Sound source locating method and apparatus, and device and storage medium

Abstract

A sound source locating method and apparatus, and a device and a storage medium. The method comprises: acquiring an environmental sound pressure by means of at least two piezoelectric vibrators (71, 73, 75) (S101); on the basis of an adjustment coefficient of each piezoelectric vibrator (71, 73, 75), calibrating the environmental sound pressure that is acquired by each piezoelectric vibrator (71, 73, 75), so as to obtain a target sound pressure for each piezoelectric vibrator (71, 73, 75) (S102); converting the target sound pressure for each piezoelectric vibrator (71, 73, 75) into a distance between each piezoelectric vibrator (71, 73, 75) and a sound source by using the sound pressure as a function of the distance (S103); and determining the location of the sound source on the basis of the location of each piezoelectric vibrator (71, 73, 75) and the distance between each piezoelectric vibrator (71, 73, 75) and the sound source (S104).

Description

A sound source localization method, device, equipment and storage medium

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application is based on the Chinese patent application with application number 202310918080.8, application date July 24, 2023, and application name “A sound source localization method, device, and storage medium”, and claims the priority of the Chinese patent application. The entire content of the Chinese patent application is hereby introduced into this application as a reference. Technical Field

[0003] The present application relates to the field of sound recognition technology, and in particular to a sound source recognition method, apparatus, device and storage medium. Background Art

[0004] The automotive industry is currently experiencing a trend toward electrification, connectivity, sharing, and intelligence. This has created new technical demands for driving styles, user interaction experiences, and vehicle usage patterns. There is an urgent need to improve current vehicle systems to meet these demands. External microphones, as sensors, need to be placed outside the vehicle to support various application scenarios, such as external voice interaction, external voice control, external noise pickup for noise reduction, and internal and external voice interaction. Voice recognition technology converts sound signals into digital signals and analyzes them through a series of algorithms to identify the type and content of the sound. This technology includes speech recognition, speech synthesis, speech analysis, and speech conversion. Related technologies are limited by the vehicle's waterproof and dustproof requirements, limiting the location of external microphones. Sound collection locations are relatively limited, resulting in a low signal-to-noise ratio (SNR) and an inability to effectively capture sound signals from multiple locations to meet user experience requirements.

[0005] Summary of the Invention

[0006] One purpose of the present application is to provide a sound source identification method, the advantage of which is that the ambient sound pressure is obtained by at least two piezoelectric vibrators; the ambient sound pressure obtained by each piezoelectric vibrator is calibrated based on the adjustment coefficient of each piezoelectric vibrator to obtain the target sound pressure of each piezoelectric vibrator; the target sound pressure of each piezoelectric vibrator is converted into the distance between each piezoelectric vibrator and the sound source by utilizing the relationship between sound pressure and distance; and the position of the sound source is determined based on the position of each piezoelectric vibrator and the distance between each piezoelectric vibrator and the sound source.

[0007] Another object of the present application is to provide a sound source identification method, the advantage of which is that the ambient sound pressure obtained by each piezoelectric vibrator is adjusted by the adjustment coefficient of each piezoelectric vibrator, so that the adjusted ambient sound pressure has the same noise sound pressure, which makes it easier to perform denoising and obtain better denoising effect.

[0008] Another object of the present application is to provide a sound source identification method, which has the advantage of determining the adjustment coefficient of each piezoelectric vibrator by using the obtained standard noise sound pressure and the noise sound pressure obtained by each piezoelectric vibrator.

[0009] Another object of the present application is to provide a sound source identification method, the advantage of which is that by respectively mounting at least two piezoelectric vibrators on the diaphragms at different positions of the vehicle panel, the adjustment coefficient of the piezoelectric vibrator mounted on the corresponding diaphragm can be changed by adjusting the material of the diaphragm. At the same time, the distance between each two piezoelectric vibrators is greater than the preset distance to obtain a better positioning effect.

[0010] Another object of the present application is to provide a sound source identification method, the advantage of which is that for each of the piezoelectric vibrators, the change in sound pressure is determined based on preset sound pressure parameters and the adjusted ambient sound pressure, and then the distance between the piezoelectric vibrator and the sound source is obtained based on the change in sound pressure.

[0011] Another object of the present application is to provide a sound source identification method, the advantage of which is that a circle with the distance between each piezoelectric vibrator and the sound source as the radius is determined with each piezoelectric vibrator as the center, and the intersection of the circles corresponding to each piezoelectric vibrator is determined, and the intersection located outside the vehicle is the sound source position.

[0012] Another advantage of the present application is that it provides a sound source identification method, which performs denoising processing based on the adjusted ambient sound pressure of at least two piezoelectric vibrators to obtain corresponding voice data, performs voice recognition based on the voice data to obtain a recognition result, and interacts based on the voice recognition result and the position of the sound source to obtain a better interaction effect.

[0013] Another object of the present application is to provide a sound source identification method, the advantage of which is to perform differential processing on the adjusted ambient sound pressure of at least two piezoelectric vibrators to eliminate the common noise part and obtain relatively pure voice data.

[0014] Another object of the present application is to provide a sound source identification device, which has the advantage that the sound source identification device includes a piezoelectric array microphone, a control unit and a diaphragm, the piezoelectric array microphone is mounted on the diaphragm at different positions on the vehicle panel, and the control panel is used to implement the above-mentioned method steps, so as to receive sounds emitted from multiple directions and then determine the position information of sound sources located at different positions of the vehicle.

[0015] Another object of the present application is to provide an electronic device, which has the advantage that the electronic device includes a memory and a processor, the memory stores a computer program that can be run on the processor, and the processor implements the steps in the above method when executing the computer program.

[0016] Another object of the present application is to provide a computer-readable storage medium, which has the advantage of storing a computer program thereon, and when the computer program is executed by a processor, the steps in the above method are implemented.

[0017] To achieve the above-mentioned purpose, the technical solution of the embodiment of the present application is implemented as follows:

[0018] In one aspect, an embodiment of the present application provides a method for identifying a sound source, the method comprising:

[0019] Ambient sound pressure is obtained through at least two piezoelectric vibrators; the ambient sound pressure obtained by each piezoelectric vibrator is calibrated based on the adjustment coefficient of each piezoelectric vibrator to obtain the target sound pressure of each piezoelectric vibrator; the target sound pressure of each piezoelectric vibrator is converted into the distance between each piezoelectric vibrator and the sound source by utilizing the relationship between sound pressure and distance; the position of the sound source is determined based on the position of each piezoelectric vibrator and the distance between each piezoelectric vibrator and the sound source.

[0020] On the other hand, an embodiment of the present application provides a sound source identification device, including a piezoelectric array microphone, a control unit and a diaphragm, wherein the piezoelectric array microphone is mounted on the diaphragm at different positions on the vehicle panel, and the control panel is used to implement the steps of the above method.

[0021] On the other hand, an embodiment of the present application provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the steps in the above method when executing the computer program.

[0022] On the other hand, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, and the computer program implements the steps in the above method when executed by a processor.

[0023] It should be understood that the above general description and the following detailed description are merely exemplary and explanatory, and do not limit the technical solutions of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] FIG1A is a schematic top view of a device for identifying external sound sources of a vehicle provided in an embodiment of the present application;

[0025] FIG1B is a flow chart of a sound source identification method according to an embodiment of the present application;

[0026] FIG2 is a flow chart of a sound source identification method provided in an embodiment of the present application;

[0027] FIG3 is a flow chart of a sound source identification method provided in an embodiment of the present application;

[0028] FIG4 is a flow chart of a sound source identification method provided in an embodiment of the present application;

[0029] FIG5 is a flow chart of a sound source identification method provided in an embodiment of the present application;

[0030] FIG6 is a flow chart of a sound source identification method provided in an embodiment of the present application;

[0031] 7A-7C are schematic structural diagrams of a piezoelectric vibrator provided in an embodiment of the present application;

[0032] FIG8 is a schematic diagram of a signal input structure of a single piezoelectric microphone according to an embodiment of the present application;

[0033] FIG9 is a schematic structural diagram of a piezoelectric microphone vibrator provided in an embodiment of the present application;

[0034] FIG10 is a schematic diagram of the structure of a piezoelectric microphone vibrator array provided in an embodiment of the present application;

[0035] FIG11 is a schematic structural diagram of a piezoelectric microphone assembly method provided in an embodiment of the present application;

[0036] FIG12 is a schematic structural diagram of a piezoelectric microphone assembly method provided in an embodiment of the present application;

[0037] FIG13 is a schematic diagram of a noise elimination method provided in an embodiment of the present application;

[0038] FIG14 is a schematic diagram of a method for determining a sound source position provided in an embodiment of the present application;

[0039] FIG15 is a schematic diagram of a method for detecting the location of a sound source when the volume of the sound source changes, provided by an embodiment of the present application;

[0040] FIG16 is a schematic diagram of the structure of a sound source identification device provided in an embodiment of the present application;

[0041] FIG17 is a schematic diagram of a hardware entity of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions of this application are further elaborated in detail below with reference to the accompanying drawings and embodiments. The described embodiments should not be regarded as limiting this application. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0043] In the following description, references to "some embodiments" describe a subset of all possible embodiments. However, it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict. The terms "first / second / third" are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It is understood that the specific order or sequence of "first / second / third" may be interchanged where permitted, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0044] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing this application only and are not intended to limit this application.

