WO2019034154A1

WO2019034154A1 - Noise reduction method and device for mobile terminal, and computer storage medium

Info

Publication number: WO2019034154A1
Application number: PCT/CN2018/101136
Authority: WO
Inventors: 于冰; 刘波; 谢姣; 孙家训
Original assignee: 西安中兴新软件有限责任公司
Priority date: 2017-08-17
Filing date: 2018-08-17
Publication date: 2019-02-21
Also published as: CN109413253A

Abstract

Provided in the present disclosure are a noise reduction method and device for a mobile terminal, and a computer storage medium. The method comprises: when a mobile terminal enters into a speaker mode, acquiring current orientation information of the mobile terminal; invoking a noise reduction parameter configuration document corresponding to the current orientation information of the mobile terminal from a pre-built noise reduction parameter calibration database; and performing an audio modification and noise reduction processing on two microphones of the mobile terminal according to the invoked noise reduction parameter configuration document.

Description

Noise reduction method, device and computer storage medium for mobile terminal

Technical field

The present disclosure relates to, but is not limited to, hand-held calling technology.

Background technique

Handheld voice calls on mobile phones are the most widely used basic functions. In a noisy environment, making a hand-held call makes it difficult for the other party to hear the spoken content. To solve this problem, a dual MIC including a primary microphone (MIC) and a secondary MIC can be used to eliminate background noise to enhance the uplink voice call effect. During the call, the mouth of the person is closer to the main MIC of the mobile phone and farther away from the auxiliary MIC. Therefore, the voice signal picked up by the main MIC is larger, and the signal picked up by the auxiliary MIC is smaller. The signal picked up by the main MIC subtracts the signal picked up by the auxiliary MIC to cancel out the background noise.

In the related art, the same noise reduction parameter is used for different postures of the user's handheld mobile phone, so that in different postures, the voice volume is too small, or the noise reduction effect is deteriorated, so that the call volume and the noise reduction effect are both The aspect cannot be balanced.

Summary of the invention

An embodiment of the present disclosure provides a method for reducing noise of a mobile terminal, including: collecting current posture information of the mobile terminal when the mobile terminal enters a handheld call mode; and calling and moving from a pre-established noise reduction parameter calibration database a noise reduction parameter configuration file corresponding to the current posture information of the terminal; and performing audio correction and noise reduction processing on the dual microphone of the mobile terminal according to the called noise reduction parameter configuration file.

An embodiment of the present disclosure further provides a noise reduction device of a mobile terminal, including: an acquisition unit, when the mobile terminal enters a handheld call mode, the collection unit collects current posture information of the mobile terminal; and a noise reduction processing unit And setting a noise reduction parameter configuration file corresponding to the current posture information of the mobile terminal from a preset noise reduction parameter calibration database, and performing audio on the dual microphone of the mobile terminal according to the invoked noise reduction parameter configuration file. Correction and noise reduction processing.

Embodiments of the present disclosure also provide a computer storage medium having stored thereon one or more programs, the one or more programs being executed by one or more processors, the one or more processors executing A noise reduction method of a mobile terminal according to the present disclosure.

DRAWINGS

The drawings are intended to provide a further understanding of the embodiments of the invention, and are in the In the drawing:

Figure 1 schematically shows the posture of a standard user-held mobile phone;

Figure 2 is a schematic illustration of the posture of a non-standard user-held mobile phone;

FIG. 3 is a flowchart of a noise reduction method of a mobile terminal according to an embodiment of the present disclosure; FIG.

4 is a structural block diagram of a noise reduction device of a mobile terminal according to an embodiment of the present disclosure;

FIG. 5 illustrates a schematic diagram of generating attitude coordinate data in accordance with an embodiment of the present disclosure;

6 shows a diagram of frequency response curve data for a primary microphone and a secondary microphone, in accordance with an embodiment of the present disclosure;

7 shows a schematic diagram of a fit curve between a pose and a parameter correction variable in accordance with an embodiment of the present disclosure;

FIG. 8 illustrates a flow chart of a noise reduction method of a mobile terminal according to an embodiment of the present disclosure.

