
In the present embodiment, the audio signal of the Ambisonic A forms can be pre-processed, the pre- place
The purpose of reason is to carry out enhancing treatment to the audio signal.Specifically, in the present embodiment, the mode of pretreatment can be wrapped
Include the average signal that calculates the audio signal Zhong Ge roads audio signal and/or the audio signal is carried out at beam forming
Reason.
Wherein, the average signal x of audio signal Zhong Ge roads audio signalave(n)：
Wherein, n is the label of sampling point in audio time domain signal, and L is the frame length of Audio Signal Processing, xiN () is the i-th road sound
The time-domain signal of frequency.
Beam forming processes xbf(n)：
Wherein, θ is the azimuth in the range of [0,360], pi(θ) is the steering vector in θ directions.
The corresponding noise energy time spectrum of the audio signal is being estimated, the audio signal Zhong Ge roads audio letter can calculated
Number average signal, then can be according to the average signal, it is determined that for the smoothing factor of estimated noise energy spectrum.It is described flat
The sliding factor can for example be represented by following formula：
αs(λ, k)=αd+(1-αd)p(λ,k)
Wherein, λ represents the label of audio signal sound intermediate frequency frame, and k represents the label of audio signal intermediate-frequeney point, αs(λ, k) table
Show corresponding smoothing factor, α at specific audio frequency frame and specified frequencydSmoothing factor is represented, value is that (λ k) is represented and referred to 0.85, p
Determine corresponding average signal at audio frame and specified frequency.So, for different audio frames and frequency, different putting down can be corresponded to
The sliding factor, the smoothing factor can be determined by average signal.
In the present embodiment, the corresponding noise energy spectrum of the audio signal can be estimated according to the smoothing factor.
Specifically, the formula of estimated noise energy spectrum can be with as follows：
D (λ, k)=αs(λ,k)D(λ-1,k)+(1-αs(λ,k))|Y(λ,k)|2
Wherein, (λ, k) represents corresponding estimated noise energy spectrum at specific audio frequency frame and specified frequency to D, and (λ k) is represented Y
Audio amplitude at specific audio frequency frame and specified frequency.
In the present embodiment, Fig. 3 is referred to, beam forming treatment can also be carried out to the audio signal.Specifically,
The steering vector (steering vector) that preset direction can be combined carries out inner product treatment with the audio signal, so that can
To strengthen the audio signal on the preset direction.So just can effectively strengthen the sound source of specific direction.
In one implementation method of the application, Fig. 4 is referred to, can be composed with reference to sound field parameters estimated noise energy.Specifically
Ground, can obtain the sound field parameters of the audio signal, and the sound field parameters include sound bearing (sound location), sound
At least one in source energy (sound power) and sound source divergence (sound diffusivity).The sound field parameters
Can be obtained by direction of arrival (Direction of Arrival, DOA) method.
In the present embodiment, smoothing factor can possess different numerical value according to different audio frames and frequency, therefore can
The smoothing factor of actual use is determined with the size between the sound source power and first threshold according to Z signals in sound field parameters.
Specifically, when the sound source power of Z signals in the sound field parameters is more than the first threshold, Second Threshold is less than using numerical value
Smoothing factor estimate the corresponding noise energy spectrum of the audio signal；When the sound source power of Z signals in the sound field parameters is small
When the first threshold, the sound is estimated more than or equal to the smoothing factor of the Second Threshold using numerical value
The corresponding noise energy spectrum of frequency signal.Specifically, the smoothing factor if less than Second Threshold has multiple, can use therein
Any one smoothing factor is estimated.Likewise, having multiple if greater than or equal to the smoothing factor of Second Threshold, also may be used
Estimated with using any one smoothing factor therein.Specifically, first threshold scope is [0.3,0.6], Second Threshold
Scope is [0.05,0.4].
Wherein, Z signals are obtained according to transition matrix A：
Wherein, the transition matrix A=[a11 a12 a13 a14], the element a of the A11,a12,......,a14Value be
Constant, is determined by different sound source scenes.
The energy of Z signals is
In the present embodiment, Fig. 5 is referred to, it is also possible to carry out beam forming treatment with reference to sound field parameters.Specifically, may be used
Goal orientation vector is adaptively determined with the sound bearing in the sound field parameters, then can be led using the target
Inner product treatment is carried out with the audio signal to vector, to obtain the audio signal of beam forming.
S3：Based on the signal that pretreatment is obtained, noise suppressed treatment is carried out to the audio signal, obtain increasing by tonequality
Strong audio signal.