[0045] The embodiments of the present application provide a method for localizing a sound source, and more particularly, a method for localizing a sound source using a vehicle-mounted piezoelectric microphone, which identifies the location of the sound source through the vehicle-mounted piezoelectric microphone and simultaneously recognizes the voice of the sound source, thereby supporting human-vehicle interaction. The method can be executed by a processor of a computer device. The computer device may refer to a server, a laptop, a tablet computer, a desktop computer, a smart TV, a set-top box, a mobile device (such as a mobile phone, a portable video player, a personal digital assistant, a dedicated messaging device, a portable gaming device), or other device with data processing capabilities.

[0046] In some embodiments, when a car is used as a vehicle capable of interacting with the outside world, microphones are required to be installed on different sides of the vehicle's exterior. Figure 1A is a top-down schematic diagram of vehicle external sound source identification in some embodiments of the present application. The location of the external sound source is determined using the external microphones. For example, microphones 1, 2, and 3 are installed around the vehicle. Microphones 1 and 2 can determine the location of sound source 1, while microphones 2 and 3 can determine the location of sound source 2.

[0047] FIG1B is a flow chart of a sound source identification method provided in an embodiment of the present application. As shown in FIG1B , the method includes the following steps S101 to S104:

[0048] S101: Acquire ambient sound pressure through at least two piezoelectric vibrators.

[0049] In an embodiment of the present application, at least two of the piezoelectric vibrators can be mounted on the diaphragm at different positions on the vehicle panel, such as the B-pillar or behind the front and rear bumpers of the vehicle. The specific installation position can be determined according to actual needs and application scenarios, and is not limited in the embodiment of the present application.

[0050] Based on the above embodiment, by installing the piezoelectric vibrator at multiple positions, it is possible to more conveniently obtain sound pressure signals from all directions, and is no longer limited to obtaining sound pressure from only fixed positions, thereby improving the user experience and the sound reception effect.

[0051] In some embodiments, when the piezoelectric vibrator is mounted behind a vehicle window, glass can be used as the diaphragm. When the piezoelectric vibrator is mounted on the vehicle's exterior, a plastic housing can be used as the diaphragm. The specific diaphragm type can vary depending on the material type of the piezoelectric vibrator's mounting location and is not limited here. For example, a dedicated diaphragm can be used to provide better sound reception and higher sensitivity, with the diaphragm's softness positively correlated with its sensitivity.

[0052] Based on the above embodiment, directly installing the microphone on the back of the vehicle structure can make minimal changes to the vehicle structure, ensuring the sealing of the vehicle body. Compared with traditional vehicle-mounted microphones, it has better dust and water resistance. The use of a dedicated diaphragm can further improve the microphone's sound reception effect and increase sensitivity. The diaphragm can be made of more durable automotive-grade materials, and the user experience is better than ordinary waterproof membranes.

[0053] In the embodiments of the present application, the distance between each two piezoelectric vibrators is greater than a preset distance. In some embodiments, the distance between each two piezoelectric vibrators is not less than 3 cm. Too small a distance between piezoelectric vibrators can lead to inaccurate sound source location recognition, so a minimum distance is set to ensure good recognition accuracy.

[0054] In an embodiment of the present application, the ambient sound pressure may be a collection of multiple sounds, such as human voices, bird calls, and the sound of cars driving. Of course, it may also include other types of sounds, depending on the actual application scenario, which is not limited in the embodiment of the present application.

[0055] In some embodiments, the piezoelectric vibrator can remain in a standby state, operating at low power consumption when no sound is detected, continuously monitoring the surrounding environment, and quickly responding and resuming normal operation when ambient sound pressure is detected. In some embodiments, when the piezoelectric vibrator detects an ambient sound pressure less than a preset sound pressure threshold, it can be treated as noise pressure that does not contain the user's voice and discarded; when the piezoelectric vibrator detects an ambient sound pressure greater than or equal to the preset sound pressure threshold, it can be treated as containing the user's voice and subsequently processed.

[0056] Based on the above embodiments, compared with ordinary MEMS (Micro-Electro-Mechanical System) microphones, the piezoelectric vibrator microphone used in this application does not require additional power support, and can achieve sound-to-electricity conversion only by relying on its own piezoelectric effect, and has a shorter response time. Therefore, the piezoelectric microphone can work in low-power mode for a long time for monitoring, realizing all-weather rapid response to user voice to determine the location of the sound source.

[0057] S102 : Calibrate the ambient sound pressure acquired by each piezoelectric vibrator based on the adjustment coefficient of each piezoelectric vibrator to obtain a target sound pressure of each piezoelectric vibrator.

[0058] In the embodiment of the present application, the adjustment coefficient of each piezoelectric vibrator is an adjustment parameter for calibrating the ambient sound pressure, which is calculated based on the standard noise parameter and the noise level obtained by the piezoelectric vibrator.

[0059] In an embodiment of the present application, after the ambient sound pressure is obtained by each of the piezoelectric vibrators, based on the method for obtaining the adjustment coefficient, when calibrating the ambient sound pressure, the adjustment coefficient can be multiplied or divided by the ambient sound pressure to obtain the calibrated ambient sound pressure, that is, the target sound pressure.

[0060] It should be noted that the calibration operation described above can be used to amplify, reduce, or maintain the acquired ambient sound pressure. Different panel materials are used in different locations on the vehicle, and the diaphragm material may need to be adjusted to match the panel material for different locations. For example, in some locations of the vehicle, a harder material may be used to ensure sealing and durability, and the diaphragm material here must also be adjusted to a harder material. A harder diaphragm will reduce the sensitivity of the microphone. Calibrating the acquired ambient sound pressure can adjust the sound pressure level of microphones with different sensitivities, facilitating subsequent noise cancellation and recognition processing.

[0061] S103 : Using the relationship between sound pressure and distance, convert the target sound pressure of each piezoelectric vibrator into the distance between each piezoelectric vibrator and the sound source.

[0062] In the embodiment of the present application, while the sound pressure level generated by the sound source remains constant, the sound pressure received by the piezoelectric vibrator decreases as the distance between the sound source and the piezoelectric vibrator increases. This means that the received ambient sound pressure is negatively correlated with the distance between the piezoelectric vibrator and the sound source. Based on this relationship, the distance between the sound source and each piezoelectric vibrator can be inferred from the target sound pressure. The relationship between sound attenuation and distance is relatively easy to measure and can be adjusted for different environments. For example, both high and low temperatures affect sound propagation. This allows for accurate sound attenuation and distance relationships to be determined for subsequent steps.

[0063] S104 : Determine the position of the sound source based on the position of each piezoelectric vibrator and the distance between each piezoelectric vibrator and the sound source.

[0064] In the embodiment of the present application, for example, the position of the sound source is determined by two piezoelectric vibrators. Based on the known positions of the piezoelectric vibrators and the distance between each piezoelectric vibrator and the sound source, the position of the sound source can be determined by constructing a geometric relationship.

[0065] For example, the position of the sound source can be determined by determining the triangle method. When the distance between the piezoelectric vibrators and the distance between the piezoelectric vibrators and the sound source are obtained, a unique triangle can be constructed with the two piezoelectric vibrators as the two vertices of the triangle. The position of the triangle vertices outside the vehicle is used as the position of the sound source to determine the exact position of the sound source.

[0066] In some embodiments, if the sound source location is determined using multiple piezoelectric vibrators, the calculation method described above can be used to select two piezoelectric vibrators at a time, resulting in a more accurate sound source location through multiple calculations. Furthermore, calculations can be performed using geometric relationships of other geometric figures, which are not listed here.

