CN106952654A - Robot noise-reduction method, device and robot - Google Patents
Robot noise-reduction method, device and robot Download PDFInfo
- Publication number
- CN106952654A CN106952654A CN201710271779.4A CN201710271779A CN106952654A CN 106952654 A CN106952654 A CN 106952654A CN 201710271779 A CN201710271779 A CN 201710271779A CN 106952654 A CN106952654 A CN 106952654A
- Authority
- CN
- China
- Prior art keywords
- signal
- audio
- noise
- audio signal
- robot
- 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
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Processing of the speech or voice signal 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
- G10L21/0316—Speech enhancement, e.g. noise reduction or echo cancellation by changing the amplitude
- G10L21/0324—Details of processing therefor
- G10L21/034—Automatic adjustment
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Processing of the speech or voice signal 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
- G10L21/0316—Speech enhancement, e.g. noise reduction or echo cancellation by changing the amplitude
- G10L21/0324—Details of processing therefor
- G10L21/0332—Details of processing therefor involving modification of waveforms
Abstract
The embodiment of the invention discloses a kind of robot noise-reduction method, device and robot, comprise the steps：Read default control source of sound sample；Obtain the audio signal of external environment, the control source of sound sample is compared with the audio signal, distinguish in the audio signal and control source of sound sample frequency identical echo signal, and the noise signal differed with the control source of sound sample frequency with described；Amplify the echo signal, scale the noise signal so that the audio power that the echo signal waveform is characterized is more than the audio power that the noise signal waveforms are characterized.By the way that the user pre-set or human audio to be renovated to scope as control source of sound sample, the control source of sound sample is compared with audio signal, distinguish echo signal and noise signal that user sends, and processing is amplified to echo signal, processing is zoomed in and out to noise signal, the power of echo signal is far longer than the power of noise signal.
Description
Technical field
The present embodiments relate to robot field, especially a kind of robot noise-reduction method, device and robot.
Background technology
With the development of mobile communication, computer and Internet technology, the miniaturization epoch of computer equipment arrive, by
In the Highgrade integration of computer software, and powerful disposal ability, it can make less intelligent apparatus that there is powerful place
Reason ability.In robot field, due to the powerful disposal ability of computer, make it possible small-sized robot of accompanying and attending to, accompany and attend to
Robot is, by voice or other interactive smart machines that can be exchanged with user, can be used as the life study of user
Good friend, in the prior art, assigns instruction or exchanges, be controlled usually using phonetic order, robot connects for the convenience of the user
The phonetic order sent by user, then performs the corresponding command, to allow the robot to accurate identifying user instruction, robot
Need to carry out noise reduction process to the mixed signal including user instruction and environmental noise after compound of receiving.
In the prior art, conventional noise-reduction method includes：Noise gate Method of Noise, sampling Method of Noise and filtering Method of Noise.Its
In, Noise gate Method of Noise sets a thresholding, such as noise peak-peak, then it is set as threshold values, less than the letter of this value
Number directly it is zeroed, the signal more than this value does not change, in this manner it is possible to which noise is eliminated.Sampling Method of Noise gathers one and made an uproar
Sound sample, it is desirable to be the simple noise of comparison, then using this sample as foundation, eliminates the noise in audio mixing.Ripple Method of Noise is first
Clearly to know that the noise in noise signal, in which frequency band, is then subtracted the signal of this frequency band by noise reduction software
It is weak, to reach noise reduction.
The inventor of the invention has found that the shortcoming of Noise gate Method of Noise is not in threshold in noise signal under study for action
Still maintain in noise in the range of value to remove.Sampling Method of Noise is abating the noise simultaneously, influences whether audio signal
In useful signal with noise same frequency, while there is certain distortion sense.Filter Method of Noise, it is same for noise with frequency range
Useful signal plays interference, therefore, and the noise reduction technology existed in the prior art can not be applied to robot noise reduction field, existing
In a high-efficient noise-reducing method that can be used in robot of exigence.
The content of the invention
The embodiment of the present invention is mainly solving the technical problems that provide robot noise-reduction method, the device for being capable of high-efficient noise-reducing
And robot.
In order to solve the above technical problems, the technical scheme that the embodiment of the invention is used is：A kind of machine is provided
Device people's noise-reduction method, comprises the steps：
Read default control source of sound sample；
The audio signal of external environment is obtained, the control source of sound sample is compared with the audio signal, distinguished
Go out in the audio signal and control source of sound sample frequency identical echo signal with described, and with the control source of sound sample frequently
The noise signal that rate is differed；
Amplify the echo signal, scale the noise signal so that the audio power that the echo signal waveform is characterized
The audio power characterized more than the noise signal waveforms.
Alternatively, the amplification echo signal, scales the noise signal so that echo signal waveform institute table
Also include before the step of audio power levied is more than the audio power of noise signal waveforms sign：
The characteristic parameter of the noise signal is obtained, the characteristic parameter includes frequency range, frequency spectrum, amplitude or power features；
According to the characteristic parameter of the noise signal, noise reduction is carried out to the noise signal.
Alternatively, the first voice acquisition device and the second voice collecting are at least included using the robot of the noise-reduction method
Device, the characteristic parameter of the acquisition noise signal, the characteristic parameter includes frequency range, frequency spectrum, amplitude or power features
The step of before also include：
The audio mixing being distributed in the direction indicated by first voice acquisition device and the detection of the second voice acquisition device
Signal, thus produces the first audio signal exported by the first voice acquisition device, and the output of the second voice acquisition device the
Two audio signals；
The difference of the amplitude of amplitude or power based on first audio signal and the second audio signal is described mixed to decay
Message number.
Alternatively, according to the characteristic parameter of the noise signal, the step of carrying out noise reduction to the noise signal includes：
The characteristic parameter for the audio signal that first audio collecting device is gathered is input in default noise reduction model；
The audio signal gathered using the noise reduction model to first audio collecting device carries out noise reduction.
Alternatively, the audio signal gathered using the noise reduction model to first audio collecting device is dropped
The step of making an uproar includes：
Detect whether the frequency range of the audio signal of the first audio collecting device collection is located in voice band limits；
If the frequency range of the audio signal of the first audio collecting device collection is located in the voice band limits, right
The audio signal of the first audio collecting device collection is amplified；
If the frequency range of the audio signal of the first audio collecting device collection is not located in the voice band limits,
The audio signal that first audio collecting device is gathered is zoomed in and out.
Alternatively, the characteristic parameter of the audio signal that first audio collecting device is gathered is input to default
Step in noise reduction model includes：
The drop is determined according to the difference of the amplitude of the amplitude or power of first audio signal and the second audio signal
Make an uproar the frequency of model.
Alternatively, the frequency of the noise reduction model is low frequency.
Alternatively, the angle between first audio collecting device and the second audio collecting device extended line is more than 60 °.
In order to solve the above technical problems, the embodiment of the present invention also provides a kind of robot denoising device, including：
Read module, for reading default control source of sound sample；
Obtain comparison module, the audio signal for obtaining external environment, by the control source of sound sample and the audio mixing
Signal is compared, and distinguishes in the audio signal and controls source of sound sample frequency identical echo signal, Yi Jiyu with described
The noise signal that the control source of sound sample frequency is differed；
Except module of making an uproar, for amplifying the echo signal, the noise signal is scaled so that the echo signal waveform institute
The audio power of sign is more than the audio power that the noise signal waveforms are characterized.
Alternatively, the robot denoising device also includes：
First acquisition submodule, the characteristic parameter for obtaining the noise signal, the characteristic parameter includes frequency range, frequency
Spectrum, amplitude or power features；
First noise reduction submodule, for the characteristic parameter according to the noise signal, noise reduction is carried out to the noise signal.
Alternatively, the robot denoising device also includes：
First voice acquisition device and the second voice acquisition device, pass through first voice acquisition device and the second voice
The audio signal that harvester detection is distributed in the direction indicated, thus produces the first sound exported by the first voice acquisition device
Frequency signal, and the second audio signal that the second voice acquisition device is exported；
First decay submodule, for amplitude or the width of power based on first audio signal and the second audio signal
The difference of degree is come the audio signal that decays.
Alternatively, the robot denoising device also includes：
First input submodule, for the characteristic parameter input for the audio signal for gathering first audio collecting device
Into default noise reduction model；
Second noise reduction submodule, for the audio mixing letter gathered using the noise reduction model to first audio collecting device
Number carry out noise reduction.
Alternatively, the robot denoising device also includes：
First monitors judging submodule, and the frequency range for detecting the audio signal that first audio collecting device is gathered is
It is no to be located in voice band limits；If the frequency range of the audio signal of the first audio collecting device collection is located at people's audio frequency
In segment limit, then the audio signal that first audio collecting device is gathered is amplified；If the first audio collection dress
The frequency range for putting the audio signal of collection is not located in the voice band limits, then first audio collecting device is gathered
Audio signal is zoomed in and out.
Alternatively, the robot denoising device also includes：
First determination sub-module, for the amplitude or the width of power according to first audio signal and the second audio signal
The difference of degree determines the frequency of the noise reduction model.
Alternatively, the frequency of the noise reduction model is low frequency.
Alternatively, the angle between first audio collecting device and the second audio collecting device extended line is more than 60 °.
In order to solve the above technical problems, the embodiment of the present invention also provides a kind of robot, including：
One or more processors；
Memory；
One or more application programs, wherein one or more of application programs are stored in the memory and quilt
It is configured to by one or more of computing devices, one or more of programs are configured to：
Read default control source of sound sample；
The audio signal of external environment is obtained, the control source of sound sample is compared with the audio signal, distinguished
Go out in the audio signal and control source of sound sample frequency identical echo signal with described, and with the control source of sound sample frequently
The noise signal that rate is differed；
Amplify the echo signal, scale the noise signal so that the audio power that the echo signal waveform is characterized
The audio power characterized more than the noise signal waveforms.
Alternatively, one or more of programs are additionally configured to：
The characteristic parameter of the noise signal is obtained, the characteristic parameter includes frequency range, frequency spectrum, amplitude or power features；
According to the characteristic parameter of the noise signal, noise reduction is carried out to the noise signal.
Alternatively, the robot at least includes the first voice acquisition device and the second voice acquisition device, the acquisition
The characteristic parameter of the noise signal；
One or more of programs are additionally configured to：
The audio mixing being distributed in the direction indicated by first voice acquisition device and the detection of the second voice acquisition device
Signal, thus produces the first audio signal exported by the first voice acquisition device, and the output of the second voice acquisition device the
Two audio signals；
The difference of the amplitude of amplitude or power based on first audio signal and the second audio signal is described mixed to decay
Message number.
Alternatively, one or more of programs are additionally configured to：
The characteristic parameter for the audio signal that first audio collecting device is gathered is input in default noise reduction model；
The audio signal gathered using the noise reduction model to first audio collecting device carries out noise reduction.
Alternatively, one or more of programs are additionally configured to：
Detect whether the frequency range of the audio signal of the first audio collecting device collection is located in voice band limits；
If the frequency range of the audio signal of the first audio collecting device collection is located in the voice band limits, right
The audio signal of the first audio collecting device collection is amplified；
If the frequency range of the audio signal of the first audio collecting device collection is not located in the voice band limits,
The audio signal that first audio collecting device is gathered is zoomed in and out.
Alternatively, one or more of programs are additionally configured to：
The drop is determined according to the difference of the amplitude of the amplitude or power of first audio signal and the second audio signal
Make an uproar the frequency of model.
Alternatively, the frequency of the noise reduction model is low frequency.
Alternatively, the angle between first audio collecting device and the second audio collecting device extended line is more than 60 °.
The embodiment of the present invention has the beneficial effect that：The present embodiment is by the way that the user pre-set or human audio are renovated
The control source of sound sample is compared as control source of sound sample, distinguishes the target that user sends by scope with audio signal
Signal and noise signal, and processing is amplified to echo signal, processing is zoomed in and out to noise signal, at a place for expanding a contracting
Under reason mode, the power of echo signal is far longer than the power of noise signal, realized in the case where not losing echo signal
Noise reduction, can make robot clearly control instruction of discriminating user in complicated external environment, improve robot control
The sensitivity of system and stability.
Brief description of the drawings
Technical scheme in order to illustrate more clearly the embodiments of the present invention, below will be to that will make needed for embodiment description
Accompanying drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for
For those skilled in the art, on the premise of not paying creative work, it can also be obtained according to these accompanying drawings other attached
Figure.
Fig. 1 is embodiment of the present invention robot noise-reduction method basic flow sheet；
Fig. 2 is a kind of basic flow sheet of embodiment of embodiment of the present invention noise-reduction method；
Fig. 3 is the basic flow sheet of the embodiment of the present invention the first audio collecting device noise reduction；
Fig. 4 is embodiment of the present invention robot denoising device basic structure block diagram；
Fig. 5 is the structured flowchart of embodiment of the present invention robot.
Embodiment
In order that those skilled in the art more fully understand the present invention program, below in conjunction with the embodiment of the present invention
Accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described.
In some flows of description in description and claims of this specification and above-mentioned accompanying drawing, contain according to
Particular order occur multiple operations, but it should be clearly understood that these operation can not herein occur according to it is suitable
Sequence is performed or performed parallel, and the sequence number such as 101,102 etc. of operation is only used for distinguishing each different operation, sequence number
Any execution sequence is not represented for itself.In addition, these flows can include more or less operations, and these operations can
To perform or perform parallel in order.It should be noted that the description such as " first ", " second " herein, is to be used to distinguish not
Same message, equipment, module etc., does not represent sequencing, it is different types also not limit " first " and " second ".
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.It is based on
Embodiment in the present invention, the every other implementation that those skilled in the art are obtained under the premise of creative work is not made
Example, belongs to the scope of protection of the invention.
Embodiment 1
Referring to Fig. 1, Fig. 1 is the present embodiment robot noise-reduction method basic flow sheet.
As shown in figure 1, a kind of robot noise-reduction method, comprises the steps：
S1100, the default control source of sound sample of reading.The default control source of sound sample in the controller of robot, control
Source of sound sample is the characteristic information by robot phonetic order of multi collect user in quiet environment, the feature ginseng
Number includes frequency range, frequency spectrum, amplitude or power features, the control source of sound sample generated using aforesaid way, with more confidentiality,
Robot is only controlled by the control of specific one or several control source of sound sample of users under general environment.In some choosings
In selecting property embodiment, control source of sound sample is generated by mankind acoustic characteristic itself, the feature letter in control source of sound sample
Breath is not the audio feature information of single or limited user, but the characteristic information that human audio has, and is had more
Universal effect is convenient to be promoted.
S1200, the audio signal for obtaining external environment, the control source of sound sample is compared with the audio signal
It is right, distinguish in the audio signal with it is described control source of sound sample frequency identical echo signal, and with the control sound
The noise signal that source sample frequency is differed.The acquisition external world of robot includes ring around effective target control signal and robot
The audio signal in border, and the characteristic information in the audio signal is compared with controlling the characteristic information of source of sound sample, point
Discern with control source of sound sample characteristics information identical and the audio signal with continuity, what the audio signal sent for user
Echo signal, with controlling the noise signal that the signal that source of sound sample characteristics information is differed is surrounding.
S1300, the amplification echo signal, scale the noise signal so that the sound that the echo signal waveform is characterized
Frequency power is more than the audio power that the noise signal waveforms are characterized.Put by the signal amplifier or signal of robot interior
Big circuit, processing is amplified to echo signal, and processing is zoomed in and out to noise signal by scaler or scaling circuit, is made
The audio power that echo signal waveform is characterized is much larger than the audio power that the noise signal waveforms are characterized, and noise reduction is realized with this
Function.
Above-mentioned embodiment by the user pre-set or human audio are renovated scope as control source of sound sample,
The control source of sound sample is compared with audio signal, echo signal and noise signal that user sends is distinguished, and to mesh
Mark signal is amplified processing, and processing is zoomed in and out to noise signal, expands one under the processing mode of a contracting, makes echo signal
Power is far longer than the power of noise signal, and noise reduction is realized in the case where not losing echo signal, and robot can be made to exist
The clearly control instruction of discriminating user in complicated external environment, improves sensitivity and the stability of robot control.
Referring to Fig. 2, basic flow sheets of the Fig. 2 for a kind of embodiment of the present embodiment noise-reduction method.
As shown in Fig. 2 also including before step S1300：
S1210, the characteristic parameter for obtaining the noise signal, the characteristic parameter include frequency range, frequency spectrum, amplitude or power
Feature.For each voice signal, the frequency of the voice signal can be obtained, can be according to adopting for each voice signal
Multiple frequencies of collection determine a frequency range, and frequency range is frequency range, and the width of sound is capable of determining that according to the amplitude of sound waveform
Degree, the calculating to whole audio signal energies can be derived that the power features of audio.
S1220, the characteristic parameter according to the noise signal, noise reduction is carried out to the noise signal.
In some selective embodiments, at least include the first voice acquisition device using the robot of the noise-reduction method
With the second voice acquisition device, detected in the direction indicated by first voice acquisition device and the second voice acquisition device
The audio signal of distribution, thus produces the first audio signal exported by the first voice acquisition device, and the second voice collecting dress
Put the second audio signal of output.Robot interior is provided with voice collecting at least two voice acquisition devices, the present embodiment and filled
Put and specifically refer to microphone.It is main microphon to determine the first voice acquisition device, and the second voice acquisition device is secondary microphone.By
In for main microphon, the voice signal that the main microphon is collected can generally include voice signal (i.e. voice) and make an uproar
Acoustical signal, and because main microphon and each secondary microphone are to gather the voice signal in environment simultaneously, therefore main microphon
The noise signal collected would generally also be gathered by other at least one secondary microphones, therefore, utilize these secondary microphones and master
The combination of microphone can get more accurately noise signal, and then main microphon is gathered using accurate noise signal
The degree of accuracy during voice signal noise reduction arrived can be higher.
Also include before step S1210：Amplitude or power based on first audio signal and the second audio signal
The difference of amplitude is come the audio signal that decays.
When being propagated in atmosphere due to sound, it can decay with propagation, namely the distance of sound transmission is longer, sound
Volume is smaller, thus the volume of voice signal that collects of microphone and the microphone to the sound source of the generation voice signal away from
From into positive correlation, that is, when microphone to the sound source for producing the voice signal it is in small distance when, the microphone is collected
Voice signal volume it is then larger, conversely, when microphone to produce the voice signal sound source it is in larger distance when, the Mike
The volume for the voice signal that wind is collected is then smaller.Measure major-minor microphone power (in each frequency range).First, second sound
The power of each passage in frequency signal is determined independently by decision maker.The power difference of major-minor microphone is calculated,
Then the difference of power is calculated by public function.And corresponding audio damping is carried out to the first audio signal according to the difference,
Then compare the power of major-minor microphone, accordingly erase handling carrying out noise reduction beyond part to power.In some selectivity
In embodiment, decay calculation is carried out to the first audio signal by audio band signal, frequency spectrum or amplitude.
Referring to Fig. 3, Fig. 3 is the basic flow sheet of the present embodiment the first audio collecting device noise reduction.
S1221, the characteristic parameter for the audio signal for gathering first audio collecting device are input to default noise reduction
In model.First audio collecting device is the main microphon in present embodiment, is that robot audio is that other main signal comes
Source, the characteristic parameter of the first audio signal that the first audio collecting device is exported is extracted, and by the characteristic parameter of extraction
It is put into noise reduction model and carries out noise reduction.Noise reduction model is by being constituted to a certain class or a certain personal audio feature information
Interval or scope Mathematical Modeling.
S1222, the audio signal gathered using the noise reduction model to first audio collecting device carry out noise reduction.Root
The audio signal cutting gathered according to the threshold interval of several characteristic informations in noise reduction model to the first audio collecting device, will
The audio signal in the threshold interval of several characteristic informations of noise reduction model does not carry out software processing, with this by the first audio
The equal specification of indices of signal reaches the purpose of noise reduction with this in certain interval range.
Specifically, detect whether the frequency range of the audio signal of the first audio collecting device collection is located at people's audio frequency range model
In enclosing；If the frequency range of the audio signal of the first audio collecting device collection is located in the voice band limits, to institute
The audio signal for stating the collection of the first audio collecting device is amplified；If the audio signal of the first audio collecting device collection
Frequency range not be located at the voice band limits in, then to first audio collecting device gather audio signal contract
Put.
In some new embodiments of selection, robot can receive the control of different signal source, including user speech and nothing
Line remote control.Pre-stored noise reduction model includes voice noise reduction model and wireless noise reduction model needs and carries out selectivity by calculating
Use noise reduction model.Institute is determined according to the difference of the amplitude of the amplitude or power of first audio signal and the second audio signal
State the frequency of noise reduction model.
Present embodiment is low frequency noise reduction model with the noise reduction model for voice de-noising.
In some selective embodiments, the folder between the first audio collecting device and the second audio collecting device extended line
Angle is more than 60 °.When angle between first audio collecting device and the second audio collecting device is more than 60 °, adopt the first audio
The deamplification difference of acquisition means and the second audio collecting device is larger, can increase noise reduction reliability to greatest extent.
A kind of robot denoising device is also provided to solve the above-mentioned technical problem present invention.Referring to Fig. 4, Fig. 4 is machine
People's denoising device basic structure block diagram.
As shown in figure 4, a kind of robot denoising device, including：Read module 2100, obtain comparison module 2200 and except making an uproar
Module 2300.Wherein, read module 2100 is used to read default control source of sound sample；Obtaining comparison module 2200 is used to obtain
The audio signal of external environment, the control source of sound sample is compared with the audio signal, distinguishes the audio mixing letter
With the control source of sound sample frequency identical echo signal in number, and with the making an uproar of differing of control source of sound sample frequency
Acoustical signal；Except module 2300 of making an uproar is used to amplify the echo signal, the noise signal is scaled so that the echo signal waveform
The audio power characterized is more than the audio power that the noise signal waveforms are characterized.
The present embodiment is by the way that the user pre-set or human audio to be renovated to scope as control source of sound sample, by this
Control source of sound sample is compared with audio signal, distinguishes echo signal and noise signal that user sends, and target is believed
Number processing is amplified, processing is zoomed in and out to noise signal, expanded one under the processing mode of a contracting, make the power of echo signal
It is far longer than the power of noise signal, noise reduction is realized in the case where not losing echo signal, robot can be made in complexity
External environment in clearly discriminating user control instruction, improve sensitivity and the stability of robot control.
In certain embodiments, robot denoising device also includes：First acquisition submodule and the first noise reduction submodule.Its
In, the first acquisition submodule is used for the characteristic parameter for obtaining the noise signal, and the characteristic parameter includes frequency range, frequency spectrum, width
Degree or power features；First noise reduction submodule is used for the characteristic parameter according to the noise signal, and the noise signal is carried out
Noise reduction.
In certain embodiments, robot denoising device also includes：First voice acquisition device, the second voice acquisition device
With the first decay submodule.First voice acquisition device and the second voice acquisition device, pass through first voice acquisition device
The audio signal being distributed in the direction indicated is detected with the second voice acquisition device, is thus produced defeated by the first voice acquisition device
The first audio signal gone out, and the second audio signal that the second voice acquisition device is exported；First decay submodule, for based on
The difference of the amplitude of first audio signal and the second audio signal or the amplitude of power is come the audio signal that decays.
In certain embodiments, robot denoising device also includes：First input submodule and the second noise reduction submodule.Its
In, the first input submodule is default for the characteristic parameter of the audio signal of first audio collecting device collection to be input to
Noise reduction model in；Second noise reduction submodule is used to mix the first audio collecting device collection using the noise reduction model
Message number carries out noise reduction.
In certain embodiments, the first monitoring judging submodule, for detecting the first audio collecting device collection
Whether the frequency range of audio signal is located in voice band limits；If the frequency of the audio signal of the first audio collecting device collection
Section is located in the voice band limits, then the audio signal that first audio collecting device is gathered is amplified；If institute
The frequency range for stating the audio signal of the first audio collecting device collection is not located in the voice band limits, then to first sound
The audio signal of frequency harvester collection is zoomed in and out.
In some embodiments, robot denoising device also includes：First determination sub-module, for according to described first
The difference of the amplitude of the amplitude or power of audio signal and the second audio signal determines the frequency of the noise reduction model.
In some embodiments, the frequency of noise reduction model is low frequency.
In certain embodiments, the angle between first audio collecting device and the second audio collecting device extended line
More than 60 °.
A kind of robot is also provided to solve above-mentioned technical problem the present embodiment.
The present embodiment provides the embodiment of robot.Specifically referring to Fig. 5, Fig. 5 is the structured flowchart of robot.
Referring to Fig. 5, robot includes：One or more processors 3110, the image collection sensor of memory 3120
3120 and voice collecting sensor 3130, wherein, image collection sensor 3120 and voice collecting sensor 3130 are connected to place
Manage on device 3110.；One or more application programs, wherein one or more application programs are stored in memory and are configured
To be performed by one or more processors, one or more programs are configured to：
Read default control source of sound sample；
The audio signal of external environment is obtained, the control source of sound sample is compared with the audio signal, distinguished
Go out in the audio signal and control source of sound sample frequency identical echo signal with described, and with the control source of sound sample frequently
The noise signal that rate is differed；
Amplify the echo signal, scale the noise signal so that the audio power that the echo signal waveform is characterized
The audio power characterized more than the noise signal waveforms.
Robot in above-described embodiment, by the way that the user pre-set or human audio are renovated into scope as control
Source of sound sample, the control source of sound sample is compared with audio signal, distinguishes echo signal and noise letter that user sends
Number, and processing is amplified to echo signal, processing is zoomed in and out to noise signal, expands one under the processing mode of a contracting, makes
The power of echo signal is far longer than the power of noise signal, and noise reduction is realized in the case where not losing echo signal, can
Make robot clearly control instruction of discriminating user in complicated external environment, improve the sensitivity of robot control with
Stability.
It is to be noted that in this implementation row, the memory memory storage of robot is used to realize that robot to drop in the present embodiment
All programs in method for de-noising, processor can call the program in the memory, perform listed by above-mentioned robot noise-reduction method
The institute of act is functional.Because the function robot noise-reduction method in the present embodiment that robot is realized has been described in detail, herein
No longer repeated.
It should be noted that the preferred embodiment of the present invention is given in the specification and its accompanying drawing of the present invention, still,
The present invention can be realized by many different forms, however it is not limited to the embodiment described by this specification, these embodiments
There is provided the purpose of these embodiments it is to make the understanding to the disclosure not as the extra limitation to present invention
It is more thorough comprehensive.Also, above-mentioned each technical characteristic continues to be mutually combined, the various embodiments not being enumerated above are formed,
It is considered as the scope of description of the invention record；Further, for those of ordinary skills, can be according to the above description
Improved or converted, and all these modifications and variations should all belong to the protection domain of appended claims of the present invention.
Claims (10)
1. a kind of robot noise-reduction method, it is characterised in that comprise the steps：
Read default control source of sound sample；
The audio signal of external environment is obtained, the control source of sound sample is compared with the audio signal, institute is distinguished
State in audio signal and control source of sound sample frequency identical echo signal with described, and with the control source of sound sample frequency not
Identical noise signal；
Amplify the echo signal, scale the noise signal so that the audio power that the echo signal waveform is characterized is more than
The audio power that the noise signal waveforms are characterized.
2. robot noise-reduction method according to claim 1, it is characterised in that the amplification echo signal, scaling
The noise signal is so that the audio power that the echo signal waveform is characterized is more than the sound that the noise signal waveforms are characterized
Also include before the step of frequency power：
The characteristic parameter of the noise signal is obtained, the characteristic parameter includes frequency range, frequency spectrum, amplitude or power features；
According to the characteristic parameter of the noise signal, noise reduction is carried out to the noise signal.
3. robot noise-reduction method according to claim 2, it is characterised in that use the robot of the noise-reduction method extremely
Include the first voice acquisition device and the second voice acquisition device less, the characteristic parameter of the acquisition noise signal is described
Also include before the step of characteristic parameter includes frequency range, frequency spectrum, amplitude or power features：
The audio signal being distributed in the direction indicated by first voice acquisition device and the detection of the second voice acquisition device,
Thus the first audio signal exported by the first voice acquisition device, and the second audio that the second voice acquisition device is exported are produced
Signal；
The difference of the amplitude of amplitude or power based on first audio signal and the second audio signal is come the audio mixing letter of decaying
Number.
4. robot noise-reduction method according to claim 3, it is characterised in that joined according to the feature of the noise signal
Number, the step of carrying out noise reduction to the noise signal includes：
The characteristic parameter for the audio signal that first audio collecting device is gathered is input in default noise reduction model；
The audio signal gathered using the noise reduction model to first audio collecting device carries out noise reduction.
5. robot noise-reduction method according to claim 4, it is characterised in that described to use the noise reduction model to described
The step of audio signal of first audio collecting device collection carries out noise reduction includes：
Detect whether the frequency range of the audio signal of the first audio collecting device collection is located in voice band limits；
If the frequency range of the audio signal of the first audio collecting device collection is located in the voice band limits, to described
The audio signal of first audio collecting device collection is amplified；
If the frequency range of the audio signal of the first audio collecting device collection is not located in the voice band limits, to institute
The audio signal for stating the collection of the first audio collecting device is zoomed in and out.
6. robot noise-reduction method according to claim 5, it is characterised in that described by first audio collecting device
The step that the characteristic parameter of the audio signal of collection is input in default noise reduction model includes：
The noise reduction mould is determined according to the difference of the amplitude of the amplitude or power of first audio signal and the second audio signal
The frequency of type.
7. robot noise-reduction method according to claim 6, it is characterised in that the frequency of the noise reduction model is low frequency.
8. the robot noise-reduction method according to claim 3~7 any one, it is characterised in that first audio is adopted
Angle between acquisition means and the second audio collecting device extended line is more than 60 °.
9. a kind of robot denoising device, it is characterised in that including：
Read module, for reading default control source of sound sample；
Obtain comparison module, the audio signal for obtaining external environment, by the control source of sound sample and the audio signal
Be compared, distinguish in the audio signal with it is described control source of sound sample frequency identical echo signal, and with it is described
The noise signal that control source of sound sample frequency is differed；
Except module of making an uproar, for amplifying the echo signal, the noise signal is scaled so that the echo signal waveform is characterized
Audio power be more than the noise signal waveforms characterize audio power.
10. a kind of robot, it is characterised in that including：
One or more processors；
Memory；
One or more application programs, wherein one or more of application programs are stored in the memory and are configured
For by one or more of computing devices, one or more of programs are configured to：
Read default control source of sound sample；
The audio signal of external environment is obtained, the control source of sound sample is compared with the audio signal, institute is distinguished
State in audio signal and control source of sound sample frequency identical echo signal with described, and with the control source of sound sample frequency not
Identical noise signal；
Amplify the echo signal, scale the noise signal so that the audio power that the echo signal waveform is characterized is more than
The audio power that the noise signal waveforms are characterized.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710271779.4A CN106952654A (en) | 2017-04-24 | 2017-04-24 | Robot noise-reduction method, device and robot |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710271779.4A CN106952654A (en) | 2017-04-24 | 2017-04-24 | Robot noise-reduction method, device and robot |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106952654A true CN106952654A (en) | 2017-07-14 |
Family
ID=59476677
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710271779.4A Pending CN106952654A (en) | 2017-04-24 | 2017-04-24 | Robot noise-reduction method, device and robot |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106952654A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107481732A (en) * | 2017-08-31 | 2017-12-15 | 广东小天才科技有限公司 | Noise-reduction method, device and terminal device in a kind of spoken test and appraisal |
| CN110097872A (en) * | 2019-04-30 | 2019-08-06 | 维沃移动通信有限公司 | A kind of audio-frequency processing method and electronic equipment |
| CN110534110A (en) * | 2018-05-25 | 2019-12-03 | 深圳市优必选科技有限公司 | A kind of robot and the method, apparatus and circuit for improving its interactive voice discrimination |
| CN111128192A (en) * | 2019-12-31 | 2020-05-08 | 深圳市优必选科技股份有限公司 | Voice recognition noise reduction method, system, mobile terminal and storage medium |
| CN111951802A (en) * | 2020-08-10 | 2020-11-17 | 山东金东数字创意股份有限公司 | Visual expression system and method based on AI voice recognition |
| WO2020238703A1 (en) * | 2019-05-28 | 2020-12-03 | 晶晨半导体(上海)股份有限公司 | Method and device for obtaining voice signals |
| CN113038338A (en) * | 2021-03-22 | 2021-06-25 | 联想(北京)有限公司 | Noise reduction processing method and device |
| CN113707140A (en) * | 2021-08-05 | 2021-11-26 | 科沃斯机器人股份有限公司 | Robot and control method thereof |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2108391U (en) * | 1991-08-21 | 1992-06-24 | 周云正 | Wireless chinese phonetics megaphone |
| CN101952884A (en) * | 2008-02-27 | 2011-01-19 | 索尼爱立信移动通讯有限公司 | Electronic devices and methods that adapt filtering of a microphone signal responsive to recognition of a targeted speaker's voice |
| CN102300140A (en) * | 2011-08-10 | 2011-12-28 | 歌尔声学股份有限公司 | Speech enhancing method and device of communication earphone and noise reduction communication earphone |
| CN102543097A (en) * | 2012-01-16 | 2012-07-04 | 华为终端有限公司 | Denoising method and equipment |
| JP5957964B2 (en) * | 2012-03-02 | 2016-07-27 | ヤマハ株式会社 | Sound processing apparatus and sound processing method |
| CN106340306A (en) * | 2016-11-04 | 2017-01-18 | 厦门盈趣科技股份有限公司 | Method and device for improving speech recognition degree |
| CN106558315A (en) * | 2016-12-02 | 2017-04-05 | 深圳撒哈拉数据科技有限公司 | Heterogeneous mike automatic gain calibration method and system |
-
2017
- 2017-04-24 CN CN201710271779.4A patent/CN106952654A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2108391U (en) * | 1991-08-21 | 1992-06-24 | 周云正 | Wireless chinese phonetics megaphone |
| CN101952884A (en) * | 2008-02-27 | 2011-01-19 | 索尼爱立信移动通讯有限公司 | Electronic devices and methods that adapt filtering of a microphone signal responsive to recognition of a targeted speaker's voice |
| CN102300140A (en) * | 2011-08-10 | 2011-12-28 | 歌尔声学股份有限公司 | Speech enhancing method and device of communication earphone and noise reduction communication earphone |
| CN102543097A (en) * | 2012-01-16 | 2012-07-04 | 华为终端有限公司 | Denoising method and equipment |
| JP5957964B2 (en) * | 2012-03-02 | 2016-07-27 | ヤマハ株式会社 | Sound processing apparatus and sound processing method |
| CN106340306A (en) * | 2016-11-04 | 2017-01-18 | 厦门盈趣科技股份有限公司 | Method and device for improving speech recognition degree |
| CN106558315A (en) * | 2016-12-02 | 2017-04-05 | 深圳撒哈拉数据科技有限公司 | Heterogeneous mike automatic gain calibration method and system |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107481732A (en) * | 2017-08-31 | 2017-12-15 | 广东小天才科技有限公司 | Noise-reduction method, device and terminal device in a kind of spoken test and appraisal |
| CN107481732B (en) * | 2017-08-31 | 2020-10-02 | 广东小天才科技有限公司 | Noise reduction method and device in spoken language evaluation and terminal equipment |
| CN110534110A (en) * | 2018-05-25 | 2019-12-03 | 深圳市优必选科技有限公司 | A kind of robot and the method, apparatus and circuit for improving its interactive voice discrimination |
| CN110097872A (en) * | 2019-04-30 | 2019-08-06 | 维沃移动通信有限公司 | A kind of audio-frequency processing method and electronic equipment |
| WO2020238703A1 (en) * | 2019-05-28 | 2020-12-03 | 晶晨半导体(上海)股份有限公司 | Method and device for obtaining voice signals |
| CN111128192A (en) * | 2019-12-31 | 2020-05-08 | 深圳市优必选科技股份有限公司 | Voice recognition noise reduction method, system, mobile terminal and storage medium |
| CN111951802A (en) * | 2020-08-10 | 2020-11-17 | 山东金东数字创意股份有限公司 | Visual expression system and method based on AI voice recognition |
| CN113038338A (en) * | 2021-03-22 | 2021-06-25 | 联想(北京)有限公司 | Noise reduction processing method and device |
| CN113707140A (en) * | 2021-08-05 | 2021-11-26 | 科沃斯机器人股份有限公司 | Robot and control method thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106952654A (en) | Robot noise-reduction method, device and robot | |
| KR101678549B1 (en) | Method and apparatus for providing user interface using surface acoustic signal, and device with the user interface | |
| JP4854533B2 (en) | Acoustic judgment method, acoustic judgment device, and computer program | |
| CN101430882B (en) | Method and apparatus for restraining wind noise | |
| CN105259459B (en) | Automation quality detecting method, device and the equipment of a kind of electronic equipment | |
| CN1443349A (en) | Method and apparatus for removing noise from electronic signals | |
| CN110444202B (en) | Composite voice recognition method, device, equipment and computer readable storage medium | |
| CN101023469A (en) | Digital filtering method, digital filtering equipment | |
| CN109285556B (en) | Audio processing method, device, equipment and storage medium | |
| CN106328130A (en) | Robot voice addressed rotation system and method | |
| CN105659631B (en) | Sound field measurement apparatus and sound field measuring method | |
| CN102512138A (en) | Cardiac sound monitoring and early warning method | |
| CN107894280A (en) | Hand-held vibration noise measuring instrument and its vibration noise measuring method | |
| CN110335593A (en) | Sound end detecting method, device, equipment and storage medium | |
| CN111508531A (en) | Audio processing method and device | |
| CN201683910U (en) | Intelligent cardiopulmonary analyzing instrument | |
| WO2017045512A1 (en) | Voice recognition method and apparatus, terminal, and voice recognition device | |
| JP5077847B2 (en) | Reverberation time estimation apparatus and reverberation time estimation method | |
| CN112135235B (en) | Quality detection method, system and computer readable storage medium | |
| CN113723521A (en) | Motor abnormal sound identification method and identification system | |
| CN106060707A (en) | Reverberation processing method and device | |
| CN103077627A (en) | Portable type visualized cardiophony teaching system | |
| CN205336536U (en) | Measurement device for short -wave radio set speaker | |
| CN110992972B (en) | Sound source noise reduction method based on multi-microphone earphone, electronic equipment and computer readable storage medium | |
| CN105577909A (en) | Denoising method and device |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170714 |