Detailed ways

Embodiments of the present disclosure will be described in detail below with reference to specific embodiments. It should be noted that the features in the embodiments of the present disclosure may be combined with each other arbitrarily without conflict.

First, the basic principle of eliminating background noise with dual MICs is described. The voice signal V _MAIN picked up by the main MIC may include a call audio signal V _{CALL_MAIN} and a background noise signal V _{NOISE_MAIN} to be filtered out. The voice signal V _AUX picked up by the secondary MIC may include a call audio signal V _{CALL_AUX} and a background noise signal V _{NOISE_AUX} to be filtered out. The calculation formula for finally obtaining the speech signal V _VOICE by subtracting the speech signal V _AUX picked up by the auxiliary MIC from the speech signal V _MAIN picked up by the main MIC is as follows:

V _VOICE =V _MAIN -V _AUX =(V _{CALL_MAIN} +V _{NOISE_MAIN} )-(V _{CALL_AUX} +V _{NOISE_AUX} ) (1)

A noise reduction algorithm is needed to implement and optimize the noise reduction function of the dual MIC. In order to effectively eliminate the noise and at the same time ensure the volume of the normal call, it is necessary to obtain a V _VOICE signal with a suitable gain. Various noise reduction algorithms each define the gain range of V _VOICE to ensure the volume of the call and effective noise reduction. The gain range is defined as [A:B]dB, ie:

AdB<V _VOICE <BdB

The gain range [A:B] provided by different codec (CODEC) chips will be slightly different. In general, if A=6, B=9, the compatibility of the MIC noise reduction algorithm of each CODEC chip can be guaranteed. .

In addition, in practical applications, different handheld gestures of the user's handheld mobile phone may also have a certain impact on the MIC noise reduction algorithm. The effects of different handheld gestures on the MIC noise reduction algorithm will be described below with reference to FIGS. 1 and 2.

Fig. 1 schematically shows the posture of a standard user hand held mobile phone, and Fig. 2 schematically shows the posture of a non-standard user hand held mobile phone.

According to the above formula (1), the following formula (2) can be obtained:

V _VOICE =(V _{CALL_MAIN} -V _{CALL_AUX} )+(V _{NOISE_MAIN} -V _{NOISE_AUX} ) (2)

Since the background noises V _{NOISE_MAIN} and V _{NOISE_AUX} are basically the same size, equation (3) can be obtained:

V _VOICE =V _{CALL_MAIN} -V _{CALL_AUX} (3)

In the standard posture as shown in Fig. 1, V _{CALL_MAIN is} large and V _{CALL_AUX is} small, so it is easy to adjust V _VOICE within the range of [A: B] dB.

However, in the non-standard posture as shown in FIG. 2, the distance from the main MIC to the human mouth becomes _longer due to the change in the hand posture, which in turn causes V _{CALL_MAIN to} become smaller. On the other hand, V _{CALL_AUX} in the language signal V _AUX picked up by the auxiliary MIC does not substantially change, so that the V _VOICE obtained by the above formula (3) also becomes small. In particular, in some hand-held postures, the V _VOICE obtained by the above formula (3) does not fall within the range of [A: B] dB.

Therefore, there are significant differences in the frequency response curves of the main MIC in different hand-held postures. However, using a fixed noise reduction parameter can only increase or attenuate the gain of V _{CALL_MAIN} and V _{CALL_AUX} as a whole.

Table 1 shows the various indicator parameters for noise reduction using fixed noise reduction parameters 0 dB and 5 dB for different hand gestures. It will be appreciated that the V _{CALL_MAIN} and V _{CALL_AUX} values in Table 1 select typical values for the frequency response curve to roughly represent the average difference in gain of the two frequency response curves.

Table 1

It can be seen from Table 1 that in the case of using a fixed noise reduction parameter of 0 dB, for the posture 1 and the posture 2, the call loudness and the noise reduction effect can satisfy the requirement, and for the posture 3 and the posture 4, the call loudness is small and cannot be satisfied. Claim. In the case of using a fixed noise reduction parameter of 5 dB, the call loudness basically satisfies the requirement (the call loudness of posture 1 and posture 2 is too large), but the noise reduction effect is weakened, and the intended double MIC noise reduction is not achieved.