[0067] Based on the above embodiment, by using relatively basic geometric relationships to infer the sound source position, the complexity of the system logic operation can be greatly reduced and the operation speed can be improved. In addition, the error of the sound source position obtained by this method is not large, and the direction of the sound source can be basically identified.

[0068] FIG2 is an optional flow diagram of a sound source localization method provided in an embodiment of the present application, which can be executed by a processor of a computer device. Based on FIG1B , step S102 in FIG1B can be updated to S201 to S202. The method includes the following steps for explanation in conjunction with the steps shown in FIG2 :

[0069] S101: Acquire ambient sound pressure through at least two piezoelectric vibrators.

[0070] S201 : Adjusting the ambient sound pressure acquired by each piezoelectric vibrator based on an adjustment coefficient of each piezoelectric vibrator to obtain an adjusted ambient sound pressure; each adjusted ambient sound pressure includes the same noise sound pressure.

[0071] In some embodiments, a corresponding adjustment coefficient is individually set for each piezoelectric vibrator, and the adjustment coefficient is a parameter used to adjust the ambient sound pressure obtained by the piezoelectric vibrator.

[0072] In some embodiments, the adjustment coefficient of the piezoelectric vibrator depends on the softness of the diaphragm material at the installation site of the piezoelectric vibrator and the selected standard noise sound pressure. When other conditions remain unchanged, for example, taking a pair of piezoelectric vibrators with diaphragms of different materials as an example, it is assumed that the ambient sound pressure after adjustment = the ambient sound pressure before adjustment * adjustment coefficient. For the sake of convenience, it is assumed that the diaphragm of the first piezoelectric vibrator is harder and the diaphragm of the second piezoelectric vibrator is softer. If the noise sound pressure obtained by the first piezoelectric vibrator is taken as the standard noise sound pressure, that is, the adjustment coefficient of the first piezoelectric vibrator is 1, and the adjustment coefficient of the second piezoelectric vibrator must be less than 1 to adjust the noise sound compression obtained by the second piezoelectric vibrator to the same level as the standard noise sound pressure. consistent; if the noise sound pressure obtained by the second piezoelectric vibrator is taken as the standard noise sound pressure, that is, the adjustment coefficient of the second piezoelectric vibrator is 1, and at this time the adjustment coefficient of the first vibrator must be greater than 1, then the noise sound pressure obtained by the first piezoelectric vibrator can be adjusted to be consistent with the standard noise sound pressure; for another example, the selected standard noise sound pressure has changed, which may be due to the adjustment of the preset standard noise sound pressure value, or the reselection of the noise sound pressure of other piezoelectric vibrators as the standard noise sound pressure. At this time, the proportional relationship between the standard noise sound pressure and the noise sound pressure of each piezoelectric vibrator has changed, and the adjustment coefficient needs to be modified so that the noise sound pressure of each piezoelectric vibrator can still be consistent with the standard noise sound pressure.

[0073] In some embodiments, in order to ensure that the sound pressure obtained by each piezoelectric vibrator contains the same level of noise pressure, thereby facilitating noise reduction processing, the sound pressure obtained by each piezoelectric vibrator needs to be adjusted. The following uses a piezoelectric vibrator as the adjustment object and provides an example to illustrate this.

[0074] For example, if the noise pressure level contained in the ambient sound pressure obtained by the piezoelectric vibrator is greater than the standard noise pressure, the ambient sound pressure obtained by the piezoelectric vibrator needs to be reduced as a whole based on the adjustment coefficient so that the noise pressure level contained in the ambient sound pressure is the same as the standard noise pressure.

[0075] For example, if the noise sound pressure level contained in the ambient sound pressure obtained by the piezoelectric vibrator is lower than the standard noise sound pressure, the ambient sound pressure obtained by the piezoelectric vibrator needs to be amplified as a whole based on the adjustment coefficient so that the noise sound pressure level contained in the ambient sound pressure is the same as the standard noise sound pressure.

[0076] For example, if the ambient sound pressure acquired by the piezoelectric vibrator includes a noise sound pressure level that is equal to the standard noise sound pressure, there is no need to change the ambient sound pressure acquired by the piezoelectric vibrator.

[0077] In some embodiments, the ambient sound pressure obtained by each piezoelectric vibrator contains the same level of noise sound pressure, and the same amount of noise can be removed by directly performing differential processing to obtain a relatively pure voice sound pressure signal.

[0078] For example, taking the use of two piezoelectric vibrators as an example, the noise sound pressure in the ambient sound pressure obtained by the first piezoelectric vibrator is used as the standard noise sound pressure. At this time, there is no need to adjust the ambient sound pressure obtained by the first piezoelectric vibrator. Instead, for the sound pressure obtained by the second piezoelectric vibrator, its noise sound pressure is obtained and a ratio operation is performed with the standard noise sound pressure. The result of the operation is used as the adjustment coefficient of the second piezoelectric vibrator, and then the ambient sound pressure obtained by the second piezoelectric vibrator is adjusted based on the adjustment coefficient of the second piezoelectric vibrator. Among them, if the noise sound pressure of the ambient sound pressure obtained by the second piezoelectric vibrator is greater than the standard noise sound pressure, the adjustment is characterized by an overall reduction in the ambient sound pressure; if the noise sound pressure is less than the standard noise sound pressure, the adjustment is characterized by an overall increase in the ambient sound pressure; if the noise sound pressure is equal to the standard noise sound pressure, the adjustment is characterized by not changing the ambient sound pressure.

[0079] Based on the above embodiment, by adjusting the sound pressure of each environment, the common noise portion in the environment sound pressure can be eliminated, thereby achieving a better dynamic noise cancellation effect.

[0080] S202 : Using the adjusted ambient sound pressure of each piezoelectric vibrator as the target sound pressure of each piezoelectric vibrator.

[0081] In some embodiments, the target sound pressure is used as the sound pressure signal finally obtained by the piezoelectric vibrator, and in subsequent processing steps, operations are performed on the target sound pressure of each piezoelectric vibrator.

[0082] S103 : Using the relationship between sound pressure and distance, convert the target sound pressure of each piezoelectric vibrator into the distance between each piezoelectric vibrator and the sound source.

[0083] S104 : Determine the position of the sound source based on the position of each piezoelectric vibrator and the distance between each piezoelectric vibrator and the sound source.

[0084] Here, the above steps S101, S103 and S104 correspond to the steps S101, S103 and S104 in FIG. 1B , respectively. When implementing, reference may be made to the specific implementations of the above steps S101, S103 and S104.

[0085] In an embodiment of the present application, by adjusting the ambient sound pressure and determining the distance between each vibrator and the sound source based on the adjusted ambient sound pressure combined with the changing relationship between sound pressure and distance, the position of the sound source is determined. While ensuring the sound pressure effect, the sound pressure of the sound source in each azimuth position can be obtained and the position of the sound source can be determined.

[0086] FIG3 is a schematic diagram of an optional flow chart of a sound source localization method provided in an embodiment of the present application, which can be executed by a processor of a computer device. Based on FIG2 , steps S301 to S302 can be added after step S101 in FIG2 , and the steps shown in FIG3 are explained. The method includes the following steps:

[0087] S101: Acquire ambient sound pressure through at least two piezoelectric vibrators.

[0088] S301: Obtain a standard noise sound pressure and a noise sound pressure in each of the ambient sound pressures.

[0089] In an embodiment of the present application, the standard noise sound pressure can be a preset noise sound pressure, or it can be an ambient sound pressure selected from multiple acquired ambient sound pressures and the noise sound pressure in the ambient sound pressure is used as the standard noise sound pressure. The standard noise sound pressure can be set according to actual needs and application scenarios, and the embodiment of the present application does not limit it.

[0090] Based on the above embodiment, by flexibly selecting an appropriate standard noise sound pressure, the sound distortion generated in the subsequent ambient sound pressure calibration operation can be reduced, thereby ensuring the accuracy of subsequent sound recognition.

[0091] In the embodiment of the present application, the noise sound pressure in each of the ambient sound pressures can be obtained through an independent component analysis (ICA) algorithm or a deep learning method. Any method that can obtain the noise sound pressure from the ambient sound pressure can be used in the embodiment of the present application and is not limited here.

[0092] S302: Determine an adjustment coefficient of each piezoelectric vibrator based on the standard noise sound pressure and the noise sound pressure acquired by each piezoelectric vibrator.