In the present embodiment, after being pre-processed to audio signal, the audio signal can be carried out at noise suppressed
Reason, so as to obtain by the enhanced audio signal of tonequality.Specifically, noise suppressed can be carried out using spectrum-subtraction, it is also possible to adopt
Noise suppressed is carried out with Wiener Filter Method.Wherein, spectrum-subtraction and Wiener Filter Method can be realized in a frequency domain.Noise suppressed
Process can be carried out in whole frequency band, it is also possible to be carried out in a sub-band.
In present embodiment kind, Fig. 6 is referred to, after beam forming is carried out to audio signal, noise suppressed can be carried out
Treatment.Specifically, the first steering vector and second guiding in opposite direction with first steering vector can be utilized respectively
Vector carries out inner product treatment to the audio signal, respectively obtains first via audio signal and the second tunnel audio after inner product treatment
Signal；Wherein, the audio signal of the pre-configured orientation in the audio signal is can obtain according to first steering vector；Then may be used
Frequency-region signal is transformed to respectively with by the first via audio signal after inner product treatment and the second tunnel, and is made an uproar in a frequency domain
Sound suppression is processed.
Specifically, beam forming is processed as：
Wherein, θ is the azimuth in the range of [0,360], pi(θ) is the steering vector in θ directions, xiN () is the i-th tunnel audio
Time-domain signal.
Time-domain signal is transformed to frequency-region signal, discrete Fourier transform DFT, Fast Fourier Transform (FFT) FFT can be used
Or Modified Discrete Cosine Transform MDCT is realized.
It should be noted that the application implementation method only can also carry out beam forming treatment to audio signal.Specifically,
The application implementation method provides a kind of audio quality enhanced method, and methods described includes：
Obtain the audio signal of preset format；
Beam forming treatment is carried out for the audio signal, wherein, waveform shaping treatment is specifically included：
Inner product treatment is carried out with the audio signal with reference to the steering vector of preset direction, is increased with the preset direction
The strong audio signal.
Fig. 7 is referred to, it is, of course, also possible to carry out noise suppressed treatment with reference to sound field parameters.Specifically, can be utilized respectively
In first steering vector and the second steering vector in opposite direction with first steering vector are carried out to the audio signal
Product treatment, respectively obtains the first via audio signal and the second tunnel audio signal after inner product treatment；Wherein, led according to described first
The audio signal of the pre-configured orientation in the audio signal is can obtain to vector；Then can be by first after inner product treatment
Road audio signal and the second tunnel audio signal are transformed to frequency-region signal, and the sound source diverging in the sound field parameters respectively
Degree, it is determined that the Dynamic gene for carrying out noise suppressed treatment, finally then can be according to the Dynamic gene for determining, to described
Audio signal carries out noise suppressed treatment.Specifically, in the sound source divergence in the sound field parameters, it is determined that for carrying out
In the step of Dynamic gene of noise suppressed treatment, it can be determined that sound source divergence in the sound field parameters and the 3rd threshold value it
Between size, when the sound source divergence be more than three threshold value when, determine numerical value more than the 4th threshold value Dynamic gene；When
When sound source divergence in the sound field parameters is less than or equal to three threshold value, determine numerical value less than or equal to described
The Dynamic gene of the 4th threshold value.Specifically, the 3rd threshold range is [0.3,0.5], the 4th threshold range is [0.05,0.5].
Fig. 8 is referred to, the application implementation method also provides a kind of audio quality enhanced device, and described device includes：
Audio signal acquiring unit 100, the audio signal for obtaining preset format；
Pretreatment unit 200, for being pre-processed for the audio signal, the pretreatment includes calculating the sound
The average signal of frequency signal Zhong Ge roads audio signal and/or beam forming treatment is carried out to the audio signal；
Noise suppressed processing unit 300, for the signal obtained based on pretreatment, noise suppression is carried out to the audio signal
System treatment, obtains by the enhanced audio signal of tonequality.
In one implementation method of the application, the pretreatment unit 200 is specifically included：
Average signal computing module, the average signal for calculating the audio signal Zhong Ge roads audio signal；
Smoothing factor determining module, for according to the average signal, it is determined that for estimated noise energy spectrum it is smooth because
Son；
Estimation block, for estimating the corresponding noise energy spectrum of the audio signal according to the smoothing factor.
The enhanced method and device of a kind of audio quality that embodiment of the present invention is proposed, can be directed to the letter of preset format
Number audio enhancing treatment is carried out, can further entered with reference to sound field parameters (sound bearing, sound source power and sound source divergence)
The treatment of row noise suppressed and beam forming treatment, can effectively lift the quality of audio, reach Expected Results.
