Earphone adaptive active noise reduction method based on environment sound wave real-time analysis

Abstract

The invention relates to the field of earphone noise reduction, and discloses an earphone adaptive active noise reduction method based on environment sound wave real-time analysis, comprising the following steps: collecting environment sound waves inside and outside an earphone and preprocessing the environment sound waves to obtain space-time synchronous multichannel environment sound waves; noise feature analysis is carried out on space-time synchronization multichannel environment sound waves, the position of short-time energy sudden change sound waves is positioned, self-adaptive active noise reduction filtering is carried out, and finally energy consumption monitoring and energy consumption optimization control are carried out on the earphone. According to the invention, adaptive active noise reduction of the earphone can be realized, hearing potential safety hazards of a user and earphone energy consumption waste are prevented, and the purposes of energy conservation and comfort of the user are realized.

Description

Technical Field

[0001] The present invention relates to the field of headphone noise reduction, and in particular to a headphone adaptive active noise reduction method based on real-time analysis of ambient sound waves. Background Art

[0002] Adaptive active noise cancellation (ANC) in headphones aims to enhance the user's auditory experience in complex environments through intelligent, real-time noise processing. This technology automatically adjusts noise cancellation parameters by analyzing ambient sound wave characteristics (frequency, intensity, and directionality) in real time. This addresses the performance degradation of traditional fixed-parameter ANC in variable noise environments. It also identifies different scenarios (such as airplane cabins, subways, and offices) to optimize the noise cancellation strategy, balancing noise suppression with the audibility of voice and warning sounds.

[0003] Without adaptive active noise cancellation (ANC), headphones that continuously use noise cancellation can pose a risk to hearing safety when outdoors, such as masking traffic warnings and increasing the risk of outdoor use. Furthermore, the constant, inverse-phase sound waves produced by static noise cancellation can easily cause ear stuffiness, reducing comfort during long-term use. Adaptive ANC overcomes the limitations of traditional "one-size-fits-all" approaches through a closed-loop control process: environmental perception, intelligent decision-making, and dynamic execution. Essentially, it elevates noise management from static defense to dynamic negotiation, achieving an intelligent auditory experience that allows for "quiet when quiet is needed and listening when listening is needed" while ensuring safety. Therefore, we propose an adaptive ANC method for headphones based on real-time analysis of ambient sound waves. Summary of the Invention

[0004] The present invention overcomes the deficiencies of the prior art and provides an adaptive active noise reduction method for headphones based on real-time analysis of ambient sound waves.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is: A first aspect of the present invention provides a method for adaptive active noise reduction of headphones based on real-time analysis of ambient sound waves, comprising the following steps: Collect multi-channel environmental sound waves, perform acoustic scene classification and spatiotemporal synchronization processing on the multi-channel environmental sound waves, and obtain spatiotemporal synchronized multi-channel environmental sound waves; Perform noise feature analysis on spatiotemporal synchronous multi-channel environmental sound waves, and detect the position of short-term energy mutation sound waves on spatiotemporal synchronous multi-channel environmental sound waves; Combined with the position of the short-time energy mutation sound wave, the environmental sound wave to be analyzed is subjected to adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering; During the real-time noise reduction environmental sound wave output, the target earphones are subjected to energy consumption monitoring and energy consumption optimization control.

[0006] Furthermore, in a preferred embodiment of the present invention, the multi-channel ambient sound waves are collected, and the multi-channel ambient sound waves are subjected to acoustic scene classification and spatiotemporal synchronization processing to obtain spatiotemporal synchronized multi-channel ambient sound waves, specifically: Obtain a headset that requires adaptive active noise reduction, mark it as a target headset, and install a microphone on the target headset; The microphone on the target earphone includes a microphone for collecting ambient sound waves outside the target earphone and a microphone for collecting ambient sound waves in the ear canal; The target headset's microphone collects ambient sound waves in real time, including ambient sound waves outside the target headset and ambient sound waves inside the ear canal, and calibrates the collected ambient sound waves into multi-channel ambient sound waves; Calculate the Mel spectrum coherence coefficient of the multi-channel ambient sound wave, and calculate the spatial and temporal characteristics of the multi-channel ambient sound wave by combining the Mel spectrum coherence coefficient of the multi-channel ambient sound wave; The spatial and temporal features of the multi-channel ambient sound waves are aligned in real time to ensure that the multi-channel ambient sound waves are spatiotemporally synchronized at different times and in different spaces, thereby obtaining spatiotemporally synchronized multi-channel ambient sound waves.

[0007] Furthermore, in a preferred embodiment of the present invention, the noise characteristic analysis is performed on the spatiotemporal synchronous multi-channel environmental sound waves, and the short-time energy mutation sound wave position is detected on the spatiotemporal synchronous multi-channel environmental sound waves, specifically: The Hanning window algorithm is introduced to perform frame-by-frame windowing processing on the spatiotemporal synchronous multi-channel environmental sound waves, and the windowed spatiotemporal synchronous multi-channel environmental sound waves are obtained and calibrated as the environmental sound waves to be analyzed. Determining a signal-to-noise ratio of the ambient sound wave to be analyzed, determining a weighted average weight of the ambient sound wave to be analyzed based on the signal-to-noise ratio of the ambient sound wave to be analyzed, performing weighted average processing, constructing a multi-window spectrum of the ambient sound wave to be analyzed, and calibrating it as a target spectrum; Perform harmonic consistency verification on the target spectrum, and select the frequency group that meets the harmonic consistency and has the highest energy in the target spectrum to obtain the main frequency of the environmental sound wave to be analyzed, and calibrate it as the target main frequency; The energy of the ambient sound wave to be analyzed is calculated based on the target main frequency and the target spectrum, and an energy threshold is preset. The position where the energy is greater than the energy threshold is determined in the ambient sound wave to be analyzed and marked as the energy mutation position to be analyzed; The mutation difference is calculated for all energy mutation positions to be analyzed, and the position with the mutation difference greater than the preset value is retrieved from all energy mutation positions to be analyzed and calibrated as the short-time energy mutation acoustic wave position.

[0008] Furthermore, in a preferred embodiment of the present invention, the adaptive transient noise reduction filtering and the adaptive steady-state noise reduction filtering are performed on the environmental sound waves to be analyzed in combination with the position of the short-time energy mutation sound waves, specifically: Obtaining a dynamic filter module, and connecting the dynamic filter module to the target earphone, so that the dynamic filter module can perform filtering processing on the target earphone; Perform position distribution analysis on the short-time energy mutation sound wave positions of the environmental sound waves to be analyzed, calculate the distribution intervals between different short-time energy mutation sound wave positions, and preset the minimum distribution interval; If the distribution intervals between the positions of different short-time energy mutation sound waves are not less than the minimum distribution interval, it is necessary to adopt adaptive steady-state noise reduction filtering for the environmental sound waves to be analyzed; if the distribution intervals between the positions of different short-time energy mutation sound waves are greater than the minimum distribution interval, it is necessary to adopt adaptive transient noise reduction filtering for the environmental sound waves to be analyzed; If adaptive transient noise reduction filtering is required for the ambient sound wave to be analyzed, the main frequency and target spectrum of the ambient sound wave to be analyzed are imported into the dynamic filtering module in real time for nonlinear phase compensation; The nonlinear phase compensation is to perform pre-distortion processing on the ambient sound wave to be analyzed, calculate the phase compensation value of the ambient sound wave to be analyzed, and apply the phase compensation value of the ambient sound wave to be analyzed to the target earphone in real time through the dynamic filtering module to achieve adaptive transient noise reduction filtering of the ambient sound wave to be analyzed; If it is necessary to adopt adaptive steady-state noise reduction filtering for the environmental sound wave to be analyzed, then the adaptive steady-state noise reduction filtering is adopted for the environmental sound wave to be analyzed, and the adaptive transient noise reduction filtering and the adaptive steady-state noise reduction filtering are mixed together.

[0009] Furthermore, in a preferred embodiment of the present invention, the environmental sound waves to be analyzed are subjected to adaptive steady-state noise reduction filtering, and adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering are mixed, specifically: An improved FxLMS algorithm is introduced into the dynamic filtering module. The improved FxLMS algorithm is used to update the target spectrum step size of the environmental sound wave to be analyzed in the dynamic filtering module, and energy leakage of the environmental sound wave to be analyzed is controlled during the target spectrum compensation update process. The energy leakage control of the environmental sound wave to be analyzed is to control the energy of the target spectrum of the environmental sound wave to be analyzed to always be maintained within the energy threshold; In the dynamic filtering module, the distribution intervals between the positions of different short-time energy mutation sound waves are analyzed in real time, and adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering are switched in real time based on the distribution intervals; The environmental sound waves to be analyzed that have passed through adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering are synthesized in multiple frequency bands, and the environmental sound waves to be analyzed after the multi-band synthesis are calibrated as real-time noise reduction environmental sound waves.

[0010] Furthermore, in a preferred embodiment of the present invention, during the real-time noise reduction ambient sound wave output, the target earphones are subjected to energy consumption monitoring and energy consumption optimization control, specifically: Spectral subtraction is introduced to eliminate residual noise from the real-time noise reduction environment sound waves, while obtaining the surrounding environment parameters of the target headphones; Calculate the real-time energy consumption change data of the target headphones in processing real-time noise reduction environmental sound waves and construct a real-time energy consumption change curve; Presetting a standard energy consumption threshold and analyzing the real-time energy consumption change curve to determine whether the real-time energy consumption of the target headset is always maintained within the standard energy consumption threshold; If so, calibrate the target earphones as qualified energy consumption target earphones, and maintain the qualified energy consumption target earphones in operation to achieve adaptive active noise reduction of ambient sound waves; If not, the target earphone is calibrated as an unqualified energy consumption target earphone, and the grey correlation method is introduced to calculate the grey correlation value between the real-time energy consumption of the unqualified energy consumption target earphone and the corresponding surrounding environment parameters, and calibrate it as the first grey correlation value; A gray correlation threshold is preset. If the first gray correlation value remains within the gray correlation threshold, it is determined that the real-time energy consumption of the unqualified energy consumption target earphone is affected by the corresponding surrounding environment parameter, and a big data network is introduced to retrieve a solution for protecting the unqualified energy consumption target earphone from being affected by the corresponding surrounding environment parameter and output the solution; If the first grey correlation value is not maintained within the grey correlation threshold, controlling the unqualified energy consumption target earphone to perform real-time decibel calculation of the ambient sound wave to be analyzed to determine the real-time decibel value of the ambient sound wave to be analyzed; A pre-set real-time decibel threshold for harm is set. If the real-time decibel value of the ambient sound wave to be analyzed is not greater than the real-time decibel threshold for harm, there is no need to perform adaptive active noise reduction on the unqualified energy consumption target headphones. If the real-time decibel value of the ambient sound wave to be analyzed is greater than the harmful real-time decibel threshold, the unqualified energy consumption target headphones are controlled to perform adaptive active noise reduction and output real-time noise reduction ambient sound waves.

[0011] A second aspect of the present invention further provides an adaptive active noise reduction system for headphones based on real-time analysis of ambient sound waves. The adaptive active noise reduction system for headphones includes a memory and a processor. The memory stores an adaptive active noise reduction method for headphones. When the adaptive active noise reduction method for headphones is executed by the processor, the following steps are implemented: Collect multi-channel environmental sound waves, perform acoustic scene classification and spatiotemporal synchronization processing on the multi-channel environmental sound waves, and obtain spatiotemporal synchronized multi-channel environmental sound waves; Perform noise feature analysis on spatiotemporal synchronous multi-channel environmental sound waves, and detect the position of short-term energy mutation sound waves on spatiotemporal synchronous multi-channel environmental sound waves; Combined with the position of the short-time energy mutation sound wave, the environmental sound wave to be analyzed is subjected to adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering; During the real-time noise reduction environmental sound wave output, the target earphones are subjected to energy consumption monitoring and energy consumption optimization control.

[0012] The present invention addresses the technical deficiencies in the background art and has the following beneficial effects: it collects ambient sound waves inside and outside the headphones and pre-processes them to obtain spatiotemporally synchronized multi-channel ambient sound waves; it performs noise feature analysis on the spatiotemporally synchronized multi-channel ambient sound waves to locate the location of short-term energy mutation sound waves, and performs adaptive active noise reduction filtering; and finally, it monitors and optimizes the energy consumption of the headphones. The present invention can achieve adaptive active noise reduction in headphones, preventing hearing safety hazards and energy waste in the headphones, thereby achieving energy conservation and user comfort. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, without paying any creative work, they can also obtain drawings of other embodiments based on these drawings.

[0014] Figure 1 A flow chart of a method for adaptive active noise reduction of headphones based on real-time analysis of ambient sound waves is shown; Figure 2 A flow chart of a method for performing adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering on an environmental sound wave to be analyzed is shown; Figure 3 A program view of a headphone adaptive active noise reduction system based on real-time analysis of ambient sound waves is shown. DETAILED DESCRIPTION

[0015] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the absence of conflict, the embodiments of the present application and the features therein can be combined with each other.

[0016] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0017] Figure 1 A flow chart of a method for adaptive active noise reduction of headphones based on real-time analysis of ambient sound waves is shown, comprising the following steps: Collect multi-channel environmental sound waves, perform acoustic scene classification and spatiotemporal synchronization processing on the multi-channel environmental sound waves, and obtain spatiotemporal synchronized multi-channel environmental sound waves; Perform noise feature analysis on spatiotemporal synchronous multi-channel environmental sound waves, and detect the position of short-term energy mutation sound waves on spatiotemporal synchronous multi-channel environmental sound waves; Combined with the position of the short-time energy mutation sound wave, the environmental sound wave to be analyzed is subjected to adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering; During the real-time noise reduction environmental sound wave output, the target earphones are subjected to energy consumption monitoring and energy consumption optimization control.

[0018] Furthermore, in a preferred embodiment of the present invention, the multi-channel ambient sound waves are collected, and the multi-channel ambient sound waves are subjected to acoustic scene classification and spatiotemporal synchronization processing to obtain spatiotemporal synchronized multi-channel ambient sound waves, specifically: Obtain a headset that requires adaptive active noise reduction, mark it as a target headset, and install a microphone on the target headset; The microphone on the target earphone includes a microphone for collecting ambient sound waves outside the target earphone and a microphone for collecting ambient sound waves in the ear canal; The target headset's microphone collects ambient sound waves in real time, including ambient sound waves outside the target headset and ambient sound waves inside the ear canal, and calibrates the collected ambient sound waves into multi-channel ambient sound waves; Calculate the Mel spectrum coherence coefficient of the multi-channel ambient sound wave, and calculate the spatial and temporal characteristics of the multi-channel ambient sound wave by combining the Mel spectrum coherence coefficient of the multi-channel ambient sound wave; The spatial and temporal features of the multi-channel ambient sound waves are aligned in real time to ensure that the multi-channel ambient sound waves are spatiotemporally synchronized at different times and in different spaces, thereby obtaining spatiotemporally synchronized multi-channel ambient sound waves.

[0019] It should be noted that the ambient sound waves include those outside the earphones and those inside the earphones. The ambient sound waves inside the earphones are the sounds inside the ear canal after the human body wears the earphones, so it is necessary to collect the sounds outside and inside the earphones, and install different microphones for collection. Since the sounds outside and inside the earphones are collected, the ambient sound waves are multi-channel ambient sound waves. The multi-channel ambient sound waves need to be pre-processed, including the spatiotemporal synchronization of the sound waves, in order to synchronize the arrival time differences of the sound waves and ensure the synchronization rate and accuracy during the noise reduction process. The Mel-spectral coherence coefficient of the multi-channel ambient sound waves is calculated in order to enhance the temporal and spatial features and to construct a feature matrix. Only after the feature matrix is constructed can spatiotemporal synchronization be performed. The method of spatiotemporal synchronization is to fuse the features to obtain spatiotemporally synchronized multi-channel ambient sound waves.

[0020] Furthermore, in a preferred embodiment of the present invention, the noise characteristic analysis is performed on the spatiotemporal synchronous multi-channel environmental sound waves, and the short-time energy mutation sound wave position is detected on the spatiotemporal synchronous multi-channel environmental sound waves, specifically: The Hanning window algorithm is introduced to perform frame-by-frame windowing processing on the spatiotemporal synchronous multi-channel environmental sound waves, and the windowed spatiotemporal synchronous multi-channel environmental sound waves are obtained and calibrated as the environmental sound waves to be analyzed. Determining a signal-to-noise ratio of the ambient sound wave to be analyzed, determining a weighted average weight of the ambient sound wave to be analyzed based on the signal-to-noise ratio of the ambient sound wave to be analyzed, performing weighted average processing, constructing a multi-window spectrum of the ambient sound wave to be analyzed, and calibrating it as a target spectrum; Perform harmonic consistency verification on the target spectrum, and select the frequency group that meets the harmonic consistency and has the highest energy in the target spectrum to obtain the main frequency of the environmental sound wave to be analyzed, and calibrate it as the target main frequency; The energy of the ambient sound wave to be analyzed is calculated based on the target main frequency and the target spectrum, and an energy threshold is preset. The position where the energy is greater than the energy threshold is determined in the ambient sound wave to be analyzed and marked as the energy mutation position to be analyzed; The mutation difference is calculated for all energy mutation positions to be analyzed, and the position with the mutation difference greater than the preset value is retrieved from all energy mutation positions to be analyzed and calibrated as the short-time energy mutation acoustic wave position.

[0021] It should be noted that the Hanning window is a cosine-squared windowing function used to process continuous audio signals. It suppresses spectral leakage of sound waves, maintains frequency resolution, and achieves smooth transitions between audio frames. In this application, Hanning windowing is performed to make the ambient sound waves smoother during the noise reduction process and avoid the generation of splicing noise. The signal-to-noise ratio of the ambient sound waves to be analyzed is determined by calculating the time domain of the sound wave signal, which is the ratio of signal to noise. The purpose of the weighted averaging of the sound waves after determining the signal-to-noise ratio is to construct a spectrum for extracting the frequency signal of the sound wave. Harmonic consistency is verified for the target spectrum. Harmonics are noise. The peak positions of the sound wave frequencies in the target spectrum are checked and then instantly calculated in combination with the fundamental frequency. If the harmonics maintain a constant value after multiple checks, harmonic consistency is met. At this point, the energy of the ambient sound waves needs to be calculated. This means that the energy of the ambient sound waves during output will be higher if noise is present. If the energy is higher at certain locations, it indicates that the noise is higher at those locations, and the sound wave needs to undergo noise reduction processing at the corresponding locations. The position of the short-time energy mutation sound wave is where the sudden energy change is large, proving that there is a lot of sudden noise. This is the position of the short-time energy mutation sound wave, and it is necessary to focus on noise reduction processing at the position of the short-time energy mutation sound wave, because the suddenly generated noise can easily affect the user experience and even cause danger.

[0022] Furthermore, in a preferred embodiment of the present invention, during the real-time noise reduction ambient sound wave output, the target earphones are subjected to energy consumption monitoring and energy consumption optimization control, specifically: Spectral subtraction is introduced to eliminate residual noise from the real-time noise reduction environment sound waves, while obtaining the surrounding environment parameters of the target headphones; Calculate the real-time energy consumption change data of the target headphones in processing real-time noise reduction environmental sound waves and construct a real-time energy consumption change curve; Presetting a standard energy consumption threshold and analyzing the real-time energy consumption change curve to determine whether the real-time energy consumption of the target headset is always maintained within the standard energy consumption threshold; If so, calibrate the target earphones as qualified energy consumption target earphones, and maintain the qualified energy consumption target earphones in operation to achieve adaptive active noise reduction of ambient sound waves; If not, the target earphone is calibrated as an unqualified energy consumption target earphone, and the grey correlation method is introduced to calculate the grey correlation value between the real-time energy consumption of the unqualified energy consumption target earphone and the corresponding surrounding environment parameters, and calibrate it as the first grey correlation value; A gray correlation threshold is preset. If the first gray correlation value remains within the gray correlation threshold, it is determined that the real-time energy consumption of the unqualified energy consumption target earphone is affected by the corresponding surrounding environment parameter, and a big data network is introduced to retrieve a solution for protecting the unqualified energy consumption target earphone from being affected by the corresponding surrounding environment parameter and output the solution; If the first grey correlation value is not maintained within the grey correlation threshold, controlling the unqualified energy consumption target earphone to perform real-time decibel calculation of the ambient sound wave to be analyzed to determine the real-time decibel value of the ambient sound wave to be analyzed; A pre-set real-time decibel threshold for harm is set. If the real-time decibel value of the ambient sound wave to be analyzed is not greater than the real-time decibel threshold for harm, there is no need to perform adaptive active noise reduction on the unqualified energy consumption target headphones. If the real-time decibel value of the ambient sound wave to be analyzed is greater than the harmful real-time decibel threshold, the unqualified energy consumption target headphones are controlled to perform adaptive active noise reduction and output real-time noise reduction ambient sound waves.

[0023] It should be noted that spectral subtraction can remove any residual noise from the real-time noise reduction ambient sound waves, ensuring a clearer, noise-free sound. Headphones also need to calculate their energy consumption. High energy consumption means faster power consumption, which is detrimental to energy conservation. Furthermore, headphones with high energy consumption indicate they are always in noise reduction mode, not adaptive noise reduction. Headphones with adaptive noise reduction should have lower energy consumption. After calculating the target headphone's energy consumption, the relationship between the energy consumption and the standard threshold is determined. If the energy consumption is within the threshold, the energy consumption is normal and the headphone meets the energy consumption standards. Headphone energy consumption may be related to ambient environmental parameters. For example, if the ambient temperature is high, the headphone will generate more heat, which will increase energy consumption and power consumption. The gray correlation method can calculate the correlation between headphone energy consumption and environmental parameters. If the correlation is greater than a preset value, it indicates that the abnormal headphone energy consumption is related to the environmental parameters. In this case, the headphone needs to be treated to prevent the influence of environmental parameters and ensure that the headphone energy consumption returns to normal. If the correlation is less than the preset value, the decibel level of the ambient sound waves is determined to determine whether noise reduction processing is necessary. If the decibel value is large, noise reduction must be performed, otherwise it will damage the user's hearing. Otherwise, it is not necessary to do so, which can achieve the purpose of reducing energy consumption.

[0024] Figure 2 A flow chart of a method for performing adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering on an environmental sound wave to be analyzed is shown, comprising the following steps: S202: performing adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering on the environmental sound waves to be analyzed based on the position of the sound waves with short-time energy mutations; S204: Adaptive steady-state noise reduction filtering is performed on the environmental sound waves to be analyzed, and adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering are mixed.

[0025] Furthermore, in a preferred embodiment of the present invention, the adaptive transient noise reduction filtering and the adaptive steady-state noise reduction filtering are performed on the environmental sound waves to be analyzed in combination with the position of the short-time energy mutation sound waves, specifically: Obtaining a dynamic filter module, and connecting the dynamic filter module to the target earphone, so that the dynamic filter module can perform filtering processing on the target earphone; Perform position distribution analysis on the short-time energy mutation sound wave positions of the environmental sound waves to be analyzed, calculate the distribution intervals between different short-time energy mutation sound wave positions, and preset the minimum distribution interval; If the distribution intervals between the positions of different short-time energy mutation sound waves are not less than the minimum distribution interval, it is necessary to adopt adaptive steady-state noise reduction filtering for the environmental sound waves to be analyzed; if the distribution intervals between the positions of different short-time energy mutation sound waves are greater than the minimum distribution interval, it is necessary to adopt adaptive transient noise reduction filtering for the environmental sound waves to be analyzed; If adaptive transient noise reduction filtering is required for the ambient sound wave to be analyzed, the main frequency and target spectrum of the ambient sound wave to be analyzed are imported into the dynamic filtering module in real time for nonlinear phase compensation; The nonlinear phase compensation is to perform pre-distortion processing on the ambient sound wave to be analyzed, calculate the phase compensation value of the ambient sound wave to be analyzed, and apply the phase compensation value of the ambient sound wave to be analyzed to the target earphone in real time through the dynamic filtering module to achieve adaptive transient noise reduction filtering of the ambient sound wave to be analyzed; If it is necessary to adopt adaptive steady-state noise reduction filtering for the environmental sound wave to be analyzed, then the adaptive steady-state noise reduction filtering is adopted for the environmental sound wave to be analyzed, and the adaptive transient noise reduction filtering and the adaptive steady-state noise reduction filtering are mixed together.

[0026] It should be noted that adaptive noise reduction processing is required for the environmental sound waves being analyzed, taking into account the locations of the short-term energy mutations. The distribution intervals between the different short-term energy mutation locations determine whether to use adaptive transient noise reduction filtering or adaptive steady-state noise reduction filtering. If the distribution intervals are small, steady-state filtering is used, as this indicates frequent noise mutations, requiring smooth and continuous noise reduction filtering. Conversely, if the distribution intervals are large, indicating that the noise typically occurs suddenly and is not very continuous, transient noise reduction filtering can achieve the purpose of adaptive noise reduction while also saving energy. Adaptive transient noise reduction filtering performs real-time phase compensation on the sound wave signal. The real-time phase and the phase difference to be compensated are determined based on the main frequency of the environmental sound wave being analyzed and the target spectrum analysis. Predistortion processing is used to calculate nonlinear phase compensation values, and the dynamic filtering module can filter high-frequency, high-energy sound waves.

[0027] Furthermore, in a preferred embodiment of the present invention, the environmental sound waves to be analyzed are subjected to adaptive steady-state noise reduction filtering, and adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering are mixed, specifically: An improved FxLMS algorithm is introduced into the dynamic filtering module. The improved FxLMS algorithm is used to update the target spectrum step size of the environmental sound wave to be analyzed in the dynamic filtering module, and energy leakage of the environmental sound wave to be analyzed is controlled during the target spectrum compensation update process. The energy leakage control of the environmental sound wave to be analyzed is to control the energy of the target spectrum of the environmental sound wave to be analyzed to always be maintained within the energy threshold; In the dynamic filtering module, the distribution intervals between the positions of different short-time energy mutation sound waves are analyzed in real time, and adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering are switched in real time based on the distribution intervals; The environmental sound waves to be analyzed that have passed through adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering are synthesized in multiple frequency bands, and the environmental sound waves to be analyzed after the multi-band synthesis are calibrated as real-time noise reduction environmental sound waves.

[0028] It should be noted that for adaptive steady-state noise reduction filtering, the improved FxLMS algorithm updates the sound waves with a variable step size. By changing the step size of the sound wave, sound wave energy leakage control is achieved, that is, preventing the sound wave energy from suddenly changing and overflowing during transmission. When leakage is always controlled, adaptive steady-state noise reduction filtering is achieved. It is necessary to switch between different filtering methods through real-time analysis of the distribution interval, because the sound wave will not always be in a state of small or large distribution interval. If there are times when the distribution interval is small, and energy leakage control has been performed before, it is necessary to perform a steady-state to transient noise reduction filtering process, thereby achieving the purpose of adaptive noise reduction and reducing energy consumption. The sound waves after noise reduction filtering with different methods need to be synthesized into the same sound wave. Multi-band synthesis can achieve this. Finally, the sound waves obtained by noise reduction filtering with different methods are combined to output a real-time noise reduction environment sound wave.

[0029] like Figure 3 As shown, the second aspect of the present invention further provides an adaptive active noise reduction system for headphones based on real-time analysis of ambient sound waves. The adaptive active noise reduction system for headphones includes a memory 31 and a processor 32. The memory 31 stores an adaptive active noise reduction method for headphones. When the adaptive active noise reduction method for headphones is executed by the processor 32, the following steps are implemented: Collect multi-channel environmental sound waves, perform acoustic scene classification and spatiotemporal synchronization processing on the multi-channel environmental sound waves, and obtain spatiotemporal synchronized multi-channel environmental sound waves; Perform noise feature analysis on spatiotemporal synchronous multi-channel environmental sound waves, and detect the position of short-term energy mutation sound waves on spatiotemporal synchronous multi-channel environmental sound waves; Combined with the position of the short-time energy mutation sound wave, the environmental sound wave to be analyzed is subjected to adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering; During the real-time noise reduction environmental sound wave output, the target earphones are subjected to energy consumption monitoring and energy consumption optimization control.

[0030] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A headphone adaptive active noise reduction method based on real-time analysis of ambient sound waves, characterized in that: The following steps are involved: Collect multi-channel environmental sound waves, perform acoustic scene classification and spatiotemporal synchronization processing on the multi-channel environmental sound waves, and obtain spatiotemporal synchronized multi-channel environmental sound waves; Perform noise feature analysis on spatiotemporal synchronous multi-channel environmental sound waves, and detect the position of short-term energy mutation sound waves on spatiotemporal synchronous multi-channel environmental sound waves; Combined with the position of the short-time energy mutation sound wave, the environmental sound wave to be analyzed is subjected to adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering; During the real-time noise reduction environmental sound wave output, the target earphones are subjected to energy consumption monitoring and energy consumption optimization control.

2. The headphone adaptive active noise reduction method based on real-time analysis of ambient sound waves according to claim 1, characterized in that: The method of collecting multi-channel ambient sound waves and performing acoustic scene classification and spatiotemporal synchronization processing on the multi-channel ambient sound waves to obtain spatiotemporal synchronized multi-channel ambient sound waves is specifically as follows: Obtain a headset that requires adaptive active noise reduction, mark it as a target headset, and install a microphone on the target headset; The microphone on the target earphone includes a microphone for collecting ambient sound waves outside the target earphone and a microphone for collecting ambient sound waves in the ear canal; The target headset's microphone collects ambient sound waves in real time, including ambient sound waves outside the target headset and ambient sound waves inside the ear canal, and calibrates the collected ambient sound waves into multi-channel ambient sound waves; Calculate the Mel spectrum coherence coefficient of the multi-channel ambient sound wave, and calculate the spatial and temporal characteristics of the multi-channel ambient sound wave by combining the Mel spectrum coherence coefficient of the multi-channel ambient sound wave; The spatial and temporal features of the multi-channel ambient sound waves are aligned in real time to ensure that the multi-channel ambient sound waves are spatiotemporally synchronized at different times and in different spaces, thereby obtaining spatiotemporally synchronized multi-channel ambient sound waves.

3. The headphone adaptive active noise reduction method based on real-time analysis of ambient sound waves according to claim 1, characterized in that: The noise characteristic analysis is performed on the spatiotemporal synchronous multi-channel environmental sound waves, and the short-time energy mutation sound wave position is detected on the spatiotemporal synchronous multi-channel environmental sound waves, specifically: The Hanning window algorithm is introduced to perform frame-by-frame windowing processing on the spatiotemporal synchronous multi-channel environmental sound waves, and the windowed spatiotemporal synchronous multi-channel environmental sound waves are obtained and calibrated as the environmental sound waves to be analyzed. Determining a signal-to-noise ratio of the ambient sound wave to be analyzed, determining a weighted average weight of the ambient sound wave to be analyzed based on the signal-to-noise ratio of the ambient sound wave to be analyzed, performing weighted average processing, constructing a multi-window spectrum of the ambient sound wave to be analyzed, and calibrating it as a target spectrum; Perform harmonic consistency verification on the target spectrum, and select the frequency group that meets the harmonic consistency and has the highest energy in the target spectrum to obtain the main frequency of the environmental sound wave to be analyzed, and calibrate it as the target main frequency; The energy of the ambient sound wave to be analyzed is calculated based on the target main frequency and the target spectrum, and an energy threshold is preset. The position where the energy is greater than the energy threshold is determined in the ambient sound wave to be analyzed and marked as the energy mutation position to be analyzed; The mutation difference is calculated for all energy mutation positions to be analyzed, and the position with the mutation difference greater than the preset value is retrieved from all energy mutation positions to be analyzed and calibrated as the short-time energy mutation acoustic wave position.

4. The headphone adaptive active noise reduction method based on real-time analysis of ambient sound waves according to claim 1, characterized in that: The adaptive transient noise reduction filtering and the adaptive steady-state noise reduction filtering are performed on the environmental sound waves to be analyzed in combination with the position of the short-time energy mutation sound waves, specifically: Obtaining a dynamic filter module, and connecting the dynamic filter module to the target earphone, so that the dynamic filter module can perform filtering processing on the target earphone; Perform position distribution analysis on the short-time energy mutation sound wave positions of the environmental sound waves to be analyzed, calculate the distribution intervals between different short-time energy mutation sound wave positions, and preset the minimum distribution interval; If the distribution intervals between the positions of different short-time energy mutation sound waves are not less than the minimum distribution interval, it is necessary to adopt adaptive steady-state noise reduction filtering for the environmental sound waves to be analyzed; if the distribution intervals between the positions of different short-time energy mutation sound waves are greater than the minimum distribution interval, it is necessary to adopt adaptive transient noise reduction filtering for the environmental sound waves to be analyzed; If adaptive transient noise reduction filtering is required for the ambient sound wave to be analyzed, the main frequency and target spectrum of the ambient sound wave to be analyzed are imported into the dynamic filtering module in real time for nonlinear phase compensation; The nonlinear phase compensation is to perform pre-distortion processing on the ambient sound wave to be analyzed, calculate the phase compensation value of the ambient sound wave to be analyzed, and apply the phase compensation value of the ambient sound wave to be analyzed to the target earphone in real time through the dynamic filtering module to achieve adaptive transient noise reduction filtering of the ambient sound wave to be analyzed; If it is necessary to adopt adaptive steady-state noise reduction filtering for the environmental sound wave to be analyzed, then the adaptive steady-state noise reduction filtering is adopted for the environmental sound wave to be analyzed, and the adaptive transient noise reduction filtering and the adaptive steady-state noise reduction filtering are mixed together.

5. The headphone adaptive active noise reduction method based on real-time analysis of ambient sound waves according to claim 4, characterized in that: The adaptive steady-state noise reduction filter is used for the environmental sound wave to be analyzed, and the adaptive transient noise reduction filter and the adaptive steady-state noise reduction filter are mixed, specifically: An improved FxLMS algorithm is introduced into the dynamic filtering module. The improved FxLMS algorithm is used to update the target spectrum step size of the environmental sound wave to be analyzed in the dynamic filtering module, and energy leakage of the environmental sound wave to be analyzed is controlled during the target spectrum compensation update process. The energy leakage control of the environmental sound wave to be analyzed is to control the energy of the target spectrum of the environmental sound wave to be analyzed to always be maintained within the energy threshold; In the dynamic filtering module, the distribution intervals between the positions of different short-time energy mutation sound waves are analyzed in real time, and adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering are switched in real time based on the distribution intervals; The environmental sound waves to be analyzed that have passed through adaptive transient noise reduction filtering and adaptive steady-state noise reduction filtering are synthesized in multiple frequency bands, and the environmental sound waves to be analyzed after the multi-band synthesis are calibrated as real-time noise reduction environmental sound waves.

6. The headphone adaptive active noise reduction method based on real-time analysis of ambient sound waves according to claim 1, characterized in that: During the real-time noise reduction ambient sound wave output, the target earphones are subjected to energy consumption monitoring and energy consumption optimization control, specifically: Spectral subtraction is introduced to eliminate residual noise from the real-time noise reduction environment sound waves, while obtaining the surrounding environment parameters of the target headphones; Calculate the real-time energy consumption change data of the target headphones in processing real-time noise reduction environmental sound waves and construct a real-time energy consumption change curve; Presetting a standard energy consumption threshold and analyzing the real-time energy consumption change curve to determine whether the real-time energy consumption of the target headset is always maintained within the standard energy consumption threshold; If so, calibrate the target earphones as qualified energy consumption target earphones, and maintain the qualified energy consumption target earphones in operation to achieve adaptive active noise reduction of ambient sound waves; If not, the target earphone is calibrated as an unqualified energy consumption target earphone, and the grey correlation method is introduced to calculate the grey correlation value between the real-time energy consumption of the unqualified energy consumption target earphone and the corresponding surrounding environment parameters, and calibrate it as the first grey correlation value; A gray correlation threshold is preset. If the first gray correlation value remains within the gray correlation threshold, it is determined that the real-time energy consumption of the unqualified energy consumption target earphone is affected by the corresponding surrounding environment parameter, and a big data network is introduced to retrieve a solution for protecting the unqualified energy consumption target earphone from being affected by the corresponding surrounding environment parameter and output the solution; If the first grey correlation value is not maintained within the grey correlation threshold, controlling the unqualified energy consumption target earphone to perform real-time decibel calculation of the ambient sound wave to be analyzed to determine the real-time decibel value of the ambient sound wave to be analyzed; A pre-set real-time decibel threshold for harm is set. If the real-time decibel value of the ambient sound wave to be analyzed is not greater than the real-time decibel threshold for harm, there is no need to perform adaptive active noise reduction on the unqualified energy consumption target headphones. If the real-time decibel value of the ambient sound wave to be analyzed is greater than the harmful real-time decibel threshold, the unqualified energy consumption target headphones are controlled to perform adaptive active noise reduction and output real-time noise reduction ambient sound waves.

7. Headphone adaptive active noise reduction system based on real-time analysis of ambient sound waves, characterized by: The headphone adaptive active noise reduction system includes a memory and a processor. The memory stores a headphone adaptive active noise reduction method program. When the headphone adaptive active noise reduction method program is executed by the processor, the headphone adaptive active noise reduction method steps according to any one of claims 1 to 6 are implemented.

Images