[0093] In an embodiment of the present application, a ratio operation is performed on the standard noise sound pressure and the noise sound pressure of each of the piezoelectric vibrators, and the obtained calculation result is used as the adjustment coefficient of each of the piezoelectric vibrators. The order of the ratio operation is not limited here, and is only used to obtain the size and proportional relationship between the noise sound pressure of the piezoelectric vibrator and the standard noise sound pressure, and to obtain the adjustment coefficient and calibrate the ambient sound pressure. Based on different operation orders, the operation method can be changed accordingly when the ambient sound pressure is subsequently adjusted based on the adjustment coefficient. For example, the adjustment coefficient = piezoelectric vibrator noise sound pressure / standard noise sound pressure, then the adjusted ambient sound pressure = the ambient sound pressure before adjustment / adjustment coefficient.

[0094] Piezoelectric vibrators are highly sensitive, and even tiny differences in vibration can produce large results. Therefore, by calculating the adjustment coefficient for each piezoelectric vibrator separately rather than using a unified adjustment coefficient for all piezoelectric vibrators, it is possible to make differentiated adjustments based on the tiny differences in each piezoelectric vibrator to obtain accurate sound pressure data, providing support for subsequent sound source location identification and speech recognition.

[0095] Based on the above embodiment, by calculating the adjustment coefficient for each piezoelectric vibrator separately, the acquired ambient sound pressure can be accurately adjusted, and the calculation method of the adjustment coefficient does not involve complex algorithms, and will not cause a large burden on the system when a large number of ambient sound pressures are involved at the same time. The adjustment coefficient can be obtained more quickly, and then the ambient sound pressure can be adjusted.

[0096] S201 : Adjusting the ambient sound pressure acquired by each piezoelectric vibrator based on an adjustment coefficient of each piezoelectric vibrator to obtain an adjusted ambient sound pressure; each adjusted ambient sound pressure includes the same noise sound pressure.

[0097] S202 : Using the adjusted ambient sound pressure of each piezoelectric vibrator as the target sound pressure of each piezoelectric vibrator.

[0098] S103 : Using the relationship between sound pressure and distance, convert the target sound pressure of each piezoelectric vibrator into the distance between each piezoelectric vibrator and the sound source.

[0099] S104 : Determine the position of the sound source based on the position of each piezoelectric vibrator and the distance between each piezoelectric vibrator and the sound source.

[0100] Here, the above-mentioned steps S101, S201, S202, S103 and S104 correspond to the above-mentioned steps S101, S201, S202, S103 and S104 in Figure 2 respectively. When implementing, the specific implementation methods of the above-mentioned steps S101, S201, S202, S103 and S104 can be referred to.

[0101] In the embodiment of the present application, the adjustment coefficient of each piezoelectric vibrator is calculated separately, which ensures the accuracy of the target sound pressure obtained, thereby improving the accuracy of subsequent sound source positioning. When a piezoelectric vibrator accessory in the vehicle is subsequently changed, it is only necessary to recalculate the adjustment coefficient of the piezoelectric vibrator separately, which improves the flexibility in the implementation of the solution.

[0102] FIG4 is a schematic diagram of an optional flow chart of a sound source localization method provided in an embodiment of the present application, which can be executed by a processor of a computer device. Based on FIG1B , step S103 in FIG1B can be updated to S401 to S402. The method includes the following steps for explanation in conjunction with the steps shown in FIG4 :

[0103] S101: Acquire ambient sound pressure through at least two piezoelectric vibrators.

[0104] S102 : Calibrate the ambient sound pressure acquired by each piezoelectric vibrator based on the adjustment coefficient of each piezoelectric vibrator to obtain a target sound pressure of each piezoelectric vibrator.

[0105] S401 : For each of the piezoelectric vibrators, determine a sound pressure variation based on a preset sound pressure parameter and the target sound pressure.

[0106] In the embodiment of the present application, the preset sound pressure parameter is the sound pressure at a preset distance from the sound source. For example, when the preset distance is 1 meter, the preset sound pressure parameter may be 66 dB.

[0107] In some embodiments, the sound pressure variation is the difference between a preset sound pressure parameter and a target sound pressure. The positive or negative value of the sound pressure variation can characterize whether the distance between the current sound source and the piezoelectric vibrator is within the preset distance or outside the preset distance. For example, taking the target sound pressure obtained by the piezoelectric vibrator as 60db as an example, the sound pressure variation at this time is determined to be 6db, indicating that the sound pressure is attenuated by 6db at this time, that is, the distance between the sound source and the piezoelectric vibrator is greater than 1m. For example, taking the sound pressure obtained by the vibrator as 71db as an example, the sound pressure variation at this time is determined to be -5db, indicating that the distance between the sound source and the piezoelectric vibrator is less than 1m.

[0108] In some embodiments, in addition to setting the preset distance to 1 meter, you can also select the sound pressure value obtained at other distances as the preset sound pressure parameter, and use the corresponding distance as the preset distance. The same sound source at the same location will obtain different changes based on different preset sound pressure parameters and preset distances, but this does not affect the subsequent determination of the sound source location. The volume of sound produced by different people varies slightly, but it is still within the normal range.

[0109] S402: Determine the distance between the piezoelectric vibrator and the sound source based on the sound pressure variation.

[0110] For example, the calculation is continued with the data given in the example of step S401. First, the distance between the sound source and the piezoelectric vibrator is calculated when the attenuation is 6 dB. That is, the distance between the sound source and the piezoelectric vibrator is 1.99m at this time.

[0111] For example, the calculation is continued with the data given in the example of step S401, and the calculation is performed for the case where the attenuation is -5db. At this time, the distance between the sound source and the piezoelectric vibrator is That is, the distance between the sound source and the piezoelectric vibrator is 0.56m at this time.

[0112] S104 : Determine the position of the sound source based on the position of each piezoelectric vibrator and the distance between each piezoelectric vibrator and the sound source.

[0113] Here, the above steps S101, S102 and S104 correspond to the steps S101, S102 and S104 in FIG. 1B , respectively. When implementing, reference may be made to the specific implementations of the above steps S101, S102 and S104.

[0114] In the embodiment of the present application, calculations are performed based on variable preset parameters, and different calculation parameters can be set for different environments. For example, for a relatively small space, the corresponding preset parameters should also be reduced accordingly to obtain a better positioning effect. In this way, the preset parameters can be flexibly set to meet the application needs of various environments without consuming additional computing power.

[0115] FIG5 is a schematic diagram of an optional flow chart of a sound source localization method provided in an embodiment of the present application, which can be executed by a processor of a computer device. Based on FIG1B , step S104 in FIG1B can be updated to S501 to S502. The method includes the following steps for explanation in conjunction with the steps shown in FIG5 :

[0116] S101: Acquire ambient sound pressure through at least two piezoelectric vibrators.

[0117] S102 : Calibrate the ambient sound pressure acquired by each piezoelectric vibrator based on the adjustment coefficient of each piezoelectric vibrator to obtain a target sound pressure of each piezoelectric vibrator.

[0118] S103 : Using the relationship between sound pressure and distance, convert the target sound pressure of each piezoelectric vibrator into the distance between each piezoelectric vibrator and the sound source.

[0119] S501 : Determine a circle corresponding to each piezoelectric vibrator, with the position of each piezoelectric vibrator as the center and the distance between each piezoelectric vibrator and the sound source as the radius.

[0120] In some embodiments, a method for determining the location of a sound source based on the characteristics of a circle is provided. When the distance between a piezoelectric vibrator and the sound source is known, a circle can be drawn with the piezoelectric vibrator as the center and the distance from the sound source as the radius, and the sound source can be determined to be on the circle. When two piezoelectric vibrators are used for the above operation, two intersection points can be determined, and when the number of piezoelectric vibrators reaches three or more, a single intersection point can be determined. Using three or more piezoelectric vibrators can achieve a more accurate positioning effect, while using two piezoelectric vibrators can reduce costs while achieving a more accurate positioning effect. In actual conditions, the number of piezoelectric vibrators used can be flexibly selected according to needs.

[0121] Based on the above example, the position of the sound source can be quickly obtained when the position of the piezoelectric vibrator and the distance to the sound source are known. In some environments where low accuracy is required, only two piezoelectric vibrators can be used for calculation, which can further improve the speed.

[0122] S502: Determine the position of the intersection of the circles corresponding to each of the piezoelectric vibrators, which is located outside the vehicle, as the sound source position.