Description to the various implementation methods of the application above is supplied to those skilled in the art with the purpose for describing.It is not
Be intended to exhaustion or be not intended to limit the invention to single disclosed embodiment.As described above, the application's is various
Substitute and change will be apparent for above-mentioned technology one of ordinary skill in the art.Therefore, although specifically beg for
The implementation method of some alternatives has been discussed, but other embodiment will be apparent, or those skilled in the art are relative
Easily draw.The application is intended to be included in this of the invention all replacement for having discussed, modification and change, and falls
Other embodiment in the spirit and scope of above-mentioned application.
Claims (10)
1. a kind of enhanced method of audio quality, it is characterised in that methods described includes：
Obtain the audio signal of preset format；
Pre-processed for the audio signal, the pretreatment includes calculating the audio signal Zhong Ge roads audio signal
Average signal and/or beam forming treatment is carried out to the audio signal；
Based on the signal that pretreatment is obtained, noise suppressed treatment is carried out to the audio signal, obtained by the enhanced sound of tonequality
Frequency signal.
2. the enhanced method of audio quality according to claim 1, it is characterised in that when the pretreatment is described to calculate
During the average signal of audio signal Zhong Ge roads audio signal, based on the signal that pretreatment is obtained, the audio signal is made an uproar
The step of sound suppresses treatment specifically includes：
According to the average signal, the corresponding noise energy spectrum of the audio signal and signal energy spectrum are determined；
According to noise energy spectrum and signal energy spectrum, noise suppressed treatment is carried out to the audio signal, obtained by sound
The enhanced audio signal of matter；
When the pretreatment is when carrying out beam forming to the audio signal to process, based on the signal that pretreatment is obtained, to institute
State audio signal and carry out the step of noise suppressed is processed and specifically include：
The first steering vector and second steering vector in opposite direction with first steering vector are utilized respectively to the sound
Frequency signal carries out inner product treatment, respectively obtains first via audio signal and the second tunnel audio signal after inner product treatment；Wherein, root
The audio signal of the pre-configured orientation in the audio signal is can obtain according to first steering vector；
According to first via audio signal and the second tunnel audio signal after inner product treatment, determine that the audio signal is corresponding
Noise energy is composed and signal energy spectrum；
According to noise energy spectrum and signal energy spectrum, noise suppression is carried out to the first via audio signal after inner product treatment
System treatment, obtains by the enhanced audio signal of tonequality.
3. the enhanced method of audio quality according to claim 1, it is characterised in that when the pretreatment is described to calculate
The average signal of audio signal Zhong Ge roads audio signal and the audio signal is carried out beam forming process when, based on pretreatment
The signal for obtaining, carries out the step of noise suppressed is processed and specifically includes to the audio signal：
Inner product treatment is carried out to the audio signal using the first steering vector, the audio signal after inner product treatment is obtained；Wherein,
The audio signal of the pre-configured orientation in the audio signal is can obtain according to first steering vector；
According to the average signal, the corresponding noise energy spectrum of the audio signal and signal energy spectrum are determined；
According to noise energy spectrum and signal energy spectrum, the audio signal after inner product treatment is carried out at noise suppressed
Reason, obtains by the enhanced audio signal of tonequality.
4. the enhanced method of audio quality according to claim 1, it is characterised in that based on the signal that pretreatment is obtained,
The step of noise suppressed is processed is carried out to the audio signal to specifically include：
The audio is believed using the first steering vector and the second steering vector in opposite direction with first steering vector
Number inner product treatment is carried out, obtain first via audio signal and the second tunnel audio signal after inner product treatment；Wherein, according to described
One steering vector can obtain the audio signal of the pre-configured orientation in the audio signal；
According to first via audio signal and the second tunnel audio signal after the average signal and inner product treatment, the audio is determined
The corresponding noise suppression factor of signal；
According to the noise suppression factor, noise suppressed treatment is carried out to the first via audio signal after inner product treatment, obtained
To by the enhanced audio signal of tonequality.
5. the enhanced method of audio quality according to claim 1, it is characterised in that carried out for the audio signal
Before pretreatment, methods described also includes：
The sound field parameters of the audio signal are obtained, the sound field parameters include the diverging of sound bearing, sound source power and sound source
At least one in degree；
Correspondingly, estimate that the corresponding noise energy spectrum of the audio signal is specifically included：
The size between the sound source power of Z signals in the sound field parameters and first threshold is judged, when Z letters in the sound field parameters
Number sound source power when being more than the first threshold, the audio signal is estimated less than the smoothing factor of Second Threshold using numerical value
Corresponding noise energy spectrum；
When Z signals in the sound field parameters sound source power be less than or equal to the first threshold when, using numerical value be more than or
The smoothing factor that person is equal to the Second Threshold estimates the corresponding noise energy spectrum of the audio signal；
Correspondingly, beam forming treatment is carried out to the audio signal to specifically include：
Sound bearing in the sound field parameters determines goal orientation vector；
Inner product treatment is carried out using the goal orientation vector and the audio signal, to obtain the audio signal of beam forming；
Correspondingly, the signal for being obtained based on pretreatment, noise suppressed treatment is carried out to the audio signal and is specifically included：
Sound source divergence in the sound field parameters, it is determined that the Dynamic gene for carrying out noise suppressed treatment；
According to the Dynamic gene for determining, noise suppressed treatment is carried out to the audio signal.