In summary, in the related art, if the non-standard posture and the standard posture of the user's hand-held mobile phone differ greatly, the volume of the call is reduced and the noise reduction effect is deteriorated, and the situation of both of these problems cannot be improved at the same time.

According to the present disclosure, the gesture information of the mobile phone in different handheld postures can be collected and different noise reduction parameter profiles can be generated in advance for calling in a hand-held call, thereby solving the problem of improving the voice volume and achieving noise reduction in a non-standard posture. problem.

FIG. 3 is a flowchart of a noise reduction method of a mobile terminal according to an embodiment of the present disclosure.

As shown in FIG. 3, the noise reduction method of the mobile terminal according to an implementation of the present disclosure may include steps S101 to S103.

In step S101, when the mobile terminal enters the handheld call mode, the current posture information of the mobile terminal is collected.

In step S102, a noise reduction parameter configuration file corresponding to the current posture information of the mobile terminal is called from the pre-established noise reduction parameter calibration database.

In step S103, audio correction and noise reduction processing are performed on the dual MIC of the mobile terminal according to the called noise reduction parameter profile.

According to an embodiment of the present disclosure, the pre-generated noise reduction parameter calibration database stores at least a noise reduction parameter profile corresponding to the posture information of the mobile terminal, wherein the posture information includes posture coordinate data of the mobile terminal, and the drop corresponding to the posture information The noise parameter profile includes a plurality of parameter correction variables for audio correction and noise reduction processing in the respective poses.

The noise reduction method of the mobile terminal according to an embodiment of the present disclosure may further include the step of establishing a noise reduction parameter calibration database in advance, the step may include: collecting various posture coordinate data of the mobile terminal; testing the mobile terminal under various posture coordinate data The main MIC and the auxiliary MIC audio data to generate frequency response curve data of the main MIC and the auxiliary MIC under different attitude coordinate data; in the generated frequency response curve data of the main MIC and the auxiliary MIC, select a plurality of typical Frequency points, and parameter correction variables are respectively set for the selected typical frequency points; and various posture coordinate data and corresponding plurality of parameter correction variables are fitted to generate a noise reduction parameter configuration file and stored.

According to an embodiment of the present disclosure, step S102 may include: searching, in the noise reduction parameter calibration database, posture coordinate data having the smallest difference from the posture coordinate data in the current posture information of the mobile terminal; and according to the obtained posture coordinate data and corresponding multiple a fitting result of the parameter correction variable generates a parameter correction variable corresponding to the posture coordinate data in the current posture information of the mobile terminal; and obtaining a noise reduction parameter configuration file corresponding to the current posture information of the mobile terminal according to the generated parameter correction variable .

According to an embodiment of the present disclosure, various attitude coordinate data of the mobile terminal may be acquired using an orientation sensor such as an acceleration transmission, a gyroscope, or the like.

4 is a structural block diagram of a noise reduction device of a mobile terminal according to an embodiment of the present disclosure.

As shown in FIG. 4, the device of the mobile terminal according to an embodiment of the present disclosure may include an acquisition unit 401 and a noise reduction processing unit 402.

When the mobile terminal enters the handheld call mode, the collecting unit 401 collects the current posture information of the mobile terminal. The noise reduction processing unit 402 is configured to call a noise reduction parameter configuration file corresponding to the current posture information of the mobile terminal from the pre-established noise reduction parameter calibration database 403, and configure the file to the dual MIC of the mobile terminal according to the called noise reduction parameter configuration file. Perform audio correction and noise reduction processing.

The frequency response curves of the main MIC and the auxiliary MIC are corrected at each frequency point to realize the main MIC and the auxiliary MIC frequency response curve for the noise reduction calculation during the call.

According to an embodiment of the present disclosure, the pre-generated noise reduction parameter calibration database 403 stores at least a noise reduction parameter profile corresponding to the posture information of the mobile terminal, wherein the posture information includes posture coordinate data of the mobile terminal, and corresponds to the posture information. The noise reduction parameter profile includes a plurality of parameter correction variables for audio correction and noise reduction processing in the respective poses.

The noise reducer according to an embodiment of the present disclosure may further include a noise reduction parameter calibration database generation unit 404. As shown in FIG. 4, the noise reduction parameter calibration database generating unit 404 may include an audio data testing module 4041, an audio parameter correction module 4042, and a noise reduction parameter calibration module 4043.

In an audio lab environment, the handset is fixed in various poses using a hand-held mode fixture, and the attitude coordinates data of the mobile terminal in various poses is acquired by the acquisition unit 401. The audio data testing module 4041 tests the audio data of the primary MIC and the secondary MIC of the mobile terminal under various attitude coordinate data to generate frequency response curve data of the primary MIC and the secondary MIC under different attitude coordinate data.

FIG. 5 illustrates a schematic diagram of generating attitude coordinate data in accordance with an embodiment of the present disclosure.

As shown in FIG. 5, various posture coordinate data of the mobile terminal can be acquired by an acceleration sensor, a gyroscope, or the like.

The audio parameter correction module 4042 is configured to select a plurality of typical frequency points in the frequency response curve data of the main MIC and the auxiliary MIC generated by the audio data test module 4041, and set a parameter correction variable Δ for the selected typical frequency points. That is, the main MIC gain and the gain of the secondary MIC are appropriately corrected.

FIG. 6 shows a diagram of frequency response curve data for a primary microphone and a secondary microphone, in accordance with an embodiment of the present disclosure. As shown in Fig. 6, the thick solid line indicates the frequency response curve of the main MIC (referred to as "frequency response curve"), the thin solid line indicates the frequency response curve before the auxiliary MIC correction, and the thin dotted line indicates the modified frequency response curve of the auxiliary MIC. .

The noise reduction parameter calibration module 4043 is configured to fit the various posture coordinate data collected by the acquisition unit 401 and the corresponding plurality of parameter correction variables Δ to generate a noise reduction parameter configuration file and store the same.

According to an embodiment of the present disclosure, the noise reduction processing unit 402 may be configured to: in the noise reduction parameter calibration database 403, look up posture coordinate data having the smallest difference from the posture coordinate data in the current posture information of the mobile terminal; The data and the corresponding fitting result of the plurality of parameter correction variables generate a parameter correction variable corresponding to the posture coordinate data in the current posture information of the mobile terminal; and obtain a parameter corresponding to the current posture information of the mobile terminal according to the generated parameter correction variable. Noise reduction parameter configuration file.

FIG. 7 shows a schematic diagram of a fit curve between a pose and a parameter correction variable in accordance with an embodiment of the present disclosure.

As shown in FIG. 7, for example, when the noise reduction parameter calibration database is established, the posture 1 to posture 6 of the mobile terminal are fitted with the parameter correction variable 1 to the parameter correction variable 6. Therefore, when the current posture information of the mobile terminal collected by the collecting unit 401 includes the coordinate data corresponding to the posture 7, the noise reduction processing unit 402 can find the minimum difference between the attitude coordinate data of the posture 7 and the posture coordinate data of the posture 7 in the noise reduction parameter calibration database 403. The attitude coordinate data, for example, finds that the posture coordinate data corresponding to the posture 3 and the posture 4 and the posture coordinate data corresponding to the posture 7 are closest to each other, and therefore, the posture 3 and the posture 4 and the parameter correction variable 3 and the parameter correction variable can be utilized. The fitting result of 4 obtains the parameter correction variable Δ corresponding to the posture 7, for example, the parameter corresponding to the posture 7 can be obtained by the local linearization calculation of the parameter correction variable 3 and the parameter correction variable 4 corresponding to the posture 3 and the posture 4. Correct the variable Δ. Based on the same principle, the parameter correction variable Δ corresponding to the posture 8 can be obtained by the fitting result of the posture 5 and the posture 6 with the parameter correction variable 5 and the parameter correction variable 6.

Table 2 shows various indicator parameters for noise reduction processing using different noise reduction parameters for different handheld gestures in accordance with an embodiment of the present disclosure.

Table 2

It can be seen from Table 2 that in the case of using different noise reduction parameters for different postures for noise reduction processing, the call loudness and noise reduction effect in each posture can satisfy the requirements.

As shown in FIG. 8, the noise reduction method of the terminal according to an embodiment of the present disclosure may include steps S801 to S809.

In step S801, the attitude coordinate data of the mobile phone is acquired by an azimuth sensor such as an acceleration sensor or a gyroscope.

In step S802, the audio data of the primary MIC and the secondary MIC are tested under different handset gestures (ie, different attitude coordinate data) by the audio lab environment to generate frequency response curve data of the primary MIC and the secondary MIC.

In step S803, a plurality of typical frequency points are selected in the frequency response curve data of the main MIC and the auxiliary MIC, and parameter correction variables are respectively set for each of the typical frequency points.

At step S804: a noise reduction parameter profile is generated by fitting the attitude coordinate data and the frequency response curve data and stored in the noise reduction parameter calibration database.

In step S805, it is judged whether or not the hand-held call mode is entered. If YES, the process proceeds to step S806, otherwise returns.

In step S806, it is judged whether the hand posture is within a reasonable range. If YES, step S807 is continued, otherwise the mobile phone is considered to be out of the hand, and the flow is ended.

It will be appreciated that the "reasonable range" referred to in step S806 includes standard hand held gestures and various non-standard hand held gestures. When it is detected that the reasonable range is exceeded, it indicates that the mobile phone has been removed from the handheld, for example, in a hands-free conversation state.

In step S807, the current posture coordinate data of the mobile phone is acquired.

In step S808, the corresponding noise reduction parameter configuration file is called according to the collected posture coordinate data, and the frequency response curve data of the main MIC and the auxiliary MIC are corrected at each frequency point.

In step S809, the noise reduction algorithm is performed using the frequency response curve data of the optimized main MIC and the auxiliary MIC to achieve noise reduction.

Embodiments of the present disclosure also provide a computer storage medium having stored thereon one or more programs, the one or more programs being executable by one or more processors such that the one or more processors are implemented A noise reduction method of a mobile terminal according to the above embodiments of the present disclosure.

According to the solution of the present disclosure, the gesture information of the mobile phone in different handheld gestures can be collected and different noise reduction parameter profiles can be generated in advance for calling in the hand-held call, thereby achieving the voice volume in a non-standard posture. Noise reduction.

A person skilled in the art can understand that all or part of the steps in the above embodiments of the present disclosure may be completed by a program to instruct related hardware, and the program may be stored in a computer readable storage medium such as a read only memory, a magnetic disk or an optical disk. . Furthermore, all or part of the steps of the above embodiments may also be implemented using one or more integrated circuits. Correspondingly, each module/unit in each of the above embodiments may be implemented in the form of hardware or in the form of a software function module. This application is not limited to any specific combination of hardware and software.

The above description is only an embodiment of the present disclosure, and is not intended to limit the scope of the disclosure. Any modifications, equivalents, improvements, etc. made within the spirit and principles of the present disclosure are intended to be included within the scope of the present disclosure.

Claims

A noise reduction method for a mobile terminal, comprising:

When the mobile terminal enters the handheld call mode, collecting current gesture information of the mobile terminal;

Recalling, from a pre-established noise reduction parameter calibration database, a noise reduction parameter configuration file corresponding to the current posture information of the mobile terminal;

Audio correction and noise reduction processing is performed on the dual microphone of the mobile terminal according to the called noise reduction parameter profile.
The noise reduction method according to claim 1, wherein the noise reduction parameter calibration database stores at least a noise reduction parameter configuration file corresponding to posture information of the mobile terminal, the posture information including posture coordinate data of the mobile terminal, and The noise reduction parameter profile corresponding to the attitude information includes a plurality of parameter correction variables for audio correction and noise reduction processing in the corresponding posture.
The noise reduction method of claim 2, further comprising: pre-establishing the noise reduction parameter calibration database, comprising:

Collecting various posture coordinate data of the mobile terminal;

Testing audio data of the primary microphone and the secondary microphone of the mobile terminal under various attitude coordinate data to generate frequency response curve data of the primary microphone and the secondary microphone under different attitude coordinate data;

In the generated frequency response curve data of the main microphone and the auxiliary microphone, selecting a plurality of typical frequency points, and respectively setting parameter correction variables for the selected typical frequency points;

The various pose coordinate data and the corresponding plurality of parameter correction variables are fitted to generate a noise reduction parameter profile and stored.
The noise reduction method according to claim 3, wherein the step of invoking a noise reduction parameter profile corresponding to the current posture information of the mobile terminal from the pre-established noise reduction parameter calibration database comprises:

Locating, in the noise reduction parameter calibration database, attitude coordinate data having the smallest difference from the attitude coordinate data in the current posture information of the mobile terminal;

Generating a parameter correction variable corresponding to the posture coordinate data in the current posture information of the mobile terminal according to the searched posture coordinate data and the fitting result of the corresponding plurality of parameter correction variables;

And obtaining a noise reduction parameter configuration file corresponding to the current posture information of the mobile terminal according to the generated parameter correction variable.
The noise reduction method according to claim 2 or 3, wherein the plurality of posture coordinate data of the mobile terminal are acquired by an orientation sensor.
The noise reduction method of claim 5, wherein the orientation sensor comprises an acceleration sensor and a gyroscope.
A noise reduction device for a mobile terminal, comprising:

a collecting unit, when the mobile terminal enters the handheld call mode, the collecting unit collects current posture information of the mobile terminal;

a noise reduction processing unit configured to call a noise reduction parameter configuration file corresponding to the current posture information of the mobile terminal from a pre-established noise reduction parameter calibration database, and to perform the movement according to the called noise reduction parameter configuration file The terminal's dual microphone performs audio correction and noise reduction processing.
The noise reduction device according to claim 7, wherein the noise reduction parameter calibration database stores at least a noise reduction parameter profile corresponding to posture information of the mobile terminal, the posture information including posture coordinate data of the mobile terminal, and The noise reduction parameter profile corresponding to the attitude information includes a plurality of parameter correction variables for audio correction and noise reduction processing in the corresponding posture.
A noise reduction device according to claim 8, further comprising a noise reduction parameter calibration database generating unit including an audio data test module, an audio parameter correction module, and a noise reduction parameter calibration module, wherein

The audio data test module tests audio data of the primary microphone and the secondary microphone of the mobile terminal collected by the acquisition unit under various attitude coordinate data to generate frequency responses of the primary microphone and the secondary microphone under different attitude coordinate data. Curve data,

The audio parameter correction module is configured to select a plurality of typical frequency points in the frequency response curve data of the primary microphone and the secondary microphone generated by the audio data test module, and set parameter corrections for the selected typical frequency points respectively. Variable, and

The noise reduction parameter calibration module is configured to fit various posture coordinate data collected by the acquisition unit and corresponding plurality of parameter correction variables to generate a noise reduction parameter configuration file and store the noise reduction parameter.
The noise reduction device of claim 9, wherein the noise reduction processing unit is configured to:

Locating, in the noise reduction parameter calibration database, attitude coordinate data having the smallest difference from the attitude coordinate data in the current posture information of the mobile terminal;

Generating a parameter correction variable corresponding to the posture coordinate data in the current posture information of the mobile terminal according to the searched posture coordinate data and the fitting result of the corresponding plurality of parameter correction variables;

And obtaining a noise reduction parameter configuration file corresponding to the current posture information of the mobile terminal according to the generated parameter correction variable.
The noise reducer of claim 7 wherein said acquisition unit comprises an orientation sensor.
The noise reducer of claim 11 wherein said orientation sensor comprises an acceleration sensor and a gyroscope.
A computer storage medium having stored thereon one or more programs, the one or more processors being executed by one or more processors, the one or more processors performing as claimed in claims 1 to 6 A noise reduction method for a mobile terminal as described.