[0123] In the embodiment of the present application, when a person is interacting with the vehicle from outside, determining the sound source using two piezoelectric transducers yields two possible sound source locations. In this scenario, the sound source location can be determined based on the scenario, eliminating the sound source inside the vehicle and using the sound source outside the vehicle as the final location. The vehicle itself makes it easier to determine whether a sound source is inside or outside the vehicle. This allows the use of only two piezoelectric transducers to eliminate the possibility of a sound source inside the vehicle and quickly and accurately determine the actual sound source location.

[0124] Here, the above steps S101, S102, and S103 correspond to the steps S101, S102, and S103 in FIG. 1B , respectively. When implementing, reference may be made to the specific implementations of the above steps S101, S102, and S103.

[0125] In an embodiment of the present application, the number of piezoelectric vibrators enabled for each positioning can be selected based on the positioning accuracy requirements, and a distinction can be made between the sounds inside and outside the vehicle to prevent misidentification of the voices of people inside the vehicle, misjudgment of the direction, or misoperation, or even danger.

[0126] FIG6 is a schematic diagram of an optional flow chart of a sound source localization method provided in an embodiment of the present application, which can be executed by a processor of a computer device. Based on FIG1B , steps S601-S603 can be added after step S104 in FIG1B . The method includes the following steps for explanation in conjunction with the steps shown in FIG6 :

[0127] S101: Acquire ambient sound pressure through at least two piezoelectric vibrators.

[0128] S102 : Calibrate the ambient sound pressure acquired by each piezoelectric vibrator based on the adjustment coefficient of each piezoelectric vibrator to obtain a target sound pressure of each piezoelectric vibrator.

[0129] S103 : Using the relationship between sound pressure and distance, convert the target sound pressure of each piezoelectric vibrator into the distance between each piezoelectric vibrator and the sound source.

[0130] S104 : Determine the position of the sound source based on the position of each piezoelectric vibrator and the distance between each piezoelectric vibrator and the sound source.

[0131] S601: Perform denoising processing based on target sound pressures of at least two piezoelectric vibrators to obtain speech data corresponding to the sound source.

[0132] In some embodiments, a spectrum analysis method may be used to analyze the sound pressure signal to obtain the frequency band where the noise is located, and then a filter of the corresponding frequency band may be set to filter the sound pressure signal to remove the noise.

[0133] In some embodiments, differential processing may be used for denoising. In which, step S601 includes:

[0134] Step S6011: Perform differential processing on the target sound pressures of at least two of the piezoelectric vibrators to obtain voice data corresponding to the sound source.

[0135] In an embodiment of the present application, after obtaining the target sound pressures of at least two piezoelectric vibrators, the target sound pressures are differentially processed to eliminate the common noise portion in each target sound pressure, and the difference signal is retained as the input of the human voice, and the retained human voice input is used as the voice data corresponding to the sound source.

[0136] Based on the above embodiment, by eliminating the common noise portion in the target sound pressure through differential processing, a relatively pure voice signal can be obtained, so that the system has a better recognition effect on the user's instructions to provide corresponding services or perform specified operations. Differential processing is a currently mature processing method and does not require the additional development of new noise reduction methods, thereby reducing development costs.

[0137] S602: Perform speech recognition based on the speech data to obtain a recognition result.

[0138] In an embodiment of the present application, after acquiring the voice data, the voice data is recognized to determine the semantics and other information of the voice data. Signal processing technology, acoustic feature extraction technology, acoustic model processing technology, voice model technology and other channels can be used to recognize the voice data. A suitable recognition method can be selected according to actual needs. As long as the method can obtain the result of voice recognition, it can be used in this embodiment and is not limited here.

[0139] In some embodiments, the accuracy and safety of vehicle voice recognition can be improved by pre-recording voice models of the car owner and / or user-preset persons. Unrecorded human voices can be marked as having failed voice recognition in the recognition results, while recorded human voices, such as the car owner's voice, can pass voice recognition.

[0140] Based on the above example, by obtaining voice data for recognition, it is possible to meet the user's usage needs while not responding to strangers' voices, preventing the vehicle from being stolen or dangerous by others. It only responds to pre-stored human voices, ensuring the safety of vehicle use and improving the user experience in many aspects.

[0141] S603: Interact based on the recognition result and the position of the sound source.

[0142] In an embodiment of the present application, after obtaining the voice recognition results, the vehicle can take corresponding actions based on the voice recognition results. For example, when the user says "open the trunk", the vehicle can open the trunk after recognizing the voice data to interact with the user. For example, when the user says "open the car door", the vehicle can open the door closest to the user based on the recognized position of the person.

[0143] In some embodiments, after recognizing the user's voice, the user is also supported to interact with the in-vehicle artificial intelligence such as ChatGPT, for example, to query weather information, road information, etc. through voice.

[0144] Based on the above embodiments, by combining interaction with the user's location, the use of the car is made intelligent, which can greatly improve the user experience in some cases where manual operation is inconvenient for the user. In addition, combined with artificial intelligence, it can have a better recognition effect on the instructions given by the user, and can meet the user's more diverse voice commands during the interaction.

[0145] Here, the above steps S101, S102, S103 and S104 correspond to steps S101, S102, S103 and S104 in FIG. 1B , respectively. When implementing, reference may be made to the specific implementations of the above steps S101, S102, S103 and S104.

[0146] In the embodiment of the present application, not only is the location of the user when the voice is uttered determined, but the user's voice is also recognized, and then interaction is performed based on the user's location, thereby realizing intelligent vehicle control.

[0147] The following describes an exemplary application of the embodiments of the present application in a practical application scenario.

[0148] With the development of intelligent connected vehicles, there is an increasing need for vehicles to be able to interact with external devices through intelligent voice communication. External microphones, as sensors, need to be installed outside the vehicle to support various application scenarios, such as external voice communication, external voice control, external noise pickup for noise reduction, and internal and external voice communication. However, the water and dust resistance ratings of current in-vehicle microphones are very low, rarely meeting IP67 requirements (i.e., lacking high-level water and dust resistance). While there are currently no standard requirements for external microphone installation, different manufacturers often require at least IP67. IP67 refers to a device with level 6 dustproofing and level 7 water-resistant ratings. IP67-rated devices provide protection against dust ingress (comprehensive protection against contact and dust penetration) and short-term liquid immersion, with protection against immersion in water from 15 cm to 1 meter for up to 30 minutes. In actual use, various extreme conditions need to be considered. For example, during ordinary rain, IP67 generally meets the microphone's water resistance requirements. However, during the car wash process, IP67-rated microphone equipment cannot withstand the impact of high-pressure water guns.

[0149] IP68 is actually the highest rating for both dust and water resistance. The "8" in IP68 refers to its ability to withstand submersion in water and sustained use for a certain depth. However, there are no specific requirements for this depth or duration, generally determined by the manufacturer. For example, a device can withstand immersion in water up to 1.5 meters for half an hour without damage, and those exceeding this standard are still designated IP68. Some IP68 products even require normal use at a depth of 100 meters for at least 10 hours. Therefore, there is no upper limit to IP68 waterproofing. Therefore, defining IP68 is meaningless, but IP69K does have clear regulations. IP69K refers to its ability to withstand the hot steam wash test defined in EN60529 and DIN40050-9. IP69K means that while completely protected from dust intrusion, it can also withstand high-pressure, high-temperature water / steam cleaning and remain operational. It can provide protection against water pressure of 100 bar (1450 psi) and temperature of 80°C. The pressure is applied directly to the equipment at a 30-degree angle (0, 30, 60 and 90 degrees) for 30 seconds at each angle for a total of two minutes to prevent water ingress.

[0150] However, IP69K support does not guarantee IP68 support. The primary significance of IP69K is high-temperature protection (for example, in northern winters, high-pressure water jets can be hot water) and high-pressure impact resistance (not long-term immersion). For microphones installed outside a vehicle, if high-pressure water jets are a concern, they must meet both IP67 and IP69K requirements.

[0151] The optimal water pressure for car washing is generally between 40 bar and 80 bar. Using higher water pressure can damage the car's paint and surface. Most IP67-rated waterproof structures utilize waterproof sealing foam or adhesive sealants (for sealing between housings). However, these sealants prevent the sound from electroacoustic components from passing through the housing after the electronics are sealed. Using waterproof membranes (waterproof breathable membranes / waterproof acoustic membranes) to achieve waterproof and breathable properties requires openings in the housing for both air and sound transmission. Currently available waterproof membranes for electronic products provide dust and water resistance, ventilation and pressure balance, moisture-proof ventilation, and heat dissipation. However, waterproof membranes alone are insufficient. Automotive-grade applications, particularly those on the exterior, are not suitable for waterproof membranes alone. Even IP69K-rated materials cannot guarantee aging characteristics. Therefore, the outdoor microphones currently on the market are basically installed inside the exterior rearview mirror and equipped with a waterproof membrane, which can prevent dust while also avoiding the impact of rain and water guns.