6. the enhanced method of audio quality according to claim 5, it is characterised in that according to the sound in the sound field parameters
Source divergence, it is determined that being specifically included for carrying out the Dynamic gene of noise suppressed treatment：
The size between the sound source divergence in the sound field parameters and the 3rd threshold value is judged, when the sound source divergence is more than institute
When stating three threshold values, Dynamic gene of the numerical value more than the 4th threshold value is determined；
When the sound source divergence in the sound field parameters is less than or equal to three threshold value, determine that numerical value is less than or waits
In the Dynamic gene of the 4th threshold value.
7. a kind of enhanced method of audio quality, it is characterised in that methods described includes：
Obtain the audio signal of preset format；
Beam forming treatment is carried out for the audio signal, is obtained by the enhanced audio signal of tonequality.
8. the enhanced method of audio quality according to claim 7, it is characterised in that beam forming treatment is specifically included：
Inner product treatment is carried out with the audio signal with reference to the steering vector of preset direction, obtains enhanced on the preset direction
Audio signal.
9. the enhanced method of audio quality according to claim 7, it is characterised in that carried out for the audio signal
Before beam forming treatment, methods described also includes：
The sound field parameters of the audio signal are obtained, the sound field parameters include the diverging of sound bearing, sound source power and sound source
At least one in degree；
Correspondingly, beam forming treatment is carried out to the audio signal to specifically include：
Sound bearing in the sound field parameters determines goal orientation vector；
Inner product treatment is carried out using the goal orientation vector and the audio signal, enhanced audio letter on target direction is obtained
Number.
10. the enhanced device of a kind of audio quality, it is characterised in that described device includes：
Audio signal acquiring unit, the audio signal for obtaining preset format；
Pretreatment unit, for being pre-processed for the audio signal, the pretreatment includes calculating the audio signal
The average signal of Zhong Ge roads audio signal and/or beam forming treatment is carried out to the audio signal；
Noise suppressed processing unit, for the signal obtained based on pretreatment, noise suppressed treatment is carried out to the audio signal,
Obtain by the enhanced audio signal of tonequality.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710064271.7A CN106816156B (en) | 2017-02-04 | 2017-02-04 | Method and device for enhancing audio quality |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710064271.7A CN106816156B (en) | 2017-02-04 | 2017-02-04 | Method and device for enhancing audio quality |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106816156A true CN106816156A (en) | 2017-06-09 |
| CN106816156B CN106816156B (en) | 2020-06-30 |
Family
ID=59111991
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710064271.7A Active CN106816156B (en) | 2017-02-04 | 2017-02-04 | Method and device for enhancing audio quality |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106816156B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107920303A (en) * | 2017-11-21 | 2018-04-17 | 北京时代拓灵科技有限公司 | A kind of method and device of audio collection |
| CN108520756A (en) * | 2018-03-20 | 2018-09-11 | 北京时代拓灵科技有限公司 | A kind of method and device of speaker's speech Separation |
| CN113077787A (en) * | 2020-12-22 | 2021-07-06 | 珠海市杰理科技股份有限公司 | Voice data identification method, device, chip and readable storage medium |
| CN113170270A (en) * | 2018-10-08 | 2021-07-23 | 诺基亚技术有限公司 | Spatial audio enhancement and reproduction |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1809105A (en) * | 2006-01-13 | 2006-07-26 | 北京中星微电子有限公司 | Dual-microphone speech enhancement method and system applicable to mini-type mobile communication devices |
| CN1953059A (en) * | 2006-11-24 | 2007-04-25 | 北京中星微电子有限公司 | A method and device for noise elimination |
| CN102227768A (en) * | 2009-01-06 | 2011-10-26 | 三菱电机株式会社 | Noise cancellation device and noise cancellation program |
| CN102801861A (en) * | 2012-08-07 | 2012-11-28 | 歌尔声学股份有限公司 | Voice enhancing method and device applied to cell phone |
| CN104065798A (en) * | 2013-03-21 | 2014-09-24 | 华为技术有限公司 | Sound signal processing method and device |
