(19)
(11)EP 3 869 821 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
26.10.2022 Bulletin 2022/43

(21)Application number: 21158473.5

(22)Date of filing:  22.02.2021
(51)International Patent Classification (IPC): 
H04R 1/10(2006.01)
(52)Cooperative Patent Classification (CPC):
H04R 2410/07; H04R 1/1083; H04R 2410/05; H04R 2460/01

(54)

SIGNAL PROCESSING METHOD AND DEVICE FOR EARPHONE, AND EARPHONE

SIGNALVERARBEITUNGSVERFAHREN UND -VORRICHTUNG FÜR KOPFHÖRER UND KOPFHÖRER

PROCÉDÉ ET DISPOSITIF DE TRAITEMENT DE SIGNAL POUR ÉCOUTEUR ET ÉCOUTEUR


(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

(30)Priority: 20.02.2020 CN 202010103438

(43)Date of publication of application:
25.08.2021 Bulletin 2021/34

(73)Proprietor: Beijing Xiaoniao Tingting Technology Co., Ltd
Beijing 100191 (CN)

(72)Inventors:
  • LIU, Song
    Beijing 100191 (CN)
  • LI, Na
    Beijing 100191 (CN)
  • LI, Bo
    Beijing 100191 (CN)

(74)Representative: Lavoix 
Bayerstraße 83
80335 München
80335 München (DE)


(56)References cited: : 
WO-A1-2019/141102
JP-A- 2007 002 393
US-A1- 2018 018 954
GB-A- 2 500 251
US-A1- 2017 374 477
  
      
    Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


    Description

    TECHNICAL FIELD



    [0001] The disclosure relates to the technical field of active noise cancellation of an earphone, and more particularly, to a signal processing method and device for an earphone, and an earphone.

    BACKGROUND



    [0002] Earphones are more and more widely used in our daily life because of a small size and easy to carry. In addition to using the earphone to listen to music, watch videos, etc., the earphone plays a role of isolating noise and provides a user with a quiet environment. The physical isolation of the earphone has limitations for mid frequency noise and low frequency noise, especially low frequency noise. Therefore, active noise cancellation is more and more widely used in the earphone.

    [0003] The principle of active noise cancellation is to generate a signal of which the amplitude is close to the amplitude of environmental noise and the phase is opposite to the phase of the environmental noise, to offset the influence of the environmental noise. However, most of the active noise cancellation technologies currently used in the earphone are relatively simple and consider fewer external influence factors. Even if there is a method for adaptively adjusting the noise cancellation level according to the amplitude of the environmental noise, the existing noise cancellation algorithm often fails to consider the presence of wind noise and the situation when there is wind noise and the user is in motion.

    [0004] For example, the user may be at a higher noise cancellation level during walking or running, besides, with the physical isolation of the earphone to the environmental noise, which causes being unable to hear the environmental noise, such as bicycles or electric vehicles, or sound or whistle of a car driving by, which causes safety problems. Therefore, it needs to adjust the noise cancellation frequency band according to whether a wearer of the earphone is in a motion state, so that the useful external noise may be perceived while ensuring noise cancellation.

    [0005] Furthermore, because the wind noise is a random signal without a fixed phase, when according to the existing noise cancellation algorithm, after a feedforward microphone simply acquires the external sound signal and superimposes it reversely, the wind noise may not be eliminated or may even be amplified, and when there is large wind noise, it will seriously affect the sense of hearing. Therefore, the wind noise also needs to be considered when performing noise cancellation processing.

    [0006] WO 2019/141102 A1 and EP 3 672 274 A4 from the same patent family disclose an adaptive audio control method and a system based on scenario identification. A motion mode of the user is determined. A wind suppression submodule is configured to filter the wind noise in the ambient sound signal. Once the big wind noises are detected, different filters may be set to deal with, so as to reduce the impact of the wind noise on the user listening to the audio signal.

    [0007] US 2017 / 0374477 A1 discloses a hearing aid device for augmenting environment sounds which comprises a sensor for recognizing motions of the hearing aid and a wind noise detector.

    SUMMARY



    [0008] The embodiments of the disclosure provide a signal processing method and device for an earphone, and an earphone, which may avoid the adverse influence of the wind noise on the sense of hearing, and reduce safety problems caused by noise cancellation in motion.

    [0009] The invention is set out in the appended set of claims.

    BRIEF DESCRIPTION OF THE DRAWINGS



    [0010] 

    FIG. 1 is a schematic flowchart of a signal processing method for an earphone according to an embodiment of the disclosure;

    FIG. 2 is a schematic flowchart of a method for detecting wind noise according to an embodiment of the disclosure;

    FIG. 3 is a schematic logical diagram of determining wind noise conditions according to an embodiment of the disclosure;

    FIG. 4 is a schematic flowchart of a signal processing method for an earphone according to another embodiment of the disclosure;

    FIG. 5 is a schematic logical diagram of controlling active noise cancellation according to an embodiment of the disclosure;

    FIG. 6 is a schematic structural diagram of a signal processing device for an earphone according to an embodiment of the disclosure;

    FIG. 7 is a schematic structural diagram of a signal processing device for an earphone according to another embodiment of the disclosure;

    FIG. 8 is a schematic flowchart of a method for detecting wind noise according to an embodiment of the disclosure; and

    FIG. 9 is a schematic structural diagram of a device for detecting wind noise according to an embodiment of the disclosure.


    DETAILED DESCRIPTION



    [0011] In order to make the objectives, technical solutions and advantages of the disclosure clearer, the embodiments of the disclosure will be further described in detail below with reference to the drawings. However, it should be understood that these descriptions are exemplary only, and are not intended to limit the scope of the disclosure. In addition, in the following descriptions, descriptions of well-known structures and technologies are omitted to avoid obscuring the concept of the disclosure unnecessarily.

    [0012] The terms used herein are intended to describe specific embodiments only, and are not intended to limit the disclosure. The words "a", "an" and "the" etc. used herein shall also include the meanings of "multiple" and "plural", unless indicated clearly by the context otherwise. Furthermore, the terms "include", "including", etc. used herein indicate the presence of the described features, steps, operations and/or components, but do not exclude the presence or addition of one or more other features, steps, operations or components.

    [0013] All the terms (including technical and scientific terms) used herein have meanings generally understood by those skilled in the art, unless defined otherwise. It should be noted that the terms used herein should be interpreted as having meanings consistent with the context of the description, and should not be interpreted in an idealized or overly rigid manner.

    [0014] Some block diagrams and/or flowcharts are shown in the drawings. It should be understood that some blocks in the block diagram and/or flowchart or combination thereof may be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that these instructions may create devices for implementing the functions/operations described in these block diagrams and/or flowcharts when being executed by the processor.

    [0015] Therefore, technologies of the disclosure may be implemented in the form of hardware and/or software (including firmware, microcode, etc.). Furthermore, the technologies of the disclosure may take the form of a computer program product on a computer-readable storage medium storing instructions, and the computer program product may be used by an instruction execution system or in combination with the instruction execution system. In the context of the disclosure, the computer-readable storage medium may be any medium that may contain, store, transfer, propagate or transmit instructions. For example, the computer-readable storage medium may include, but is not limited to, an electric, magnetic, optical, electromagnetic, infrared or semiconductor system, device, component or propagation medium. Specific examples of the computer-readable storage medium include: magnetic storage devices such as magnetic tape or hard disk (HDD); optical storage devices such as optical disk (CD-ROM); memory such as Random Access Memory (RAM) or flash memory; and/or wired/wireless communication link.

    [0016] An active noise cancellation earphone usually has a fixed noise cancellation mode, and the noise cancellation effect thereof cannot be selected intelligently according to actual conditions. When a user is in motion, especially when the user selects a higher noise cancellation level, it will cause a greater level of noise cancellation added based on the physical isolation. When the user forgets to decrease the noise cancellation level, it often makes the user completely unable to hear the external environmental sound, which may cause safety problems. Therefore, it needs to control a noise cancellation filter automatically according to a motion state, so that the user may perceive useful external noise during the motion, thereby avoiding safety problems. Furthermore, when there is wind noise, since the wind noise is a random signal and has no fixed phase characteristics, reverse superposition of feedforward noise cancellation will cause the wind noise to be amplified at some moments, which will affect the user's sense of hearing seriously. Therefore, it needs to control a feedforward filter automatically according to the intensity of the wind noise of different frequency bands, to avoid the adverse influence of the wind noise on the sense of hearing.

    [0017] FIG. 1 is a schematic flowchart of a signal processing method for an earphone according to an embodiment of the disclosure, as shown in FIG. 1, the signal processing method for an earphone according to the embodiment includes operations 110 to 130:
    In operation 110, a motion state of a wearer of the earphone is detected by using an acceleration sensor arranged inside the earphone.

    [0018] The motion state of the wearer of the earphone may be acquired by a triaxial acceleration sensor arranged inside the earphone. The currently integrated triaxial acceleration sensor usually has an output of the number of steps, or step counting statistics may be performed according to the triaxial acceleration signal, and then it is determined, according to the output of the number of steps, whether there is a motion. According to an embodiment, when there is step counting continuously for 3 seconds (s) and the total number of steps is greater than 8, then it may be determined that the wearer of the earphone is in a motion state at the moment; and when there is no step counting continuously for 3 seconds (s), then it may be determined that the wearer of the earphone is in a non-motion state at the moment.

    [0019] In operation 120, a first microphone and a second microphone both arranged outside the earphone detect wind noise conditions corresponding to different frequency bands.

    [0020] The wind noise condition is detected mainly by signal of microphones arranged outside the earphone. In the disclosure, two microphones are both arranged outside the earphone: a first microphone and a second microphone. The positions for arranging the two microphones are slightly different, herein the first microphone is located far away from the mouth of the wearer of the earphone, and is mainly used to collect the environmental noise around the wearer of the earphone, also known as a feedforward microphone; and the second microphone is located close to the mouth of the wearer of the earphone, and is mainly used to collect the voice signal of the wearer of the earphone, also known as a talk microphone.

    [0021] Furthermore, a feedback microphone is usually arranged inside the earphone too, which is located in a coupling cavity between the human ear and the earphone, and is configured to collect residual noise in the coupling cavity.

    [0022] Frequency bands of the signals are divided into multiple different frequency bands, and the presence or absence of the wind noise and the intensity of the wind noise are detected for each of the frequency band respectively, to detect the frequency band polluted by the wind noise, so that in the motion state, a suitable frequency band is selected to amplify the external environmental sound and avoid safety problems to the greatest extent.

    [0023] It should be noted that the above operations 110 and 120 are performed synchronously, and what is acquired is the motion state of the wearer of the earphone and the wind noise conditions corresponding to different frequency bands at the same time.

    [0024] In operation 130, according to the motion state of the wearer of the earphone and the wind noise conditions corresponding to different frequency bands, operating modes of a feedforward filter and a feedback filter inside the earphone are adjusted, herein the feedforward filter and the feedback filter are configured for active noise cancellation of the earphone.

    [0025] In this operation, at least the coefficient of the feedforward filter or the coefficient of the feedback filter is adjusted according to the motion state of the wearer of the earphone and the wind noise conditions corresponding to different frequency bands, so that they have appropriate noise cancellation effects at different scenarios and different frequency bands.

    [0026] The principle of active noise cancellation of the feedforward filter is to control a loudspeaker to generate noise with close energy and an opposite phase, according to the ambient environmental noise collected by the feedforward microphone (the first microphone), thereby achieving the noise cancellation effect. The principle of active noise cancellation of the feedback filter is to control the loudspeaker to generate a signal of which energy is close to and phase is opposite to the noise in the coupling cavity, according to the residual noise in the coupling cavity collected by the feedback microphone, thereby achieving the noise cancellation effect.

    [0027] It should be noted that when only feedforward noise cancellation is used, it may be considered that the feedback microphone is not arranged, and the algorithm of the embodiment is also applicable.

    [0028] Compared with the related art, the signal processing method and device for an earphone and the earphone provided by the embodiments consider the motion state of the wearer and wind noise conditions corresponding to different frequency bands together, when performing active noise cancellation control on the filter group inside the earphone, so that the user in motion may perceive the useful outside noise, by adjusting the operating modes of the feedforward filter and the feedback filter inside the earphone, thereby avoiding safety problems; and when there is wind noise, the noise cancellation processing of the feedforward filter is controlled to avoid the frequency band where the wind noise is located, thereby avoiding the adverse influence of the wind noise on the sense of hearing.

    [0029] The above operation 120 and 130 are described in detail below.

    [0030] In order to detect the wind noise condition effectively, the embodiment obtains a method for detecting wind noise conditions corresponding to different frequency bands effectively according to the correlation of signals of the feedforward microphone and the talk microphone at different frequency bands, and energy of the signal of the feedforward microphone at different frequency bands. Since the intensity of the wind noise is different, the frequency band that the wind noise may reach is also different, for example, in the case of no wind noise or weak wind noise in the low frequency band, it may be considered that there is no wind noise in the mid frequency band and the high frequency band either, while in the case of strong wind noise in the low frequency band, the mid frequency band may have weak wind noise or strong wind noise. Moreover, the intensity of the wind noise and the frequency band where it is located change in real time. In order to ensure that the user may hear the external environmental sound in the motion state, it needs to consider amplifying signals in the mid and high frequency bands (horns and alarms are usually in the mid and high frequency bands), but when the wind noise is present, the frequency band where the wind noise is present cannot be amplified to avoid affecting the sense of hearing. Therefore, in the motion, in order to avoid amplifying the wind noise and to amplify the external environmental sound in mid and high frequency bands as much as possible, the embodiment divides the determination of the wind noise into three frequency bands: low frequency band, mid frequency band and high frequency band. The presence or absence of the wind noise in the three frequency bands and the intensity of the wind noise are determined respectively, so that in a motion mode, a suitable frequency band may be selected to amplify the external environmental sound, avoid safety problems to the greatest extent while avoid the adverse effect of the wind noise.

    [0031] The above operation 120 includes: the average correlation of signals of the two microphones corresponding to each of three frequency bands including a low frequency band, a mid frequency band and a high frequency band respectively, and the average energy of the signals of the first microphone are acquired according to signals collected by the first microphone and the second microphone; and then it is determined, according to the acquired average correlations and average energies, whether there is wind noise in the three frequency bands, and intensities of the wind noise in the three frequency bands.

    [0032] The division of the above three frequency bands including a low frequency band, a mid frequency band and a high frequency band considers a frequency range where the wind noise is mainly located and an approximate frequency range where some environmental noise is located, such as the frequency range during operation of electric vehicles, cars, bicycles, etc., and the frequency of whistle, etc.

    [0033] In some embodiments, the "acquiring, according to signals collected by the first microphone and the second microphone, an average correlation of signals of the two microphones corresponding to each of three frequency bands including a low frequency band, a mid frequency band and a high frequency band respectively, and an average energy of the signals of the first microphone corresponding to each of the three frequency bands" includes:
    according to time domain signals collected by the first microphone and the second microphone, frequency domain signals of the first microphone and the second microphone at each of frequency points are acquired; and frequency bands of the signals are divided into three frequency bands including a low frequency band, a mid frequency band and a high frequency band, and according to the acquired frequency domain signals of the first microphone and the second microphone at each of frequency points, the average correlation of signals of the two microphones corresponding to each of the three frequency bands including the low frequency band, the mid frequency band and the high frequency band respectively, and the average energy of the signals of the first microphone corresponding to each of the three frequency bands are acquired.

    [0034] In some embodiments, the "determining, according to the acquired average correlations and average energies, whether there is wind noise in the three frequency bands, and the intensities of the wind noise in the three frequency bands" includes:

    for each frequency band, if the average correlation is less than a first correlation threshold, it is determined that there is strong wind noise in the frequency band responsive to the average energy being greater than a first energy threshold, it is determined that there is weak wind noise in the frequency band responsive to the average energy being less than the first energy threshold but greater than a second energy threshold, and it is determined that there is no wind noise in the frequency band responsive to the average energy being less than the second energy threshold;

    for each frequency band, if the average correlation is greater than the first correlation threshold but less than a second correlation threshold, it is determined that there is weak wind noise in the frequency band responsive to the average energy being greater than the second energy threshold, and it is determined that there is no wind noise in the frequency band responsive to the average energy being less than the second energy threshold; and

    for each frequency band, if the average correlation is greater than the second correlation threshold, it is determined that there is no wind noise in the frequency band.



    [0035] The above determination logic is applicable to the three frequency bands including the low frequency band, the mid frequency band and the high frequency band.

    [0036] It should be noted that when whether there is wind noise in the three frequency bands, and the intensities of the wind noise in the three frequency bands are determined, whether there is wind noise in the low frequency band is first determined, and when it is determined that there is wind noise in the low frequency band, whether there is wind noise in the mid frequency band is then determined, and when it is determined that there is wind noise in the mid frequency band, whether there is wind noise in the high frequency band is then determined. This is because: since the intensity of the wind noise is different, the frequency band that the wind noise may reach is also different. The stronger the wind noise, the higher the frequency that may be reached, and the wider the frequency band involved. When the wind noise is weak, it is only present at low frequencies. Therefore, when it is determined that there is no wind noise in the low frequency band, it is also considered that there is no wind noise in the mid frequency band and the high frequency band, thus it does not need to determine whether there is wind noise in the mid frequency band and the high frequency band; when it is determined that there is no wind noise in the mid frequency band, it is also considered that there is no wind noise in the high frequency band, thus it does not need to determine whether there is wind noise in the high frequency band. With the sequence of determining from the low frequency band, the mid frequency band to the high frequency band, when there is no wind noise in the low frequency band or the mid frequency band, the number of determination may be reduced without affecting the accuracy of the result of determination.

    [0037] FIG. 2 is a schematic flowchart of a method for detecting wind noise according to an embodiment of the disclosure. With reference to FIG. 2, the method for detecting wind noise given by the embodiment includes operations 201 to 207:
    In operation 201, time domain signals collected by two microphones respectively are acquired.

    [0038] For example, the time domain signal collected by the first microphone (the feedforward microphone) is denoted as x=[x(0),x(1),......,x(N-1)], and the time domain signal collected by the second microphone (the talk microphone) is denoted as y=[y(0),y(1),......,y(N-1)].

    [0039] Furthermore, in order to filter out the influence of DC signal, it usually needs to perform high-pass filtering on the time domain signals of the two microphones.

    [0040] In operation 202, frequency domain signals of the two microphones at each of frequency points are acquired.

    [0041] This operation is to add analysis windows and perform Fourier transform to the time domain signals of the two microphones after high-pass filtering respectively, to obtain the frequency domain signals.

    [0042] The added analysis window such as a Hamming window (w=[w(0), w(1),...,w(N-1)]). The frequency domain signals of the two microphones at each of frequency points are acquired, denoted as X(k), Y(k) respectively, as shown in the following formula:

    ; and



    [0043] Herein, N represents the number of Fourier transform points, n represents the sampling point of signal sequence, and k represents the frequency point.

    [0044] In operation 203, normalized correlation calculation on the frequency domain signals of the two microphones is performed.

    [0045] This operation is to perform amplitude normalization processing on the frequency domain signals of the two microphones respectively, as shown in the following formula, to obtain normalized signals Xnorm(k) and Ynorm(k):



    [0046] Then the normalized signals are used by the following formula to obtain the correlation signal:



    [0047] In operation 204, frequency bands are divided.

    [0048] This operation is to divide frequency bands of the signals into three frequency bands corresponding to low frequency, mid frequency and high frequency. For example, according to actual conditions and empirical data, in order to avoid mutual interference between different frequency bands, the frequency bands of the signals are divided into the following three discontinuous frequency bands: 200 Hz-600 Hz, 700 Hz-1200 Hz, and 1300 Hz-3000 Hz. Of course, the frequency bands of the signals may also be divided into three continuous frequency bands according to actual needs.

    [0049] The conversion formula between frequency point in frequency domain and frequency in time domain is as follows:



    [0050] Herein, bin represents frequency point in the frequency domain (or referred to as frequency window), f represents actual frequency in the time domain, such as 200 Hz, 600 Hz, 700 Hz, 1200 Hz, 1300 Hz and 3000 Hz, etc., fs=8000 Hz represents the sampling rate, the length of time for processing signals is 0.016s, N=80000.016=256, which represents the number of Fast Fourier Transformation (FFT) points.

    [0051] In operation 205, the average correlation of each of different frequency bands is calculated.

    [0052] The following method may be used to calculate the average correlation of signals of the two microphones at different frequency bands:



    [0053] Herein, Cohband_i represents the average correlation of the i-th frequency band, bandi represents the i-th frequency band (for example, the first frequency band is the low frequency band, the second frequency band is the mid frequency band, and the third frequency band is the high frequency band), i_binstart represents the serial number of the frequency point where the i-th frequency band starts, i_binend represents the serial number of the frequency point where the i-th frequency band ends, and Coh (k) represents the correlation value at the k-th frequency point.

    [0054] In operation 206, the average energy of each of different frequency bands is calculated.

    [0055] The following method may be used to calculate the average energy of the signals of the first microphone in each of different frequency bands:



    [0056] Herein, X(k) represents the amplitude of the frequency domain signal of the first microphone at each of frequency points, and Xpowband_i represents the average energy of the i-th frequency band.

    [0057] It should be noted that the above operations 205 and 206 are in a parallel relationship, the two operations may be performed in parallel or sequentially, for example, the frequency domain signal of the first microphone may be divided into different frequency bands, and the average energy of the first microphone at different frequency bands is first calculated, and then the relevant signals of signals of the two microphones are also divided into the same frequency band, and then the average correlation level of signals of the two microphones at a corresponding frequency band is calculated.

    [0058] In operation 207, according to the acquired average correlations and average energies, the wind noise condition of the corresponding frequency band is determined.

    [0059] This operation is to determine, according to the acquired average correlation of signals of the two microphones corresponding to each of the three frequency bands including the low frequency band, the mid frequency band and the high frequency band respectively and the average energy of the signals of the first microphone corresponding to each of the three frequency bands, whether there is wind noise at the low frequency, mid frequency and high frequency and the intensity of the wind noise respectively.

    [0060] According to the randomness of the wind noise, the correlation between the wind noise collected by two microphones is very low, and when the wind noise is large, the energy of the signal of the first microphone is usually large, thus whether there is wind noise at different frequency bands and the intensity of the wind noise may be obtained according to the two characteristics. By considering the energy value of the signal of the microphone when determining the wind noise, incorrect determination that may occur when considering the correlation value of signals of the first and second microphones only may be avoided.

    [0061] FIG. 3 is a schematic logical diagram of determining wind noise conditions according to an embodiment of the disclosure, the determination logic is applicable to any frequency band of low frequency, mid frequency and high frequency. As shown in FIG. 3, the logic for determining the wind noise condition at a certain frequency band given by the embodiment includes operations 301 to 350:

    [0062] In operation 301, it is determined whether the average correlation is less than a first correlation threshold; if yes, that is, when the average correlation is less than the first correlation threshold, it goes to operation 302, and if not, that is, when the average correlation is greater than the first correlation threshold, it goes to operation 312.

    [0063] In operation 302, it is determined whether the average energy is greater than a first energy threshold; if yes, that is, when the average energy is greater than the first energy threshold, it goes to operation 330 in which it is determined that there is strong wind noise. If not, it goes to operation 303.

    [0064] In operation 303, it is determined whether the average energy is greater than a second energy threshold; if yes, that is, when the average energy is less than the first energy threshold but greater than the second energy threshold, it goes to operation 340 in which it is determined that there is weak wind noise. If not, that is, when the average energy is less than the second energy threshold, it goes to operation 350 in which it is determined that there is no wind noise.

    [0065] In operation 312, it is determined whether the average correlation is less than a second correlation threshold; if yes, that is, when the average correlation is greater than the first correlation threshold but less than the second correlation threshold, it goes to operation 322, and if not, that is, when the average correlation is greater than the second correlation threshold, it goes to operation 350 in which it is determined that there is no wind noise.

    [0066] In operation 322, it is determined whether the average energy is greater than the second energy threshold; if yes, that is, when the average energy is greater than the second energy threshold, it goes to operation 340 in which it is determined that there is weak wind noise. If not, that is, when the average energy is less than the second energy threshold, it goes to operation 350 in which it is determined that there is no wind noise.

    [0067] The correlation thresholds and energy thresholds of different frequency bands may be obtained according to actual test and statistics. For example, it is obtained according to actual test and statistics that corresponding to the low frequency band, the first correlation threshold is 0.35, the second correlation threshold is 0.5, the first energy threshold is 0.0032, and the second energy threshold is 0.0015.

    [0068] Due to the randomness of the wind noise, when the wind noise is small, or in the case of a quieter environment without wind noise, when the energy of the environmental noise is small, the correlation between two microphones outside the earphone is also relatively low, misjudgment easily occurs by using the correlation only at the moment. In the method for detecting wind noise given by the embodiment, in addition to considering correlation of signals, the energy of the first microphone (the feedforward microphone) is also considered, and when determining the wind noise condition, correlation of signals and energy of signals are considered at the same time to effectively reduce the misjudgment of the wind noise and ensure that the presence of the wind noise is detected only when there is a large wind noise; and the detection result of the wind noise may not only distinguish between weak wind noise and no wind noise, but also avoid determining weak wind noise as strong wind noise.

    [0069] In some embodiments not being part of the invention, the above operation 130 includes:

    when the wearer of the earphone is in the motion state and there is weak wind noise in the low frequency band, and no wind noise in the mid frequency band and the high frequency band, coefficient of the feedforward filter is adjusted to allow the feedforward filter to amplify signals in mid and high frequency bands, and the feedback filter is kept unchanged;

    when the wearer of the earphone is in the motion state and there is strong wind noise in the low frequency band, there is weak wind noise in the mid frequency band and no wind noise in the high frequency band, the coefficients of the feedforward filter are adjusted to allow the feedforward filter to amplify the signal in the high frequency band only, and the feedback filter is kept unchanged;

    when the wearer of the earphone is in the motion state and there is strong wind noise in the mid frequency band, and weak wind noise or strong wind noise in the high frequency band, the feedforward filter is turned off, and the feedback filter is kept unchanged; and

    when the wearer of the earphone is in the non-motion state and there is strong wind noise in the low frequency band, the feedforward filter is turned off, and the feedback filter is kept unchanged.



    [0070] In particular, because the physical isolation mainly isolates mid frequency noise and high frequency noise, especially high frequency noise, and the running noise and whistle noise of cars, electric vehicles or bicycles are also mainly located in mid and high frequency bands, for safety reasons, the user needs to hear such noise during motion, thus the feedforward filter needs to amplify the noise in the frequency band so that the user may hear it. In response to the above requirements, the reasons for making the above adjustments to the operating modes of the feedforward filter and the feedback filter in operation 130 are as follows:

    [0071] When the wearer of the earphone is in the motion state and there is weak wind noise in the low frequency band, and no wind noise in the mid frequency band and the high frequency band, then the coefficients of the feedforward filter are adjusted at the moment, the frequency response curve of the feedforward filter is raised in mid and high frequency bands, so that the noise at the mid and high frequency bands is amplified to let the user hear the external environmental sound, ensuring safety; and the feedback filter is kept unchanged at the moment, that is, the noise cancellation effect of the feedback filter on the low frequency noise is kept.

    [0072] When the wearer of the earphone is in the motion state and there is strong wind noise in the low frequency band, weak wind noise in the mid frequency band and no wind noise in the high frequency band, then the coefficients of the feedforward filter are adjusted at the moment, the frequency response curve of the feedforward filter is raised at high frequency, so that the noise in the high frequency band is amplified, and it has no effect on the noise at the low and mid frequency bands, which may avoid the adverse influence of the wind noise on the sense of hearing, and may also ensure safety to some extent at the same time. The feedback filter is also kept unchanged at the moment, that is, the noise cancellation effect of the feedback filter on the low frequency noise is kept.

    [0073] When the wearer of the earphone is in the motion state and there is strong wind noise in the mid frequency band, and weak wind noise or strong wind noise in the high frequency band, then the feedforward filter is turned off directly, and the feedback filter is kept unchanged, that is, only the noise cancellation effect of the feedback filter on the low frequency noise is kept.

    [0074] When the wearer of the earphone is in the non-motion state, but there is strong wind noise at the current frequency band, that is, he/she is in a wind noise environment, since there are no safety problems at the moment, the feedforward filter may be turned off, and only the noise cancellation effect of the feedback filter on the low frequency noise is kept.

    [0075] FIG. 4 a schematic flowchart of a signal processing method for an earphone according to another embodiment of the disclosure, as shown in FIG. 4, the signal processing method for an earphone according to the embodiment includes operations 410 to 440:

    [0076] In operation 410, a motion state of a wearer of the earphone is detected by using a triaxial acceleration sensor arranged inside the earphone.

    [0077] In operation 420, a first microphone and a second microphone both arranged outside the earphone detect wind noise conditions corresponding to different frequency bands.

    [0078] In operation 430, an energy magnitude of environmental noise is detected by using the first microphone.

    [0079] In operation 440, according to the motion state of the wearer of the earphone, the wind noise conditions corresponding to different frequency bands and the energy magnitude of the environmental noise, operating modes of a feedforward filter and a feedback filter inside the earphone are adjusted, herein the feedforward filter and the feedback filter are configured for active noise cancellation of the earphone.

    [0080] Compared with the signal processing method for an earphone as shown in FIG. 1, the signal processing method for an earphone as shown in FIG. 4 adds operation 430, operation 430 and operations 410, 420 are performed synchronously, that is, what is acquired is the motion state of the wearer of the earphone, the wind noise conditions corresponding to different frequency bands and the energy magnitude of the environmental noise at the same time; FIG. 4 replaces operation 130 shown in FIG. 1 with operation 440, the energy magnitude of the environmental noise is also considered when adjusting the operating modes of the feedforward filter and the feedback filter inside the earphone; operations 410 and 420 shown in FIG. 4 are the same as operations 110 and 120 shown in FIG. 1 respectively.

    [0081] In some embodiment, the above operation 430 includes: exponential smoothing is performed on the frequency domain signal of the first microphone at any of frequency points, and a minimum value of the smoothed signals is taken within a set length of time to obtain the environmental noise signal of the frequency point in the current time frame; and the environmental noise signals of all the frequency points in the current time frame are superimposed to obtain a total energy of the environmental noises.

    [0082] In particular, traditional methods for estimating noise such as the method for estimating noise with minimum statistics may be used to calculate the energy magnitude of the environmental noise, perform exponential smoothing on the frequency domain signal X(k) of the first microphone at each frequency point k, then take a minimum value of the smoothed signals within a set length of time to obtain the environmental noise signal Noise(k) shown in the following formula, herein λ is the current time frame, W is the time length, k is the frequency point, and P(λ, k) is the smoothed signal spectrum:





    [0083] Then for the estimated environmental noise signal Noise(k), the following formula is used to calculate the total energy NoisePow.



    [0084] The method for estimating the energy magnitude of the environmental noise given by the embodiment may avoid the interference of transient signals. In addition, when there is wind noise, the noise estimation is not updated, which may avoid the influence of the wind noise on the calculation of energy of the environmental noise.

    [0085] When adjusting the operating modes of the feedforward filter and the feedback filter inside the earphone, the above operation 440 needs to consider the energy magnitude of the environmental noise to include the following two situations:

    [0086] First situation: when the wearer of the earphone is in the motion state and there is no wind noise in the low frequency band, then the coefficients of the feedforward filter are adjusted to allow the feedforward filter to amplify signals in mid and high frequency bands, and a feedforward filter and a feedback filter with appropriate noise cancellation levels are selected according to the energy magnitude of the environmental noise in the low frequency band, so that there is an appropriate level of noise cancellation in the low frequency band;

    [0087] Second situation: when the wearer of the earphone is in the non-motion state, and there is no wind noise or weak wind noise in the low frequency band, a feedforward filter and a feedback filter with appropriate noise cancellation levels are selected according to the energy magnitude of the environmental noise in the entire frequency bands, to perform noise cancellation on the entire frequency bands.

    [0088] In particular, when the wearer of the earphone is in the motion state, for noise control in the low frequency band, the environmental noise may be pre-divided into different levels according to the energy magnitude of the environmental noise in the low frequency band, and the corresponding preset coefficients of the feedforward filter and the feedback filter may be selected according to different noise levels, so that there is an appropriate level of noise cancellation in the low frequency band. When the wearer of the earphone is in the non-motion state, for noise control in the entire frequency bands, the environmental noise may be pre-divided into different levels according to the energy magnitude of the environmental noise in the entire frequency bands, and the corresponding preset coefficients of the feedforward filter and the feedback filter may be selected according to different noise levels, to perform noise cancellation on the entire frequency bands. Further description will be made hereinafter.

    [0089] The feedforward filter processes the signal of the feedforward microphone, and the feedback filter processes the signal of the feedback microphone, both of them may be implemented by software on a chip. The applicant found that the feedforward filter has a good noise cancellation effect mainly on about 500Hz-1000Hz, but has a limited noise cancellation effect on the low frequency band below 500Hz, but the feedback filter has a noise cancellation effect mainly on the low frequency band below 500Hz. Therefore, considering that the feedforward filter and the feedback filter have different noise cancellation effects on different frequency bands, the operating modes of the feedforward filter and the feedback filter may be adjusted according to the energy magnitude of the environmental noise to ensure the level of noise cancellation at the corresponding frequency band. For example, when the wearer of the earphone is in the motion state and there is no wind noise in the low frequency band at the moment, safety is considered, and the energy of the environmental noise is usually strong at low frequency, thus it needs to amplify the environmental noises in the mid and high frequency bands and perform a proper noise cancellation on the low frequency. For the selection of the noise cancellation filter, the noise cancellation level of the noise cancellation filter may be divided into different levels according to the low frequency energy magnitude of the environmental noise (each level has its corresponding preset coefficients of the feedforward filter and the feedback filter), when the low frequency energy of the environmental noise is high, then a noise cancellation filter with a high noise cancellation level is selected, otherwise, a noise cancellation filter with a low noise cancellation level is selected. When there is wind noise in the low frequency band, the filter is not adjusted according to the low frequency energy of the environmental noise any more. When the user is in the non-motion state, and there is no wind noise or there is little wind noise in the low frequency band at the moment, the noise cancellation may be divided into different levels according to the energy of the environmental noise in the entire frequency bands, and each level has its corresponding preset coefficients of the feedforward filter and the feedback filter. When the energy of the environmental noise is high, then a high level noise cancellation filter is selected, otherwise, a low level noise cancellation filter is selected. When there is strong wind noise in the low frequency band, the filter is not adjusted according to the energy of the environmental noise.

    [0090] By considering the four control situations included in operation 130 of FIG. 1, operation 440 of FIG. 4 "adjusting, according to the motion state of the wearer of the earphone, the wind noise conditions corresponding to different frequency bands and the energy magnitude of the environmental noise, operating modes of the feedforward filter and the feedback filter inside the earphone" includes at least six control situations.

    [0091] FIG. 5 is a schematic logical diagram of controlling active noise cancellation according to an embodiment of the disclosure. As shown in FIG. 5, the logic of controlling active noise cancellation given by the embodiment includes operations 501 to 580:
    In operation 501, a motion state of a wearer of an earphone is acquired.

    [0092] In operation 502, wind noise conditions corresponding to low frequency band, mid frequency band and high frequency band are acquired.

    [0093] In operation 503, energy magnitude of environmental noise is acquired.

    [0094] In operation 511, it is determined whether the wearer of the earphone is in the motion state; if yes, that is, the wearer of the earphone is in the motion state, it goes to operation 521, and if not, that is, the wearer of the earphone is in a non-motion state, it goes to operation 531.

    [0095] In operation 521, it is determined whether there is wind noise in the low frequency band, and the intensity of the wind noise; when it is determined that there is no wind noise in the low frequency band, it goes to operation 540, and when it is determined that there is weak wind noise in the low frequency band, it goes to operation 550, and when it is determined that there is strong wind noise in the low frequency band, it goes to operation 522.

    [0096] In operation 522, the intensity of the wind noise in the mid frequency band is determined, when it is determined that there is weak wind noise in the mid frequency band, it goes to operation 560, and when it is determined that there is strong wind noise in the mid frequency band, it goes to operation 570.

    [0097] In operation 531, it is determined whether there is wind noise in the low frequency band, and the intensity of the wind noise; when it is determined that there is no wind noise or weak wind noise in the low frequency band, it goes to operation 580, and when it is determined that there is strong wind noise in the low frequency band, it goes to operation 570.

    [0098] In operation 540, the coefficients of the feedforward filter are adjusted to allow the feedforward filter to amplify signals in mid and high frequency bands, and according to the energy magnitude of the environmental noise in the low frequency band, a feedforward filter and a feedback filter with appropriate noise cancellation levels are selected, so that there is enough level of noise cancellation in the low frequency band. For example, when the environmental noise is large in the low frequency band, a higher noise cancellation level may be selected.

    [0099] Note: in the case of no wind noise in the low frequency band, it may be considered that there is no wind noise in the mid frequency band and the high frequency band, either.

    [0100] In operation 550, the coefficients of the feedforward filter are adjusted to allow the feedforward filter to amplify signals in mid and high frequency bands, and the feedback filter is kept unchanged.

    [0101] Note: in the case of weak wind noise in the low frequency band, it may be considered that there is no wind noise in the mid frequency band and the high frequency band.

    [0102] In operation 560, the coefficients of the feedforward filter are adjusted to allow the feedforward filter to amplify the signal in the high frequency band only, and the feedback filter is kept unchanged.

    [0103] Note: in the case of strong wind noise in the low frequency band, there may be weak wind noise or strong wind noise in the mid frequency band. In the case of weak wind noise in the mid frequency band, it may be considered that there is no wind noise in the high frequency band.

    [0104] In operation 570, the feedforward filter is turned off and the feedback filter is kept unchanged.

    [0105] Note: in the case of strong wind noise in the mid frequency band, there may be weak wind noise or strong wind noise in the high frequency band.

    [0106] In operation 580, according to the energy magnitude of the environmental noise in the entire frequency bands, a feedforward filter and a feedback filter with appropriate noise cancellation levels are selected to perform noise cancellation on the entire frequency bands, so that there is enough level of noise cancellation in the entire frequency bands. For example, the noise cancellation level is divided into different levels according to the energy magnitude of the environmental noise, and each level has its corresponding preset coefficients of the feedforward filter and the feedback filter. When the energy of the environmental noise is high, then a noise cancellation filter with high noise cancellation level is selected, otherwise, a noise cancellation filter with low noise cancellation level is selected.

    [0107] It should be noted that there are two control situations for the control method corresponding to the above operation 570 "turning off the feedforward filter and keeping the feedback filter unchanged": one is in the non-motion state, as long as there is strong wind noise in the low frequency band, the feedforward filter is turned off directly, and this operation is performed no matter whether there is wind noise in the mid frequency band or the high frequency band, thus it does not need to consider the wind noise conditions of the mid and high frequency bands. The other is in the motion state, for safety reasons, it needs to amplify the environmental noise, but the frequency bands affected by the level of the wind noise are different, it needs to determine, according to the frequency bands to which the wind noise relates, the frequency bands that may be amplified, thus it needs to consider the wind noise conditions of the mid and high frequency bands. However, as long as there is strong wind noise in the mid frequency band, the feedforward filter is also turned off directly to avoid the adverse influence of the wind noise on the sense of hearing.

    [0108] So far, the implementation of the signal processing method for an earphone according to the embodiment has been described in detail. The signal processing method for an earphone according to the embodiment adjusts the operating modes of the feedforward filter and the feedback filter inside the earphone by considering the motion state of the wearer of the earphone, the wind noise conditions corresponding to different frequency bands and the energy magnitude of the environmental noise together, to achieve controlling the noise cancellation effect automatically, while obtaining a better effect on the sense of hearing at different scenarios, and reducing safety problems caused by noise cancellation during motion.

    [0109] FIG. 6 is a schematic structural diagram of a signal processing device for an earphone according to an embodiment of the disclosure, as shown in FIG. 6, the signal processing device for an earphone according to the embodiment includes:

    a motion state detection module 610 is configured to detect a motion state of a wearer of the earphone by using an acceleration sensor arranged inside the earphone; for example, an output of the number of steps is acquired by a triaxial acceleration sensor integrated inside the earphone, or step counting statistics may be performed according to the triaxial acceleration signal, and then it is determined, according to the output of the number of steps, whether there is a motion.

    a wind noise detection module 620 is configured to detect wind noise conditions corresponding to different frequency bands by using a first microphone and a second microphone both arranged outside the earphone; and

    a noise cancellation control module 640 is configured to adjust, according to the motion state of the wearer of the earphone and the wind noise conditions corresponding to different frequency bands, operating modes of a feedforward filter and a feedback filter inside the earphone, herein the feedforward filter and the feedback filter are configured for active noise cancellation of the earphone.



    [0110] Herein the motion state detection module 610 and the wind noise detection module 620 are connected in parallel to the noise cancellation control module 640.

    [0111] In some embodiments, also as shown in FIG. 6, the wind noise detection module 620 includes:

    a band-wise processing unit 621 is configured to acquire, according to signals collected by the first microphone and the second microphone, an average correlation of signals of the two microphones corresponding to each of three frequency bands including a low frequency band, a mid frequency band and a high frequency band respectively, and an average energy of the signals of the first microphone corresponding to each of the three frequency bands; and

    a wind noise determination unit 622 is configured to determine, according to the acquired average correlations and average energies, whether there is wind noise in the three frequency bands, and the intensities of the wind noise in the three frequency bands.



    [0112] In some embodiments, the band-wise processing unit 621 is specifically configured to:

    acquire, according to time domain signals collected by the first microphone and the second microphone, frequency domain signals of the first microphone and the second microphone at each of frequency points; and

    divide frequency bands of the signals into three frequency bands including a low frequency band, a mid frequency band and a high frequency band, and acquire, according to the acquired frequency domain signals of the first microphone and the second microphone at each of frequency points, the average correlation of signals of the two microphones corresponding to each of the three frequency bands including the low frequency band, the mid frequency band and the high frequency band respectively, and the average energy of the signals of the first microphone corresponding to each of the three frequency bands.



    [0113] In some embodiments, the wind noise determination unit 622 is specifically configured to:

    for each frequency band, if the average correlation is less than a first correlation threshold, determine that there is strong wind noise in the frequency band responsive to the average energy being greater than a first energy threshold, determine that there is weak wind noise in the frequency band responsive to the average energy being less than the first energy threshold but greater than a second energy threshold, and determine that there is no wind noise in the frequency band responsive to the average energy being less than the second energy threshold;

    for each frequency band, if the average correlation is greater than the first correlation threshold but less than a second correlation threshold, determine that there is weak wind noise in the frequency band responsive to the average energy being greater than the second energy threshold, and determine that there is no wind noise in the frequency band responsive to the average energy being less than the second energy threshold; and

    for each frequency band, if the average correlation is greater than the second correlation threshold, determine that there is no wind noise in the frequency band.



    [0114] In some embodiments, the noise cancellation control module 640 is specifically configured to:

    when the wearer of the earphone is in the motion state and there is weak wind noise in the low frequency band, and no wind noise in the mid frequency band and the high frequency band, adjust coefficients of the feedforward filter to allow the feedforward filter to amplify signals in mid and high frequency bands, and keep the feedback filter unchanged;

    when the wearer of the earphone is in the motion state and there is strong wind noise in the low frequency band, weak wind noise in the mid frequency band and no wind noise in the high frequency band, adjust the coefficients of the feedforward filter to allow the feedforward filter to amplify the signal in the high frequency band only, and keep the feedback filter unchanged;

    when the wearer of the earphone is in the motion state and there is strong wind noise in the mid frequency band, and weak wind noise or strong wind noise in the high frequency band, turn off the feedforward filter, and keep the feedback filter unchanged;

    when the wearer of the earphone is in a non-motion state and there is strong wind noise in the low frequency band, turn off the feedforward filter, and keep the feedback filter unchanged.



    [0115] FIG. 7 is a schematic structural diagram of a signal processing device for an earphone according to another embodiment of the disclosure, as shown in FIG. 7, the signal processing device for an earphone according to the embodiment includes:

    a motion state detection module 710 is configured to detect, by a triaxial acceleration sensor arranged inside the earphone, a motion state of a wearer of the earphone;

    a wind noise detection module 720 is configured to detect wind noise conditions corresponding to different frequency bands by using a first microphone and a second microphone both arranged outside the earphone;

    an environmental noise detection module 730 is configured to detect an energy magnitude of environmental noise by using the first microphone; and

    a noise cancellation control module 740 is configured to adjust, according to the motion state of the wearer of the earphone, the wind noise conditions corresponding to different frequency bands and the energy magnitude of the environmental noise, operating modes of a feedforward filter and a feedback filter inside the earphone, herein the feedforward filter and the feedback filter are configured for active noise cancellation of the earphone.



    [0116] Compared with the signal processing device for an earphone shown in FIG. 6, the signal processing device for an earphone shown in FIG. 7 further includes the environmental noise detection module 730, the environmental noise detection module 730 and the motion state detection module 710, the wind noise detection module 720 are connected in parallel to the noise cancellation control module 740 together; compared with the noise cancellation control module 640 shown in FIG. 6, the noise cancellation control module 740 shown in FIG. 7 also considers the energy magnitude of the environmental noise when adjusting the operating modes of the feedforward filter and the feedback filter inside the earphone; the motion state detection module 710 and the wind noise detection module 720 shown in FIG. 7 are the same as the motion state detection module 610 and the wind noise detection module 620 shown in FIG. 6 respectively.

    [0117] In some embodiments, the environmental noise detection module 730 is specifically configured to:

    perform exponential smoothing on the frequency domain signal of the first microphone at any of frequency points, and take a minimum value of the smoothed signals within a set length of time to obtain the environmental noise signal of the frequency point in the current time frame; and

    superimpose the environmental noise signals of all the frequency points in the current time frame to obtain a total energy of the environmental noises.



    [0118] In some embodiments, compared with the noise cancellation control module 640 shown in FIG. 6, the noise cancellation control module 740 is also specifically configured to:

    when the wearer of the earphone is in the motion state and there is no wind noise in the low frequency band, adjust the coefficients of the feedforward filter to allow the feedforward filter to amplify signals in mid and high frequency bands, and select, according to the energy magnitude of the environmental noise in the low frequency band, a feedforward filter and a feedback filter with appropriate noise cancellation levels, so that there is an appropriate level of noise cancellation in the low frequency band; and

    when the wearer of the earphone is in a non-motion state, and there is no wind noise or weak wind noise in the low frequency band, select a feedforward filter and a feedback filter with appropriate noise cancellation levels according to the energy magnitude of the environmental noise in the entire frequency bands, to perform noise cancellation on the entire frequency bands.



    [0119] For the device embodiment, since it substantially corresponds to the method embodiment, the relevant descriptions thereof may refer to descriptions of the part of the method embodiment. The device embodiment as described above is merely illustrative, herein modules or units described as separate components may be or may not be physically separated, that is, they may be located at one place, or they may be distributed to multiple modules or units. Some or all of the modules may be selected according to actual needs to achieve the objective of the solution of the embodiment. Those ordinarily skilled in the art may understand and implement them without paying any creative work

    [0120] Finally, the disclosure also provides an embodiment of earphone. The earphone provided by the embodiment of the disclosure includes an acceleration sensor arranged inside the earphone, a first microphone and a second microphone both arranged outside the earphone, and a feedforward filter and a feedback filter inside the earphone; the earphone is further provided therein with the above signal processing device for an earphone.

    [0121] The earphone may be a wireless earphone or a wired earphone. When it is a wireless earphone, the earphone on each side needs to have the acceleration sensor arranged inside the earphone and two microphones both arranged outside the earphone; when it is a wired earphone, the earphones on two sides just need to be provided with a set of acceleration sensors arranged inside the earphone and two microphones both arranged outside the earphone, they may be located on the same side of the wired earphone, or they may be distributed on different sides of the wired earphone. The acceleration sensor inside the earphone may be an integrated triaxial acceleration sensor.

    [0122] Furthermore, the disclosure also provides an embodiment of a method for detecting wind noise, which detects wind noise conditions corresponding to different frequency bands according to the signal related characteristics of the talk microphone and the feedforward microphone and energy of the signal of the feedforward microphone, reduces the misjudgment of the wind noise effectively and increases the accuracy of detecting the wind noise.

    [0123] FIG. 8 is a schematic flowchart of a method for detecting wind noise according to an embodiment of the disclosure, as shown in FIG. 8, the method for detecting wind noise according to the embodiment includes operations 810 to 830:

    [0124] In operation 810, according to time domain signals collected by a first microphone and a second microphone arranged outside an earphone, frequency domain signals of the first microphone and the second microphone at each of frequency points are acquired;

    [0125] In operation 820, frequency bands of the signals are divided into three frequency bands including a low frequency band, a mid frequency band and a high frequency band, and according to the acquired frequency domain signals of the first microphone and the second microphone at each of frequency points, an average correlation of signals of the two microphones corresponding to each of the three frequency bands including the low frequency band, the mid frequency band and the high frequency band respectively, and average energy of the signals of the first microphone corresponding to each of the three frequency bands are acquired; and

    [0126] In operation 830, it is determined, according to the acquired average correlations and average energies, whether there is wind noise in the three frequency bands, and the intensities of the wind noise in the three frequency bands.

    [0127] Herein the implementation of operation 810 may refer to the above descriptions of operations 201 to 203 in FIG. 2, and the implementation of operation 820 may refer to the above descriptions of operations 204 to 206 in FIG. 2, which will not be repeated here.

    [0128] According to an embodiment, the first microphone is a feedforward microphone and the second microphone is a talk microphone.

    [0129] Herein the "determining, according to the acquired average correlations and average energies, whether there is wind noise in the three frequency bands, and the intensities of the wind noise in the three frequency bands" of operation 830 includes:

    for each frequency band, if the average correlation is less than a first correlation threshold, it is determined that there is strong wind noise in the frequency band responsive to the average energy being greater than a first energy threshold, it is determined that there is weak wind noise in the frequency band responsive to the average energy being less than the first energy threshold but greater than a second energy threshold, and it is determined that there is no wind noise in the frequency band responsive to the average energy being less than the second energy threshold;

    for each frequency band, if the average correlation is greater than the first correlation threshold but less than a second correlation threshold, it is determined that there is weak wind noise in the frequency band responsive to the average energy being greater than the second energy threshold, and it is determined that there is no wind noise in the frequency band responsive to the average energy being less than the second energy threshold; and

    for each frequency band, if the average correlation is greater than the second correlation threshold, it is determined that there is no wind noise in the frequency band.



    [0130] The above determination logic of operation 830 may refer to the schematic logical diagram of determining wind noise conditions shown in FIG. 3, which will not be repeated here either.

    [0131] FIG. 9 is a schematic structural diagram of a device for detecting wind noise according to an embodiment of the disclosure, as shown in FIG. 9, the device for detecting wind noise according to the embodiment includes:

    a signal acquiring unit 910 is configured to acquire, according to time domain signals collected by a first microphone and a second microphone arranged outside an earphone, frequency domain signals of the first microphone and the second microphone at each of frequency points;

    a frequency band dividing unit 920 is configured to divide frequency bands of the signals into three frequency bands including a low frequency band, a mid frequency band and a high frequency band;

    an average correlation acquiring unit 930 is configured to acquire, according to the acquired frequency domain signals of the first microphone and the second microphone at each of frequency points, an average correlation of signals of the two microphones corresponding to each of the three frequency bands including the low frequency band, the mid frequency band and the high frequency band respectively;

    an average energy acquiring unit 940 is configured to acquire, according to the acquired frequency domain signals of the first microphone and the second microphone at each of frequency points, an average energy of the signals of the first microphone corresponding to each of the three frequency bands including the low frequency band, the mid frequency band and the high frequency band respectively; and

    a wind noise determination unit 950 is configured to determine, according to the acquired average correlations and average energies, whether there is wind noise in the three frequency bands, and the intensities of the wind noise in the three frequency bands.



    [0132] Herein the average correlation acquiring unit 930 and the average energy acquiring unit 940 are in a parallel relationship, and are connected in parallel to the wind noise determination unit 950, and the wind noise determination unit 950 needs to perform a logical determination of the wind noise condition according to the output results of the two units simultaneously.

    [0133] According to an embodiment, the first microphone is a feedforward microphone and the second microphone is a talk microphone.

    [0134] Herein the wind noise determination unit 950 is specifically configured to:

    for each frequency band, if the average correlation is less than a first correlation threshold, determine that there is strong wind noise in the frequency band responsive to the average energy being greater than a first energy threshold, determine that there is weak wind noise in the frequency band responsive to the average energy being less than the first energy threshold but greater than a second energy threshold, and determine that there is no wind noise in the frequency band responsive to the average energy being less than the second energy threshold;

    for each frequency band, if the average correlation is greater than the first correlation threshold but less than a second correlation threshold, determine that there is weak wind noise in the frequency band responsive to the average energy being greater than the second energy threshold, and determine that there is no wind noise in the frequency band responsive to the average energy being less than the second energy threshold; and

    for each frequency band, if the average correlation is greater than the second correlation threshold, determine that there is no wind noise in the frequency band.



    [0135] Each of the units of the device for detecting wind noise shown in FIG. 9 substantially corresponds to each of the operations of the method embodiment for detecting wind noise shown in FIG. 2, thus the relevant descriptions thereof may refer to descriptions of the part of the method embodiment of FIG. 2. Furthermore, the determination logic of the wind noise determination unit 950 may refer to the schematic logical diagram of determining wind noise conditions shown in FIG. 3. The method and device for detecting wind noise provided by the embodiments of the disclosure consider both the correlation of signals of the two microphones and energy of signals of the microphones when band-wise determining wind noise, band-wise detect wind noise conditions according to the correlation of signals of the feedforward microphone and the talk microphone at different frequency bands as well as energy of the signal of the feedforward microphone at a corresponding frequency band, which reduces the misjudgment of the wind noise effectively and increases the accuracy of detecting the wind noise.

    [0136] The above descriptions are only specific implementations of the disclosure, and under the above teachings of the disclosure, those skilled in the art may make other improvements or modifications based on the above embodiments. Those skilled in the art should understand that the above specific descriptions are only for better explaining the objective of the disclosure, and the protection scope of the disclosure should be subject to the protection scope of the claims.


    Claims

    1. A signal processing method for an earphone, comprising:

    detecting a motion state of a wearer of the earphone by using an acceleration sensor arranged inside the earphone (110);

    detecting wind noise conditions corresponding to different frequency bands by using a first microphone and a second microphone both arranged outside the earphone (120); and

    adjusting, according to the motion state of the wearer of the earphone and the wind noise conditions corresponding to different frequency bands, operating modes of a feedforward filter and a feedback filter inside the earphone, wherein the feedforward filter and the feedback filter are configured for active noise cancellation of the earphone (130),

    characterized in that, detecting the wind noise conditions corresponding to different frequency bands comprises:

    acquiring, according to signals collected by the first microphone and the second microphone, an average correlation of signals of the two microphones corresponding to each of three frequency bands including a low frequency band, a mid frequency band and a high frequency band respectively, and an average energy of the signals of the first microphone corresponding to each of the three frequency bands, wherein the first microphone is a feedforward microphone and the second microphone is a talk microphone; and

    determining, according to the acquired average correlations and average energies, whether there is wind noise in the three frequency bands, and intensities of the wind noise in the three frequency bands.


     
    2. The signal processing method for an earphone of claim 1, wherein the acquiring step comprises:

    acquiring, according to time domain signals collected by the first microphone and the second microphone, frequency domain signals of the first microphone and the second microphone at each of frequency points; and

    dividing frequency bands of the signals into three frequency bands including a low frequency band, a mid frequency band and a high frequency band, and acquiring, according to the acquired frequency domain signals of the first microphone and the second microphone at each of frequency points, the average correlation of signals of the two microphones corresponding to each of the three frequency bands including the low frequency band, the mid frequency band and the high frequency band respectively, and the average energy of the signals of the first microphone corresponding to each of the three frequency bands.


     
    3. The signal processing method for an earphone of claim 1, wherein the determining, step comprises:

    for each frequency band, if the average correlation is less than a first correlation threshold, determining that there is strong wind noise in the frequency band responsive to the average energy being greater than a first energy threshold, determining that there is weak wind noise in the frequency band responsive to the average energy being less than the first energy threshold but greater than a second energy threshold, and determining that there is no wind noise in the frequency band responsive to the average energy being less than the second energy threshold;

    for each frequency band, if the average correlation is greater than the first correlation threshold but less than a second correlation threshold, determining that there is weak wind noise in the frequency band responsive to the average energy being greater than the second energy threshold, and determining that there is no wind noise in the frequency band responsive to the average energy being less than the second energy threshold; and

    for each frequency band, if the average correlation is greater than the second correlation threshold, determining that there is no wind noise in the frequency band.


     
    4. The signal processing method for an earphone of any one of claims 1 to 3, further comprising: detecting an energy magnitude of environmental noise by using the first microphone, and considering the energy magnitude of the environmental noise when adjusting the operating modes of the feedforward filter and the feedback filter inside the earphone.
     
    5. The signal processing method for an earphone of claim 4, wherein detecting the energy magnitude of the environmental noise by using the first microphone comprises:

    performing exponential smoothing on the frequency domain signal of the first microphone at any of frequency points, and taking a minimum value of the smoothed signals within a set length of time to obtain the environmental noise signal of the frequency point in a current time frame; and

    superimposing the environmental noise signals of all the frequency points in the current time frame to obtain a total energy of the environmental noises.


     
    6. The signal processing method for an earphone of claim 4, wherein considering the energy magnitude of the environmental noise when adjusting the operating modes of the feedforward filter and the feedback filter inside the earphone comprises:
    when the wearer of the earphone is in a motion state and there is no wind noise in the low frequency band, selecting a higher noise cancellation level when the environmental noise is large in the low frequency band.
     
    7. The signal processing method for an earphone of claim 4, wherein considering the energy magnitude of the environmental noise when adjusting the operating modes of the feedforward filter and the feedback filter inside the earphone comprises:
    when the wearer of the earphone is in the non-motion state and there is no wind noise or weak wind noise in the low frequency band, dividing the noise cancellation level into different levels according to the energy magnitude of the environmental noise, wherein each level has its corresponding preset coefficients of the feedforward filter and the feedback filter; when the energy of the environmental noise is high, selecting a noise cancellation filter with high noise cancellation level, otherwise, selecting a noise cancellation filter with low noise cancellation level.
     
    8. A signal processing device for an earphone, comprising:

    a motion state detection module (610) configured to detect a motion state of a wearer of the earphone by using an acceleration sensor arranged inside the earphone;

    a wind noise detection module (620) configured to detect wind noise conditions corresponding to different frequency bands by using a first microphone and a second microphone both arranged outside the earphone; and

    a noise cancellation control module (640) configured to adjust, according to the motion state of the wearer of the earphone and the wind noise conditions corresponding to different frequency bands, operating modes of a feedforward filter and a feedback filter inside the earphone, wherein the feedforward filter and the feedback filter are configured for active noise cancellation of the earphone,

    characterized in that, the wind noise detection module (620) comprises:

    a band-wise processing unit (621) configured to acquire, according to signals collected by the first microphone and the second microphone, an average correlation of signals of the two microphones corresponding to each of three frequency bands including a low frequency band, a mid frequency band and a high frequency band respectively, and an average energy of the signals of the first microphone corresponding to each of the three frequency bands, wherein the first microphone is a feedforward microphone and the second microphone is a talk microphone; and

    a wind noise determination unit (622) configured to determine, according to the acquired average correlations and average energies, whether there is wind noise in the three frequency bands, and intensities of the wind noise in the three frequency bands.


     
    9. The signal processing device for an earphone of claim 8, wherein the band-wise processing unit (621) is specifically configured to:

    acquire, according to time domain signals collected by the first microphone and the second microphone, frequency domain signals of the first microphone and the second microphone at each of frequency points; and

    divide frequency bands of the signals into three frequency bands including a low frequency band, a mid frequency band and a high frequency band, and acquire, according to the acquired frequency domain signals of the first microphone and the second microphone at each of frequency points, the average correlation of signals of two microphones corresponding to each of the three frequency bands including the low frequency band, the mid frequency band and the high frequency band respectively, and the average energy of the signals of the first microphone corresponding to each of the three frequency bands.


     
    10. The signal processing device for an earphone of claim 8, wherein the wind noise determination unit (622) is specifically configured to:

    for each frequency band, if the average correlation is less than a first correlation threshold, determine that there is strong wind noise in the frequency band responsive to the average energy being greater than a first energy threshold, determine that there is weak wind noise in the frequency band responsive to the average energy being less than the first energy threshold but greater than a second energy threshold, and determine that there is no wind noise in the frequency band responsive to the average energy being less than the second energy threshold;

    for each frequency band, if the average correlation is greater than the first correlation threshold but less than a second correlation threshold, determine that there is weak wind noise in the frequency band responsive to the average energy being greater than the second energy threshold, and determine that there is no wind noise in the frequency band responsive to the average energy being less than the second energy threshold; and

    for each frequency band, if the average correlation is greater than the second correlation threshold, determine that there is no wind noise in the frequency band.


     
    11. The signal processing device for an earphone of any one of claims 8 to 10, further comprising:

    an environmental noise detection module (730) configured to detect an energy magnitude of environmental noise by using the first microphone;

    the noise cancellation control module (640) further configured to consider the energy magnitude of the environmental noise when adjusting the operating modes of the feedforward filter and the feedback filter inside the earphone.


     
    12. The signal processing device for an earphone of claim 11, wherein the environmental noise detection module (730) is specifically configured to:

    perform exponential smoothing on the frequency domain signal of the first microphone at any of frequency points, and take a minimum value of the smoothed signals within a set length of time to obtain the environmental noise signal of the frequency point in a current time frame; and

    superimpose the environmental noise signals of all the frequency points in the current time frame to obtain a total energy of the environmental noises.


     
    13. The signal processing device for an earphone of claim 11, wherein the noise cancellation control module (640) is specifically configured to:
    when the wearer of the earphone is in the motion state and there is no wind noise in the low frequency band, select a higher noise cancellation level when the environmental noise is large in the low frequency band.
     
    14. The signal processing device for an earphone of claim 11, wherein the noise cancellation control module (640) is specifically configured to:
    when the wearer of the earphone is in the non-motion state, and there is no wind noise or weak wind noise in the low frequency band, divide the noise cancellation level into different levels according to the energy magnitude of the environmental noise, wherein each level has its corresponding preset coefficients of the feedforward filter and the feedback filter; when the energy of the environmental noise is high, select a noise cancellation filter with high noise cancellation level, otherwise, select a noise cancellation filter with low noise cancellation level.
     
    15. An earphone, comprising: an acceleration sensor arranged inside the earphone, a first microphone and a second microphone both arranged outside the earphone, and a feedforward filter and a feedback filter inside the earphone, wherein the earphone is further provided therein with the signal processing device for an earphone of claim any one of claims 8 to 14.
     


    Ansprüche

    1. Signalverarbeitungsverfahren für einen Ohrhörer, umfassend:

    Erfassen eines Bewegungszustandes eines Trägers des Ohrhörers durch Verwenden eines Beschleunigungssensors, der im Inneren des Ohrhörers (110) angeordnet ist;

    Erfassen von Windgeräuschbedingungen, die verschiedenen Frequenzbändern entsprechen, durch Verwenden eines ersten Mikrofons und eines zweiten Mikrofons, die beide außerhalb des Ohrhörers (120) angeordnet sind; und

    Einstellen, entsprechend dem Bewegungszustand des Trägers des Ohrhörers und den Windgeräuschbedingungen, die verschiedenen Frequenzbändern entsprechen, von Betriebsarten eines Mitkopplungsfilters und eines Rückkopplungsfilters innerhalb des Ohrhörers, wobei der Mitkopplungsfilter und der Rückkopplungsfilter für eine aktive Geräuschunterdrückung des Ohrhörers (130) konfiguriert sind,

    dadurch gekennzeichnet, dass ein Erfassen der Windgeräuschbedingungen, die verschiedenen Frequenzbändern entsprechen, Folgendes umfasst:

    Erfassen, gemäß Signalen, die von dem ersten Mikrofon und dem zweiten Mikrofon gesammelt werden, einer durchschnittlichen Korrelation von Signalen der zwei Mikrofone, die jedem der drei Frequenzbänder entsprechen, die ein Niederfrequenzband, ein Mittelfrequenzband bzw. ein Hochfrequenzband beinhalten, und einer durchschnittlichen Energie der Signale des ersten Mikrofons, die jedem der drei Frequenzbänder entsprechen, wobei das erste Mikrofon ein Mitkopplungsmikrofon und das zweite Mikrofon ein Sprechmikrofon ist; und

    Bestimmen, gemäß der erfassten durchschnittlichen Korrelationen und den durchschnittlichen Energien, ob es in den drei Frequenzbändern Windgeräusche gibt, und von Intensitäten der Windgeräusche in den drei Frequenzbändern.


     
    2. Signalverarbeitungsverfahren für einen Ohrhörer nach Anspruch 1, wobei der Erfassungsschritt Folgendes umfasst:

    Erfassen, gemäß den von dem ersten Mikrofon und dem zweiten Mikrofon erfassten Zeitbereichssignalen, von Frequenzbereichssignalen des ersten Mikrofons und des zweiten Mikrofons an jedem der Frequenzpunkte; und

    Unterteilen von Frequenzbändern der Signale in drei Frequenzbänder, die ein Niederfrequenzband, ein Mittelfrequenzband bzw. ein Hochfrequenzband beinhalten, und Erfassen, gemäß den erfassten Frequenzbereichssignalen des ersten Mikrofons und des zweiten Mikrofons an jedem der Frequenzpunkte, der durchschnittlichen Korrelation von Signalen der zwei Mikrofone, die jedem der drei Frequenzbänder, entsprechen, die das Niederfrequenzband, das Mittelfrequenzband bzw. das Hochfrequenzband beinhalten, und der durchschnittlichen Energie der Signale des ersten Mikrofons, die jedem der drei Frequenzbänder entsprechen.


     
    3. Signalverarbeitungsverfahren für einen Ohrhörer nach Anspruch 1, wobei der Bestimmungsschritt Folgendes umfasst:

    für jedes Frequenzband, wenn die durchschnittliche Korrelation kleiner als ein erster Korrelationsschwellenwert ist, Bestimmen, dass es starkes Windgeräusch in dem Frequenzband gibt, als Reaktion darauf, dass die durchschnittliche Energie größer ist als ein erster Energieschwellenwert, Bestimmen, dass es schwaches Windgeräusch in dem Frequenzband gibt, als Reaktion darauf, dass die durchschnittliche Energie kleiner als der erste Energieschwellenwert aber größer als ein zweiter Energieschwellenwert ist, und Bestimmen, dass es kein Windgeräusch in dem Frequenzband gibt, als Reaktion darauf, dass die durchschnittliche Energie kleiner ist als der zweite Energieschwellenwert;

    für jedes Frequenzband, wenn die durchschnittliche Korrelation größer als der erste Korrelationsschwellenwert, aber kleiner als ein zweiter Korrelationsschwellenwert ist, Bestimmen, dass es schwaches Windgeräusch in dem Frequenzband gibt, als Reaktion darauf, dass die durchschnittliche Energie größer ist als der zweite Energieschwellenwert, und Bestimmen, dass es kein Windgeräusch in dem Frequenzband gibt, als Reaktion darauf, dass die durchschnittliche Energie kleiner ist als der zweite Energieschwellenwert; und

    für jedes Frequenzband, wenn die durchschnittliche Korrelation größer ist als der zweite Korrelationsschwellenwert, Bestimmen, dass es in dem Frequenzband kein Windgeräusch gibt.


     
    4. Signalverarbeitungsverfahren für einen Ohrhörer nach einem der Ansprüche 1 bis 3, ferner umfassend: Erfassen einer Energiegröße von Umgebungsgeräusch durch Verwenden des ersten Mikrofons und Berücksichtigen der Energiegröße des Umgebungsgeräuschs beim Einstellen der Betriebsarten des Mitkopplungsfilters und des Rückkopplungsfilters innerhalb des Ohrhörers.
     
    5. Signalverarbeitungsverfahren für einen Ohrhörer nach Anspruch 4, wobei ein Erfassen der Energiegröße des Umgebungsgeräuschs durch Verwenden des ersten Mikrofons Folgendes umfasst:

    Ausführen eines exponentiellen Glättens des Frequenzbereichssignals des ersten Mikrofons an einem beliebigen der Frequenzpunkte und Nehmen eines Minimalwertes der geglätteten Signale innerhalb einer festgelegten Zeitspanne, um das Umgebungsgeräuschsignal des Frequenzpunkts in einem aktuellen Zeitrahmen zu erlangen; und

    Überlagern der Umweltgeräuschsignale aller Frequenzpunkte in dem aktuellen Zeitrahmen, um eine Gesamtenergie der Umweltgeräusche zu erlangen.


     
    6. Signalverarbeitungsverfahren für einen Ohrhörer nach Anspruch 4, wobei ein Berücksichtigen der Energiegröße des Umgebungsgeräuschs beim Einstellen der Betriebsarten des Mitkopplungsfilters und des Rückkopplungsfilters innerhalb des Ohrhörers Folgendes umfasst:
    wenn der Träger des Ohrhörers in einem Bewegungszustand ist und es kein Windgeräusch in dem Niederfrequenzband gibt, Auswählen eines höheren Geräuschunterdrückungspegels, wenn das Umgebungsgeräusch in dem Niederfrequenzband stark ist.
     
    7. Signalverarbeitungsverfahren für einen Ohrhörer nach Anspruch 4, wobei ein Berücksichtigen der Energiegröße des Umgebungsgeräuschs beim Einstellen der Betriebsarten des Mitkopplungsfilters und des Rückkopplungsfilters innerhalb des Ohrhörers Folgendes umfasst:
    wenn der Träger des Ohrhörers in dem Nicht-Bewegungszustand ist und es kein Windgeräusch oder ein schwaches Windgeräusch in dem Niederfrequenzband gibt, Unterteilen des Geräuschunterdrückungspegels in verschiedene Pegel gemäß der Energiegröße des Umgebungsgeräuschs, wobei jeder Pegel seine entsprechenden voreingestellten Koeffizienten des Mitkopplungsfilters und des Rückkopplungsfilters aufweist; wenn die Energie des Umgebungsgeräuschs hoch ist, Auswählen eines Geräuschunterdrückungsfilters mit hohem Geräuschunterdrückungspegel, sonst Auswählen eines Geräuschunterdrückungsfilters mit niedrigem Geräuschunterdrückungspegel.
     
    8. Signalverarbeitungsvorrichtung für einen Ohrhörer, umfassend:

    ein Bewegungszustand-Erfassungsmodul (610), das konfiguriert ist, um einen Bewegungszustand eines Trägers des Ohrhörers durch Verwenden eines im Inneren des Ohrhörers angeordneten Beschleunigungssensors zu erfassen;

    ein Windgeräusch-Erfassungsmodul (620), das konfiguriert ist, um Windgeräuschbedingungen, die verschiedenen Frequenzbändern entsprechen, durch Verwenden eines ersten Mikrofons und eines zweiten Mikrofons, die beide außerhalb des Ohrhörers angeordnet sind, zu erfassen; und

    ein Geräuschunterdrückung-Steuermodul (640), das konfiguriert ist, um gemäß dem Bewegungszustand des Trägers des Ohrhörers und den Windgeräuschbedingungen, die verschiedenen Frequenzbändern entsprechen, Betriebsarten eines Mitkopplungsfilters und eines Rückkopplungsfilters innerhalb des Ohrhörers einzustellen, wobei der Mitkopplungsfilter und der Rückkopplungsfilter für eine aktive Geräuschunterdrückung des Ohrhörers konfiguriert sind,

    dadurch gekennzeichnet, dass das Windgeräusch-Erfassungsmodul (620) Folgendes umfasst:

    eine bandweise Verarbeitungseinheit (621), die konfiguriert ist, um, gemäß Signalen, die von dem ersten Mikrofon und dem zweiten Mikrofon gesammelt werden, eine durchschnittliche Korrelation von Signalen der zwei Mikrofone, die jedem der drei Frequenzbänder entsprechen, die ein Niederfrequenzband, ein Mittelfrequenzband bzw. ein Hochfrequenzband beinhalten, und eine durchschnittliche Energie der Signale des ersten Mikrofons, die jedem der drei Frequenzbänder entsprechen, wobei das erste Mikrofon ein Mitkopplungsmikrofon und das zweite Mikrofon ein Sprechmikrofon ist, zu erfassen; und

    eine Windgeräusch-Bestimmungseinheit (622), die konfiguriert ist, um gemäß den erfassten durchschnittlichen Korrelationen und durchschnittlichen Energien zu bestimmen, ob es in den drei Frequenzbändern Windgeräusche gibt, und von Intensitäten der Windgeräusche in den drei Frequenzbändern.


     
    9. Signalverarbeitungsvorrichtung für einen Ohrhörer nach Anspruch 8, wobei die bandweise Verarbeitungseinheit (621) spezifisch zu Folgendem konfiguriert ist:

    Erfassen, gemäß den von dem ersten Mikrofon und dem zweiten Mikrofon erfassten Zeitbereichssignalen, von Frequenzbereichssignalen des ersten Mikrofons und des zweiten Mikrofons an jedem der Frequenzpunkte; und

    Unterteilen von Frequenzbändern der Signale in drei Frequenzbänder, die ein Niederfrequenzband, ein Mittelfrequenzband bzw. ein Hochfrequenzband beinhalten, und Erfassen, gemäß den erfassten Frequenzbereichssignalen des ersten Mikrofons und des zweiten Mikrofons an jedem der Frequenzpunkte, der durchschnittlichen Korrelation von Signalen von zwei Mikrofonen, die jedem der drei Frequenzbänder, entsprechen, die das Niederfrequenzband, das Mittelfrequenzband bzw. das Hochfrequenzband beinhalten, und der durchschnittlichen Energie der Signale des ersten Mikrofons, die jedem der drei Frequenzbänder entsprechen.


     
    10. Signalverarbeitungsvorrichtung für einen Ohrhörer nach Anspruch 8, wobei die Windgeräusch-Bestimmungseinheit (622) spezifisch zu Folgendem konfiguriert ist:

    für jedes Frequenzband, wenn die durchschnittliche Korrelation kleiner als ein erster Korrelationsschwellenwert ist, Bestimmen, dass es starkes Windgeräusch in dem Frequenzband gibt, als Reaktion darauf, dass die durchschnittliche Energie größer ist als ein erster Energieschwellenwert, Bestimmen, dass es schwaches Windgeräusch in dem Frequenzband gibt, als Reaktion darauf, dass die durchschnittliche Energie kleiner als der erste Energieschwellenwert aber größer als ein zweiter Energieschwellenwert ist, und Bestimmen, dass es kein Windgeräusch in dem Frequenzband gibt, als Reaktion darauf, dass die durchschnittliche Energie kleiner ist als der zweite Energieschwellenwert;

    für jedes Frequenzband, wenn die durchschnittliche Korrelation größer als der erste Korrelationsschwellenwert, aber kleiner als ein zweiter Korrelationsschwellenwert ist, Bestimmen, dass es schwaches Windgeräusch in dem Frequenzband gibt, als Reaktion darauf, dass die durchschnittliche Energie größer ist als der zweite Energieschwellenwert, und Bestimmen, dass es kein Windgeräusch in dem Frequenzband gibt, als Reaktion darauf, dass die durchschnittliche Energie kleiner ist als der zweite Energieschwellenwert; und

    für jedes Frequenzband, wenn die durchschnittliche Korrelation größer ist als der zweite Korrelationsschwellenwert, Bestimmen, dass es in dem Frequenzband kein Windgeräusch gibt.


     
    11. Signalverarbeitungsvorrichtung für einen Ohrhörer nach einem der Ansprüche 8 bis 10, ferner umfassend:

    ein Umgebungsgeräusch-Erfassungsmodul (730), das konfiguriert ist, um eine Energiegröße von Umgebungsgeräusch durch Verwenden des ersten Mikrofons zu erfassen;

    wobei das Geräuschunterdrückung-Steuermodul (640) ferner konfiguriert ist, um beim Einstellen der Betriebsarten des Mitkopplungsfilters und des Rückkopplungsfilters innerhalb des Ohrhörers die Energiegröße des Umgebungsgeräuschs zu berücksichtigen.


     
    12. Signalverarbeitungsvorrichtung für einen Ohrhörer nach Anspruch 11, wobei das Umgebungsgeräusch-Erfassungsmodul (730) spezifisch zu Folgendem konfiguriert ist:

    Ausführen eines exponentiellen Glättens des Frequenzbereichssignals des ersten Mikrofons an einem beliebigen der Frequenzpunkte und Nehmen eines Minimalwerts der geglätteten Signale innerhalb einer festgelegten Zeitspanne, um das Umgebungsgeräuschsignal des Frequenzpunkts in einem aktuellen Zeitrahmen zu erlangen; und

    Überlagern der Umweltgeräuschsignale aller Frequenzpunkte in dem aktuellen Zeitrahmen, um eine Gesamtenergie der Umweltgeräusche zu erlangen.


     
    13. Signalverarbeitungsvorrichtung für einen Ohrhörer nach Anspruch 11, wobei das Geräuschunterdrückung-Steuermodul (640) spezifisch zu Folgendem konfiguriert ist:
    wenn der Träger des Ohrhörers in dem Bewegungszustand ist und es kein Windgeräusch in dem Niederfrequenzband gibt, Auswählen eines höheren Geräuschunterdrückungspegels, wenn das Umgebungsgeräusch in dem Niederfrequenzband stark ist.
     
    14. Signalverarbeitungsvorrichtung für einen Ohrhörer nach Anspruch 11, wobei das Geräuschunterdrückung-Steuermodul (640) spezifisch zu Folgendem konfiguriert ist:
    wenn der Träger des Ohrhörers in dem Nicht-Bewegungszustand ist und es kein Windgeräusch oder ein schwaches Windgeräusch in dem Niederfrequenzband gibt, Unterteilen des Geräuschunterdrückungspegels in verschiedene Pegel gemäß der Energiegröße des Umgebungsgeräuschs, wobei jeder Pegel seine entsprechenden voreingestellten Koeffizienten des Mitkopplungsfilters und des Rückkopplungsfilters aufweist; wenn die Energie des Umgebungsgeräuschs hoch ist, Auswählen eines Geräuschunterdrückungsfilters mit hohem Geräuschunterdrückungspegel, sonst Auswählen eines Geräuschunterdrückungsfilters mit niedrigem Geräuschunterdrückungspegel.
     
    15. Ohrhörer, umfassend: einen Beschleunigungssensor, der innerhalb des Ohrhörers angeordnet ist, ein erstes Mikrofon und ein zweites Mikrofon, die beide außerhalb des Ohrhörers angeordnet sind, und einen Mitkopplungsfilter und einen Rückkopplungsfilter innerhalb des Ohrhörers, wobei der Ohrhörer ferner mit der Signalverarbeitungsvorrichtung für einen Ohrhörer nach einem der Ansprüche 8 bis 14 versehen ist.
     


    Revendications

    1. Procédé de traitement de signal pour un écouteur, comprenant :

    la détection d'un état de mouvement d'un porteur de l'écouteur en utilisant un capteur d'accélération disposé à l'intérieur de l'écouteur (110) ;

    la détection des conditions de bruit du vent correspondant à différentes bandes de fréquence en utilisant un premier microphone et un second microphone tous deux disposés à l'extérieur de l'écouteur (120) ; et

    le réglage, selon l'état de mouvement du porteur de l'écouteur et les conditions de bruit du vent correspondant à différentes bandes de fréquence, des modes de fonctionnement d'un filtre d'anticipation et d'un filtre de rétroaction à l'intérieur de l'écouteur, dans lequel le filtre d'anticipation et le filtre de rétroaction sont configurés pour une annulation active du bruit de l'écouteur (130),

    caractérisé en ce que la détection des conditions de bruit du vent correspondant à différentes bandes de fréquence comprend :

    l'acquisition, en fonction des signaux collectés par le premier microphone et le second microphone, d'une corrélation moyenne des signaux des deux microphones correspondant à chacune de trois bandes de fréquences comprenant une bande de basses fréquences, une bande de moyennes fréquences et une bande de hautes fréquences respectivement, et d'une énergie moyenne des signaux du premier microphone correspondant à chacune des trois bandes de fréquences, dans lequel le premier microphone est un microphone d'anticipation et le second microphone est un microphone de conversation ; et

    la détermination, en fonction des corrélations moyennes et des énergies moyennes acquises, s'il existe un bruit de vent dans les trois bandes de fréquences, et la détermination des intensités du bruit de vent dans les trois bandes de fréquences.


     
    2. Procédé de traitement de signal pour un écouteur selon la revendication 1, dans lequel l'étape d'acquisition comprend :

    l'acquisition, selon les signaux du domaine temporel collectés par le premier microphone et le second microphone, de signaux du domaine fréquentiel du premier microphone et du second microphone à chacun des points de fréquence ; et

    la division des bandes de fréquences des signaux en trois bandes de fréquences comprenant une bande de basses fréquences, une bande de moyennes fréquences et une bande de hautes fréquences, et l'acquisition, selon les signaux acquis dans le domaine de fréquence du premier microphone et du second microphone à chacun des points de fréquence, de la corrélation moyenne des signaux des deux microphones correspondant à chacune des trois bandes de fréquences comprenant la bande de basses fréquences, la bande de moyennes fréquences et la bande de hautes fréquences respectivement, et de l'énergie moyenne des signaux du premier microphone correspondant à chacune des trois bandes de fréquences.


     
    3. Procédé de traitement de signal pour un écouteur selon la revendication 1, dans lequel l'étape de détermination comprend :

    pour chaque bande de fréquences, si la corrélation moyenne est inférieure à un premier seuil de corrélation, la détermination qu'il y a un fort bruit de vent dans la bande de fréquences en réponse au fait que l'énergie moyenne est supérieure à un premier seuil d'énergie, la détermination qu'il y a un faible bruit de vent dans la bande de fréquences en réponse au fait que l'énergie moyenne est inférieure au premier seuil d'énergie, mais supérieure à un second seuil d'énergie, et la détermination qu'il n'y a pas de bruit de vent dans la bande de fréquences en réponse au fait que l'énergie moyenne est inférieure au second seuil d'énergie ;

    pour chaque bande de fréquences, si la corrélation moyenne est supérieure au premier seuil de corrélation, mais inférieure à un second seuil de corrélation, la détermination qu'il y a un faible bruit de vent dans la bande de fréquences en réponse à l'énergie moyenne supérieure au second seuil d'énergie, et la détermination qu'il n'y a pas de bruit de vent dans la bande de fréquences en réponse à l'énergie moyenne inférieure au second seuil d'énergie ; et

    pour chaque bande de fréquences, si la corrélation moyenne est supérieure au second seuil de corrélation, la détermination qu'il n'y a pas de bruit de vent dans la bande de fréquences.


     
    4. Procédé de traitement de signal pour un écouteur selon l'une quelconque des revendications 1 à 3, comprenant en outre : la détection d'une amplitude d'énergie d'un bruit ambiant en utilisant le premier microphone, et la prise en compte de l'amplitude d'énergie du bruit ambiant lors du réglage des modes de fonctionnement du filtre d'anticipation et du filtre de rétroaction à l'intérieur de l'écouteur.
     
    5. Procédé de traitement de signal pour un écouteur selon la revendication 4, dans lequel la détection de l'amplitude d'énergie du bruit ambiant en utilisant le premier microphone comprend :

    la réalisation d'un lissage exponentiel sur le signal de domaine de fréquence du premier microphone à l'un quelconque des points de fréquence, et la prise d'une valeur minimale des signaux lissés dans une durée déterminée pour obtenir le signal de bruit ambiant du point de fréquence dans une trame de temps actuelle ; et

    la superposition des signaux de bruit ambiant de tous les points de fréquence dans la trame de temps actuelle pour obtenir une énergie totale des bruits ambiants.


     
    6. Procédé de traitement de signal pour un écouteur selon la revendication 4, dans lequel la prise en compte de l'amplitude d'énergie du bruit ambiant lors du réglage des modes de fonctionnement du filtre d'anticipation et du filtre de rétroaction à l'intérieur de l'écouteur comprend :
    lorsque le porteur de l'écouteur se trouve dans un état de mouvement et qu'il n'y a pas de bruit de vent dans la bande de basses fréquences, la sélection d'un niveau d'annulation de bruit plus élevé lorsque le bruit ambiant est important dans la bande de basses fréquences.
     
    7. Procédé de traitement de signal pour un écouteur selon la revendication 4, dans lequel la prise en compte de l'amplitude d'énergie du bruit ambiant lors du réglage des modes de fonctionnement du filtre d'anticipation et du filtre de rétroaction à l'intérieur de l'écouteur comprend :
    lorsque le porteur de l'écouteur se trouve dans l'état d'absence de mouvement et qu'il n'y a pas de bruit de vent ou qu'il y a un faible bruit de vent dans la bande de basses fréquences, la division du niveau d'annulation de bruit en différents niveaux selon l'amplitude d'énergie du bruit ambiant, dans lequel chaque niveau a ses coefficients prédéfinis correspondants du filtre d'anticipation et du filtre de rétroaction ; lorsque l'énergie du bruit ambiant est élevée, la sélection d'un filtre d'annulation de bruit avec un niveau d'annulation de bruit élevé, sinon, la sélection d'un filtre d'annulation de bruit avec un niveau d'annulation de bruit faible.
     
    8. Dispositif de traitement de signal pour un écouteur, comprenant :

    un module de détection d'état de mouvement (610) configuré pour détecter un état de mouvement d'un porteur de l'écouteur en utilisant un capteur d'accélération disposé à l'intérieur de l'écouteur ;

    un module de détection du bruit du vent (620) configuré pour détecter les conditions de bruit du vent correspondant à différentes bandes de fréquences en utilisant un premier microphone et un second microphone tous deux disposés à l'extérieur de l'écouteur ; et

    un module de commande d'annulation de bruit (640) configuré pour régler, selon l'état de mouvement du porteur de l'écouteur et les conditions de bruit du vent correspondant à différentes bandes de fréquences, les modes de fonctionnement d'un filtre d'anticipation et d'un filtre de rétroaction à l'intérieur de l'écouteur, dans lequel le filtre d'anticipation et le filtre de rétroaction sont configurés pour une annulation active du bruit de l'écouteur,

    caractérisé en ce que, le module de détection du bruit du vent (620) comprend :

    une unité de traitement par bande (621) configurée pour acquérir, selon les signaux collectés par le premier microphone et le second microphone, une corrélation moyenne des signaux des deux microphones correspondant à chacune des trois bandes de fréquences comprenant une bande de basses fréquences, une bande de moyennes fréquences et une bande de hautes fréquences respectivement, et une énergie moyenne des signaux du premier microphone correspondant à chacune des trois bandes de fréquences, dans lequel le premier microphone est un microphone d'anticipation et le second microphone est un microphone de conversation ; et

    une unité de détermination de bruit de vent (622) configurée pour déterminer, selon les corrélations moyennes et les énergies moyennes acquises, s'il y a un bruit de vent dans les trois bandes de fréquences, et les intensités du bruit de vent dans les trois bandes de fréquences.


     
    9. Dispositif de traitement de signal pour un écouteur selon la revendication 8, dans lequel l'unité de traitement par bande (621) est spécifiquement configurée pour :

    acquérir, selon les signaux du domaine temporel collectés par le premier microphone et le second microphone, les signaux du domaine fréquentiel du premier microphone et du second microphone à chacun des points de fréquence ; et

    diviser les bandes de fréquences des signaux en trois bandes de fréquences comprenant une bande de basses fréquences, une bande de moyennes fréquences et une bande de hautes fréquences, et acquérir, selon les signaux acquis dans le domaine de fréquence du premier microphone et du second microphone à chacun des points de fréquence, la corrélation moyenne des signaux de deux microphones correspondant à chacune des trois bandes de fréquences comprenant la bande de basses fréquences, la bande de moyennes fréquences et la bande de hautes fréquences respectivement, et l'énergie moyenne des signaux du premier microphone correspondant à chacune des trois bandes de fréquences.


     
    10. Dispositif de traitement de signal pour un écouteur selon la revendication 8, dans lequel l'unité de détermination de bruit de vent (622) est spécifiquement configurée pour :

    pour chaque bande de fréquences, si la corrélation moyenne est inférieure à un premier seuil de corrélation, déterminer qu'il y a un fort bruit de vent dans la bande de fréquences en réponse au fait que l'énergie moyenne est supérieure à un premier seuil d'énergie, déterminer qu'il y a un faible bruit de vent dans la bande de fréquences en réponse au fait que l'énergie moyenne est inférieure au premier seuil d'énergie, mais supérieure à un second seuil d'énergie, et déterminer qu'il n'y a pas de bruit de vent dans la bande de fréquences en réponse au fait que l'énergie moyenne est inférieure au second seuil d'énergie ;

    pour chaque bande de fréquences, si la corrélation moyenne est supérieure au premier seuil de corrélation, mais inférieure à un second seuil de corrélation, déterminer qu'il y a un faible bruit de vent dans la bande de fréquences en réponse à l'énergie moyenne supérieure au second seuil d'énergie, et déterminer qu'il n'y a pas de bruit de vent dans la bande de fréquences en réponse à l'énergie moyenne inférieure au second seuil d'énergie ; et

    pour chaque bande de fréquences, si la corrélation moyenne est supérieure au second seuil de corrélation, déterminer qu'il n'y a pas de bruit de vent dans la bande de fréquences.


     
    11. Dispositif de traitement de signal pour un écouteur selon l'une quelconque des revendications 8 à 10, comprenant en outre :

    un module de détection de bruit ambiant (730) configuré pour détecter une amplitude d'énergie du bruit ambiant en utilisant le premier microphone ;

    le module de commande d'annulation de bruit (640) configuré en outre pour prendre en compte l'amplitude d'énergie du bruit ambiant lors du réglage des modes de fonctionnement du filtre d'anticipation et du filtre de rétroaction à l'intérieur de l'écouteur.


     
    12. Dispositif de traitement de signal pour un écouteur selon la revendication 11, dans lequel le module de détection de bruit ambiant (730) est spécifiquement configuré pour :

    effectuer un lissage exponentiel sur le signal de domaine de fréquence du premier microphone à l'un quelconque des points de fréquence, et prendre une valeur minimale des signaux lissés dans une durée déterminée pour obtenir le signal de bruit ambiant du point de fréquence dans une trame temporelle actuelle ; et

    superposer les signaux de bruit ambiant de tous les points de fréquence dans la trame temporelle actuelle pour obtenir une énergie totale des bruits ambiants.


     
    13. Dispositif de traitement de signal pour un écouteur selon la revendication 11, dans lequel le module de commande d'annulation de bruit (640) est spécifiquement configuré pour :
    lorsque le porteur de l'écouteur se trouve dans un état de mouvement et qu'il n'y a pas de bruit de vent dans la bande de basses fréquences, sélectionner un niveau d'annulation de bruit plus élevé lorsque le bruit ambiant est important dans la bande de basses fréquences.
     
    14. Dispositif de traitement de signal pour un écouteur selon la revendication 11, dans lequel le module de commande d'annulation de bruit (640) est spécifiquement configuré pour :
    lorsque le porteur de l'écouteur se trouve dans l'état d'absence de mouvement et qu'il n'y a pas de bruit de vent ou qu'il y a un faible bruit de vent dans la bande de basses fréquences, diviser le niveau d'annulation de bruit en différents niveaux selon l'amplitude d'énergie du bruit ambiant, dans lequel chaque niveau a ses coefficients prédéfinis correspondants du filtre d'anticipation et du filtre de rétroaction ; lorsque l'énergie du bruit ambiant est élevée, sélectionner un filtre d'annulation de bruit avec un niveau d'annulation de bruit élevé, sinon, sélectionner un filtre d'annulation de bruit avec un niveau d'annulation de bruit faible.
     
    15. Écouteur, comprenant: un capteur d'accélération disposé à l'intérieur de l'écouteur, un premier microphone et un second microphone tous deux disposés à l'extérieur de l'écouteur, et un filtre d'anticipation et un filtre de rétroaction à l'intérieur de l'écouteur, dans lequel l'écouteur est en outre pourvu à l'intérieur du dispositif de traitement de signal pour un écouteur selon l'une quelconque des revendications 8 à 14.
     




    Drawing




















    Cited references

    REFERENCES CITED IN THE DESCRIPTION



    This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

    Patent documents cited in the description