[0152] When a car is used as a vehicle that can interact with the outside world, microphones need to be installed on different sides of the outside of the vehicle. Considering the waterproof level requirements, there is currently no suitable installation location for the microphone. The usual way is to install the microphone inside the exterior rearview mirror. The problem with this is that the direction of the outside person's speech should be close to the direction of the exterior rearview mirror. In addition, when the exterior rearview mirror is replaced with a smaller exterior rearview camera, it is impossible to install an exterior microphone. At the same time, other locations of the vehicle, such as the front and rear, still need to install exterior microphones to support voice input from different directions outside the vehicle. In order to solve the above problems, the embodiment of the present application provides a microphone that does not require a pickup hole but can still pick up sounds outside the vehicle, thereby solving problems such as waterproofing, dustproofing, and corrosion resistance, and at the same time determining the position of the speaker.

[0153] In the embodiment of the present application, the microphone utilizes the piezoelectric properties of a piezoelectric vibrator. When there is air vibration outside the vehicle, the piezoelectric vibrator senses the vibration caused by the diaphragm through its back diaphragm and converts the vibration signal into an electrical signal, which is the input microphone audio signal. Among them, the array of multiple piezoelectric vibrators can have the following characteristics: (1) the sampling sound frequency of the piezoelectric microphone is wider (compared to traditional microphones); (2) better resistance to environmental noise level (based on the ability of the microphone array); (3) waterproof, dustproof, and high temperature resistant; (4) lower cost (multi-microphone array can still use one electronic control unit ECU) and better scalability; (5) no need for a complex microphone mounting bracket; (6) support sound positioning; (7) has a sound wake-up function (can work in low power mode), which can be woken up by external sound. When the external sound pressure exceeds the threshold, the microphone component is actively woken up, and the microphone component can wake up the vehicle ECU to work.

[0154] In some embodiments, the exterior microphone can be installed within the interior structure of the vehicle's exterior panel, in a suitable location to support exterior voice interaction. For example, on the side of the vehicle, this can be done within the plastic backing of the B-pillar or C-pillar, or in a suitable location on the rear of the vehicle, such as the tailgate. These locations can be equipped with diaphragms made of similar materials and structures. Because the diaphragms face directly outward, they can achieve waterproof and dustproof properties while maintaining a similar appearance.

[0155] In some embodiments, the installation methods include: (1) mounting on the rear of the diaphragm, the structural assembly of the diaphragm and the piezoelectric vibrator is mounted outside the vehicle and seamlessly integrated with the external panel. The diaphragm is made of special materials, and its color and appearance are consistent with the surrounding materials, but the diaphragm material has better receiving sensitivity and can better receive the sound outside the vehicle. This diaphragm material is usually not a metal material, and a softer material would be more appropriate. (2) mounting on the rear of the car window glass, mounting the piezoelectric vibrator on the back of the car window, so that the external sound will vibrate the glass, and the glass will transmit the vibration to the piezoelectric vibrator. The piezoelectric vibrator converts the pressure vibration into an electrical signal, and the piezoelectric vibrator microphone module can analyze the piezoelectric signal into a sound audio signal. Usually, due to the thickness of the glass, its signal receiving sensitivity is not high enough, so the final sound signal-to-noise ratio is not good enough, which affects the reception effect. Therefore, verification and calibration are required, and the signal-to-noise ratio is improved through the microphone array. However, since the signal-to-noise ratio of the environment outside the car is relatively poor, the voltage obtained by the piezoelectric vibrator will be low when passing through the glass-type diaphragm with low sensitivity. (3) Mounted behind the exterior plastic panel: The piezoelectric vibrator can be mounted directly behind the B-pillar or the front and rear bumpers. When encountering panels made of harder materials, the signal reception sensitivity will be insufficient, resulting in a poor signal-to-noise ratio, which affects the reception quality. Therefore, verification and calibration are required, and the device must be mounted in the appropriate location to improve the signal-to-noise ratio through a microphone array.

[0156] As shown in Figures 7A-7C, the voltage generated by the piezoelectric vibrator changes accordingly with its deformation. The piezoelectric vibrator is highly sensitive and can sense very slight vibration pressure. By converting the vibration pressure into an electrical signal, the captured vibration signal can be converted into an electrical sound signal. In Figure 7A, voltmeter 72 detects a reverse voltage when piezoelectric vibrator 71 is stretched. Figure 7B shows that piezoelectric vibrator 73 is not deformed, at which point voltmeter 74 indicates no voltage is generated. Figure 7C shows that piezoelectric vibrator 75 is compressed, at which point voltmeter 76 detects a forward voltage.

[0157] Among them, the piezoelectric vibrator relies on the piezoelectric effect to convert sound into electric energy. When the piezoelectric ceramic sheet is subjected to external pressure, the ceramic sheet undergoes mechanical changes, and its polarization intensity decreases accordingly, causing a portion of the charge attached to the surface of the ceramic sheet to be released, resulting in a discharge phenomenon. When the pressure is removed, it returns to its original state, the polarization intensity increases, and a portion of the charge is adsorbed on the electrode again, resulting in a charging phenomenon. This phenomenon of converting mechanical energy into electrical energy is called the "positive piezoelectric effect." Positive piezoelectric effect: When subjected to an external force, the polarization of the piezoelectric body changes and releases corresponding electrical energy, and the released electrical energy is proportional to the applied force. Therefore, after selecting the piezoelectric vibrator, the piezoelectric vibrator is mounted on the back of the external diaphragm (the hard protective film of the panel). As shown in Figures 8 to 10, Figure 8 is a schematic diagram of the signal input structure of a single piezoelectric microphone, where 81 is the positive voltage output and 82 is the negative voltage output; Figure 9 is a schematic diagram of the structure of a piezoelectric microphone vibrator assembly, where only one piezoelectric microphone vibrator 92 is installed on the diaphragm panel 91; Figure 10 is a schematic diagram of the structure of a piezoelectric microphone vibrator array, where two piezoelectric microphone vibrators 94 are installed on the same diaphragm panel 93.

[0158] In some embodiments, the function of the piezoelectric microphone array back assembly is to parse the microphone's analog voltage signal into an audio signal and output it to other ECU modules in the vehicle via I2S (Inter IC Sound bus, also known as integrated circuit built-in audio bus).

[0159] As shown in Figures 11-12, Figure 11 is a structural schematic diagram of a piezoelectric microphone combination method, where the piezoelectric microphone array diaphragm panel 1101 is arranged on the vehicle panel hard protective film 1103 in the vehicle exterior panel 1102 (such as the vehicle exterior B-pillar panel), and the wiring harness interface 1104 is directly connected to the piezoelectric microphone array diaphragm panel 1101. At this time, a backplate assembly is not set for each piezoelectric vibration microphone; Figure 12 is a structural schematic diagram of another piezoelectric microphone combination method, where the piezoelectric microphone array diaphragm panel 1201 is arranged on the vehicle panel hard protective film 1203 in the vehicle exterior panel (such as the vehicle exterior B-pillar panel) 1202, the backplate assembly 1205 is arranged on the piezoelectric microphone array diaphragm panel 1201, and the wiring harness interface 1204 is connected to the backplate assembly 1205. At this time, a backplate assembly is set for each piezoelectric microphone.

[0160] In some embodiments, not every piezoelectric microphone assembly requires a backplate assembly. The backplate assembly can be integrated into other ECU modules, or all piezoelectric microphone assemblies can share one backplate assembly. In this way, the analog signal on each piezoelectric microphone assembly will be connected to other ECU modules or one of the backplate assemblies.

[0161] As shown in FIG13 , FIG13 is a schematic diagram of a noise elimination method. FIG13 takes two piezoelectric microphones as an example to illustrate how to adjust the received ambient sound pressure, which will be described below in conjunction with an embodiment.

[0162] In some embodiments, the sound pressure signal can be processed by the following method. The principle of piezoelectric microphone array signal processing is to perform differential processing on a single microphone signal, obtain the difference signal as the input of the human voice, and eliminate other signals as external vehicle noise. The premise is that the horizontal distance between each microphone needs to be far enough, for example, not less than 3 cm. When the diaphragm materials used by microphone 1 and microphone 2 are different, the sound pressure received at the microphone end for the same external sound pressure will be different. The softer the diaphragm material, the higher the sensitivity of the microphone, that is, the higher the sound pressure received. The harder the diaphragm material, the lower the sensitivity of the microphone, that is, the lower the sound pressure received. Therefore, the sound pressure level of the microphone can be adjusted at this time, such as increasing or decreasing the sound pressure Z received by microphone 2. When the noise sound pressure level X of microphone 2 is 噪 The noise sound pressure level Y of microphone 1 噪 When the adjustment coefficient of microphone 2 is consistent, a=Y 噪 / Z 噪 , then the modulated sound pressure of microphone 2 is X=a*Z. Through differential processing, the pressure difference XY between the two microphones can be obtained. This pressure difference is the voice pressure difference D of the person outside the car, that is, a*ZY, where. The voice sound pressure Y received by microphone 1 is the sound pressure received by the person outside the car when speaking at a specific position outside the car. The voice sound pressure X received by microphone 2 is the sound pressure received by the person outside the car when speaking at this same position, that is, X=Y+D=a*Z. The above method can improve the microphone receiving sensitivity. The main method is to use the modulated sound pressure amplitude to eliminate common noise and analyze the human voice sound pressure input of different microphones. The advantage of this method is that it has better dynamic characteristics and can perform better dynamic noise cancellation when there is a large range of random noise outside.

[0163] The above-described method can improve microphone reception sensitivity. The main approach utilizes modulated sound pressure amplitude to eliminate common noise and parse the vocal pressure input from different microphones. This method offers the advantage of better dynamic characteristics, enabling effective dynamic noise cancellation when subjected to a wide range of random external noise.

[0164] As shown in FIG14 , FIG14 is a schematic diagram of a method for determining the location of a sound source, and the specific method will be described below in conjunction with an embodiment.

[0165] In some embodiments, the position of the sound source can be determined by the following method: when the microphone 1401 and the microphone 1402 provided in the vehicle panel 1400 are in the same environment, it can be assumed that the noise environment is basically the same, so only the change in the position of the speaker will cause the sound pressure to be different on different microphones. The louder the voice pressure of the person outside the vehicle, the greater the sound pressure received by the microphone. When the person outside the vehicle speaks at the same sound pressure level, the closer to the vehicle, the closer the sound pressure received by the microphone. The position of the person outside the vehicle can be located by using the voice sound pressure intensity of the person outside the vehicle (using the intersection of the radius circle to take the intersection point). In this way, the horizontal relative position of the person outside the vehicle and the center point of the microphone 1401 and the microphone 1402 can be determined.

[0166] As shown in FIG1A , FIG1A is a top view schematic diagram for identifying external sound sources of a vehicle, and a specific method is described below in conjunction with an embodiment.

[0167] The specific implementation method is as follows. To make the solution easier to understand, we first explain the terms and values ​​involved in the solution. Sound pressure level (SPL): The human ear's perception of sound intensity is reflected through sound pressure level, and the unit is dB. Sound pressure level: LP = 20lg (P / P0) where P is sound pressure (Pa Pa), P0 is the reference sound pressure, which is 2*10 -5Pa. This value refers to the sound pressure at a frequency of 1000Hz, just at the point where the human ear can hear it, i.e., the minimum sound pressure. During normal conversation, the sound power of a person is defined as 1uW. When speaking loudly, the sound power of a person's voice can increase to 1mW, which means that the sound pressure measured at 1m is approximately 65-69dB. Sound attenuation (L): At a temperature of 20°C, the attenuation at a distance from the sound source is: L = 20logD. Where D is the front distance from the sound source (M meters). From this, it can be deduced that the attenuation at 1m from the sound source is L = 20log1 = 0dB. The attenuation at 2m is L = 20log2 = 6dB, and so on. In the above embodiment, the accuracy of the microphone's sound source localization does not need to be very precise; it only needs to confirm the approximate direction (distance and direction from the vehicle). Therefore, it can be preliminarily estimated that the sound pressure of a person speaking outside the vehicle (at a distance of 1 meter) is 66dB (reference value). At this time, the car-end microphones at different distances can measure different sound pressures of the input speakers, thereby inferring the distance relationship to the sound source. Using the sound pressure attenuation L = 20lgD, it is found that at 1 meter, the attenuation is 0, and the sound pressure at this time is 66dB. Therefore, when the distance is D, the sound pressure attenuation is L = LP (microphone) - 66dB = 20lgD. The measured LP can be used to determine the distance a and b between the corresponding microphone and the speaker. Because the spacing c of the microphone array is a fixed value, the positional relationship of the triangle abc is confirmed. The position of the person outside the car can also be known, that is, the sound source 1 of the person outside the car. This is the method of realizing vehicle positioning using the microphone array positioning method.

[0168] To improve the accuracy of off-vehicle positioning, it is recommended that the distance between the off-vehicle microphone array should not be too small. Otherwise, due to the small distance c, the position of the person outside the vehicle cannot be measured (the two circles almost overlap), and only the distance between the person outside the vehicle and the microphone can be determined. In this case, the distance position needs to be matched with the distance position captured by the off-vehicle camera or the off-vehicle digital key.

[0169] Of course, when a person outside the vehicle stands at the position of sound source 2, the array composed of microphone 2 and microphone 3 can identify the position of the person outside the vehicle.

[0170] As shown in Figure 15, a schematic diagram of a method for detecting the location of a sound source when the sound source volume changes (for ease of presentation, the distance ratios in the figure have been distorted and do not represent the distance ratios in a real-world scenario). To account for the varying volume of different people speaking, the following method can be used. Assuming the person's relative position relative to the vehicle remains unchanged, meaning the sound source location remains unchanged, but the volume changes, with a sound pressure change of ΔdB, microphone 1501 measures a change in distance from the sound source by Δa, and microphone 1502 measures a change in distance from the sound source by Δb. Consequently, the calculated position of the person outside the vehicle will also change. However, this change in sound pressure has little effect on position and orientation. For example, assuming that the distance between microphone 1501 and microphone 1502 is 0.05 meters, and the sound source is located between microphone 1501 and microphone 1502, the calculation is based on the preset parameter that the sound pressure of a person speaking at a distance of 1 meter is 66dB. Assuming that the distance between the sound source and microphone 1501 and microphone 1502 is 1 meter and 1.02 meters respectively, it can be obtained that the sound pressure measured by microphone 1501 is 66dB, and the sound pressure measured by microphone 1502 is approximately 65.828dB. At this time, the predicted distance between the sound source position 1503 and the vehicle is 0.925m. If the sound pressure of the sound source increases by 2dB at this time, recalculation can obtain the new distance between the sound source and microphone 1501 is 0.794m, and the distance from microphone 1502 is approximately 0.81m. At this time, the predicted distance between the sound source position 1504 and the vehicle is 0.759m, and the distance change from the vehicle is 0.166m. It can be seen that the distance change affected is not large. Therefore, sound source localization can still be performed based on a reference volume of 66dB. Changes in the volume of a human voice source have little impact on localization. Sound source localization allows the horizontal position of the speaker to be determined, enabling more robust speech recognition, such as better support for chatGPT (who is speaking to whom and where the speaker is standing). The vehicle can thus track the speaker's position and output vehicle information to the corresponding side. When external cameras cannot be used for external surveillance due to privacy protections, sound source localization can be used to enable better and more seamless communication with people outside the vehicle. When a piezoelectric microphone array is installed outside the vehicle, the sound field location of the person speaking outside the vehicle can be localized and tracked, further supporting external speech recognition and voice interaction. When multiple microphones are installed outside the vehicle, the external microphones can collectively capture the ambient noise outside the vehicle. Since the person speaking outside the vehicle is located at a specific location, the difference in the sound input from the external microphones can be used to modulate the person's speech, thereby improving the signal-to-noise ratio. The difference is not noticeable, and the direction can be largely identified.

[0171] In some embodiments, piezoelectric microphones require no power supply. Leveraging their inherent piezoelectric effect, they achieve a very short response time when sensing sound pressure. Consequently, piezoelectric microphones can operate in low-power modes. Existing technologies, such as MEMS (Micro-Electro-Mechanical System) microphones, require power for continuous monitoring. This demonstrates the unparalleled advantages of piezoelectric microphones.

[0172] In some embodiments, the control can be performed through the following process: the microphone analog signal enters the A / D analog-to-digital converter and is converted into a digital audio signal. The microphone's voice signal can be obtained through noise reduction processing inside the digital signal processor (DSP). At the same time, the DSP transmits the processed audio signal to the external ECU module. The microcontroller unit (MCU) or DSP processes the sound source positioning and sends the positioning information to the vehicle module through the corresponding transceiver method, that is, it can be sent to the vehicle module through interfaces such as UART (Universal Asynchronous Receiver-Transmitter), CAN (Controller Area Network), and LIN (Local Interconnect Network).

[0173] The microphone backplane array control unit can also be integrated into other ECUs. The piezoelectric microphone array and microphone backplane array control unit, along with noise cancellation and location tracking, form a complete piezoelectric microphone array integrated module solution that supports voice input in an off-vehicle environment.

[0174] The methods provided in the above embodiments can solve the following problems: (1) A piezoelectric array microphone structure for use in an off-vehicle environment is designed, which is dustproof, waterproof, and corrosion-resistant. (2) A noise cancellation system based on an off-vehicle piezoelectric vibrator microphone is designed, which improves the signal-to-noise ratio, and furthermore, the position of a person speaking can be determined based on multiple microphones. (3) The piezoelectric vibrator microphone has a low-power operation mode, while the piezoelectric microphone does not require a power supply and fully utilizes its own piezoelectric effect, resulting in a very short response time.

[0175] As shown in Figure 16, Figure 16 is a schematic diagram of the structure of a sound source identification device. The device includes a piezoelectric array microphone 1610, a control unit 1620, and a diaphragm (integrated into the microphone, not shown in the figure). The piezoelectric array microphone 1610 is mounted on the diaphragm at different locations on the vehicle panel. The piezoelectric array microphone 1610 and the control unit 1620 are connected. The functions of the control unit 1620 can be implemented by a processor in a computer device or by a specific logic circuit. In the implementation process, the processor can be a central processing unit (CPU), a microprocessor (MPU), a digital signal processor, or a field programmable gate array (FPGA). The control unit 1620 is used to implement some or all of the steps in the above method. Exemplarily, the control unit 1620 includes an A / D converter 1621 for converting the microphone analog signal into a digital audio signal. The digital signal processor 1622 is used to perform noise reduction processing on the digital audio signal to obtain the microphone voice signal and transmit the processed audio signal to an external ECU module (not shown in the figure). The microcontroller unit 1623 is used to process the sound source location and send the location information to the vehicle module through the corresponding transceiver. The power management chip (Power Management IC, PMIC) 1624 is used to receive power input and perform power management on the control unit 1620.

[0176] As shown in Figure 17, an embodiment of the present application provides an electronic device 1700, including a memory 1702 and a processor 1701, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the program, it implements some or all of the steps in the above method.

[0177] The present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements some or all of the steps in the above method. The computer-readable storage medium may be transient or non-transient.

[0178] An embodiment of the present application provides a computer program, including computer-readable code. When the computer-readable code is run in a computer device, a processor in the computer device executes some or all of the steps for implementing the above method.

[0179] An embodiment of the present application provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and when the computer program is read and executed by a computer, implements some or all of the steps in the above method. The computer program product can be implemented specifically by hardware, software, or a combination thereof. In some embodiments, the computer program product is embodied as a computer storage medium. In other embodiments, the computer program product is embodied as a software product, such as a software development kit (SDK), etc.

[0180] It should be noted that the descriptions of the various embodiments above tend to emphasize the differences between the various embodiments, and their similarities or similarities can be referenced to each other. The descriptions of the above device, storage medium, computer program, and computer program product embodiments are similar to the descriptions of the above method embodiments and have similar beneficial effects as the method embodiments. For technical details not disclosed in the embodiments of the device, storage medium, computer program, and computer program product of this application, please refer to the description of the method embodiments of this application for understanding. Industrial Applicability

[0181] The embodiments of the present application disclose a sound source identification method, apparatus, device, and storage medium, wherein the method includes: obtaining ambient sound pressure through at least two piezoelectric vibrators; calibrating the ambient sound pressure obtained by each piezoelectric vibrator based on the adjustment coefficient of each piezoelectric vibrator to obtain the target sound pressure of each piezoelectric vibrator; using the relationship between sound pressure and distance, converting the target sound pressure of each piezoelectric vibrator into the distance between each piezoelectric vibrator and the sound source; and determining the position of the sound source based on the position of each piezoelectric vibrator and the distance between each piezoelectric vibrator and the sound source. The method provided by the above embodiment can greatly reduce the complexity of system logic operations and improve the operation speed during positioning.

Claims

1. A sound source identification method, applied to a sound source identification device, the method comprising: Acquiring ambient sound pressure through at least two piezoelectric vibrators; Calibrate the ambient sound pressure acquired by each piezoelectric vibrator based on the adjustment coefficient of each piezoelectric vibrator to obtain the target sound pressure of each piezoelectric vibrator; Using the relationship between the change in sound pressure and distance, convert the target sound pressure of each of the piezoelectric vibrators into the distance between each of the piezoelectric vibrators and the sound source; The position of the sound source is determined based on the position of each of the piezoelectric vibrators and the distance between each of the piezoelectric vibrators and the sound source.

2. The method according to claim 1, wherein: The step of calibrating the ambient sound pressure acquired by each piezoelectric vibrator based on the adjustment coefficient of each piezoelectric vibrator to obtain the target sound pressure of each piezoelectric vibrator includes: Adjusting the ambient sound pressure obtained by each piezoelectric vibrator based on the adjustment coefficient of each piezoelectric vibrator to obtain an adjusted ambient sound pressure; each of the adjusted ambient sound pressures includes the same noise sound pressure; The adjusted ambient sound pressure of each of the piezoelectric vibrators is used as the target sound pressure of each of the piezoelectric vibrators.

3. The method according to claim 2, wherein: Before calibrating the ambient sound pressure acquired by each piezoelectric vibrator based on the adjustment coefficient of each piezoelectric vibrator to obtain the target sound pressure of each piezoelectric vibrator, the method further includes: Acquiring a standard noise sound pressure and a noise sound pressure in each of the environmental sound pressures; An adjustment coefficient of each of the piezoelectric vibrators is determined based on the standard noise sound pressure and the noise sound pressure acquired by each of the piezoelectric vibrators.

4. The method according to claim 3, wherein: The at least two piezoelectric vibrators are mounted on the diaphragm at different positions on the vehicle panel, the adjustment coefficient is related to the softness of the diaphragm, and the distance between any two of the at least two piezoelectric vibrators is greater than or equal to a preset distance.

5. The method according to any one of claims 1 to 4, wherein: The method of converting the target sound pressure of each of the piezoelectric vibrators into the distance between each of the piezoelectric vibrators and the sound source by utilizing the relationship between the sound pressure and the distance includes: For each of the piezoelectric vibrators, a sound pressure variation is determined based on a preset sound pressure parameter and the target sound pressure; and a distance between the piezoelectric vibrator and a sound source is determined based on the sound pressure variation.

6. The method according to any one of claims 1 to 4, wherein: The determining the position of the sound source based on the position of each of the piezoelectric vibrators and the distance between each of the piezoelectric vibrators and the sound source comprises: Determine a circle corresponding to each piezoelectric vibrator with the position of each piezoelectric vibrator as the center and the distance between each piezoelectric vibrator and the sound source as the radius; The position of the intersection point located outside the vehicle among the intersection points of the circles corresponding to each of the piezoelectric vibrators is determined as the sound source position.

7. The method according to any one of claims 1 to 4, wherein: The method further comprises: Performing denoising processing based on target sound pressures of at least two piezoelectric vibrators to obtain speech data corresponding to the sound source; Performing speech recognition based on the speech data to obtain a recognition result; Interaction is performed based on the recognition result and the position of the sound source.

8. The method according to claim 7, wherein: The denoising is performed based on the target sound pressure of at least two piezoelectric vibrators to obtain the speech data corresponding to the sound source, including: The target sound pressures of at least two of the piezoelectric vibrators are differentially processed to obtain speech data corresponding to the sound source.

9. A sound source identification device, wherein: It comprises a piezoelectric array microphone, a control unit and a diaphragm, wherein the piezoelectric array microphone is mounted on the diaphragm at different positions on the vehicle panel, and the piezoelectric array microphone is connected to the control unit, wherein: The control unit is used to execute the method according to any one of claims 1 to 8.

10. An electronic device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor implements the steps of the method according to any one of claims 1 to 8 when executing the computer program.

11. A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method according to any one of claims 1 to 8.

Images