| WO2016147020A1 (en) * | 2015-03-19 | 2016-09-22 | Intel Corporation | Microphone array speech enhancement |
-
2017
- 2017-02-04 CN CN201710064271.7A patent/CN106816156B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1809105A (en) * | 2006-01-13 | 2006-07-26 | 北京中星微电子有限公司 | Dual-microphone speech enhancement method and system applicable to mini-type mobile communication devices |
| CN1953059A (en) * | 2006-11-24 | 2007-04-25 | 北京中星微电子有限公司 | A method and device for noise elimination |
| CN102227768A (en) * | 2009-01-06 | 2011-10-26 | 三菱电机株式会社 | Noise cancellation device and noise cancellation program |
| CN102801861A (en) * | 2012-08-07 | 2012-11-28 | 歌尔声学股份有限公司 | Voice enhancing method and device applied to cell phone |
| CN104065798A (en) * | 2013-03-21 | 2014-09-24 | 华为技术有限公司 | Sound signal processing method and device |
| WO2016147020A1 (en) * | 2015-03-19 | 2016-09-22 | Intel Corporation | Microphone array speech enhancement |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107920303A (en) * | 2017-11-21 | 2018-04-17 | 北京时代拓灵科技有限公司 | A kind of method and device of audio collection |
| CN107920303B (en) * | 2017-11-21 | 2019-12-24 | 北京时代拓灵科技有限公司 | Audio acquisition method and device |
| CN108520756A (en) * | 2018-03-20 | 2018-09-11 | 北京时代拓灵科技有限公司 | A kind of method and device of speaker's speech Separation |
| CN108520756B (en) * | 2018-03-20 | 2020-09-01 | 北京时代拓灵科技有限公司 | Method and device for separating speaker voice |
| CN113170270A (en) * | 2018-10-08 | 2021-07-23 | 诺基亚技术有限公司 | Spatial audio enhancement and reproduction |
| US11363403B2 (en) | 2018-10-08 | 2022-06-14 | Nokia Technologies Oy | Spatial audio augmentation and reproduction |
| US11729574B2 (en) | 2018-10-08 | 2023-08-15 | Nokia Technologies Oy | Spatial audio augmentation and reproduction |
| CN113077787A (en) * | 2020-12-22 | 2021-07-06 | 珠海市杰理科技股份有限公司 | Voice data identification method, device, chip and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106816156B (en) | 2020-06-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7011075B2 (en) | Target voice acquisition method and device based on microphone array | |
| CN106816156A (en) | A kind of enhanced method and device of audio quality | |
| US10856094B2 (en) | Method and device for sound source localization | |
| CN100524465C (en) | A method and device for noise elimination | |
| US8947978B2 (en) | System and method for estimating the direction of arrival of a sound | |
| US9552828B2 (en) | Audio signal processing device | |
| CN102402987A (en) | Noise suppression device, noise suppression method, and program | |
| US20180277135A1 (en) | Audio signal quality enhancement based on quantitative snr analysis and adaptive wiener filtering | |
| CN111081267B (en) | Multi-channel far-field speech enhancement method | |
| CN112242148B (en) | Headset-based wind noise suppression method and device | |
| US20100111329A1 (en) | Sound Processing Apparatus, Sound Processing Method and Program | |
| CN107346664A (en) | A kind of ears speech separating method based on critical band | |
| CN107742521A (en) | The coding method of multi-channel signal and encoder | |
| CN103680512B (en) | The horizontal lifting system of speech recognition and its method of vehicle array microphone | |
| CN107369460A (en) | Speech sound enhancement device and method based on acoustics vector sensor space sharpening technique | |
| CN105845150A (en) | Voice enhancement method and system adopting cepstrum to correct | |
| CN111951818B (en) | Dual-microphone voice enhancement method based on improved power difference noise estimation algorithm | |
| CN108520756A (en) | A kind of method and device of speaker's speech Separation | |
| CN113903353A (en) | Directional noise elimination method and device based on spatial discrimination detection | |
| CN103824563A (en) | Hearing aid denoising device and method based on module multiplexing | |
| CN104143337B (en) | A kind of method and apparatus improving sound signal tonequality | |
| CN113223552B (en) | Speech enhancement method, device, apparatus, storage medium, and program | |
| CN114189781A (en) | Noise reduction method and system for double-microphone neural network noise reduction earphone | |
| CN106128480B (en) | The method that a kind of pair of noisy speech carries out voice activity detection | |
| CN105719658B (en) | Wavelet packet voice de-noising method based on new threshold function table and adaptive threshold |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |