Audio equipment intelligent noise reduction method and system based on environmental noise

Abstract

The invention relates to the technical field of environmental noise monitoring and control, and discloses an intelligent noise reduction method and system for audio equipment based on environmental noise, and the method comprises the steps: obtaining a real-time original audio signal, achieving the synchronous transmission of data, and obtaining a real-time noise data stream; performing audio signal spectrum analysis according to the real-time noise data stream to obtain frequency domain distribution characteristics; according to the frequency domain distribution characteristics, noise characteristic comparison is carried out, and a noise classification result is obtained; according to the noise classification result, corresponding suppression parameters are extracted, and a preliminary suppression scheme is obtained; performing preliminary noise reduction according to the real-time noise data stream and the preliminary suppression scheme, and analyzing the noise data stream subjected to preliminary noise reduction to obtain a noise intensity real-time fluctuation trend; and according to the noise intensity real-time fluctuation trend and the preliminary suppression scheme, dynamically adjusting suppression parameters, and generating a final noise reduction control instruction. According to the method, accurate classification and dynamic suppression of noise can be realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of environmental noise monitoring and control, and particularly relates to an audio device intelligent noise reduction method and system based on environmental noise. BACKGROUND

[0002] Currently, environmental noise monitoring and control is an important issue for modern city management and quality of life improvement, especially in the fields of transportation, industry and public places, which is directly related to people's health and comfort. With the acceleration of urbanization, noise pollution has become a social problem that needs to be solved urgently, and its governance not only requires technological innovation, but also has far-reaching significance for environmental protection and public welfare.

[0003] In one prior art, the audio device noise reduction method based on environmental noise usually collects noise signals through a single device and uses a fixed noise reduction algorithm for processing. These methods usually use simple audio filtering techniques to perform spectral decomposition on the collected noise signals, and then reduce the noise intensity through pre-set filter parameters. For example, some devices will filter out noise in a specific frequency range to reduce interference in the environment. However, this method often lacks multi-device coordination and real-time data analysis capabilities when dealing with complex and variable environmental noise, making it difficult to accurately distinguish noise from different sources such as traffic noise, construction noise or human voices. In addition, since these methods rely on fixed noise reduction parameters, it is difficult to dynamically adjust strategies according to real-time changes in noise intensity, resulting in poor noise reduction effect when facing sudden noise or complex scenarios.

[0004] In summary, the prior art has the problem of being unable to achieve dynamic suppression of noise. SUMMARY

[0005] The present application provides an audio device intelligent noise reduction method and system based on environmental noise to achieve dynamic suppression of noise.

[0006] In a first aspect, to solve the above technical problems, the present application provides an audio device intelligent noise reduction method based on environmental noise, comprising: Obtaining real-time original audio signals of noise collection terminals distributed at different positions, using a pre-set communication protocol to realize data synchronization transmission, and obtaining unified real-time noise data stream; Performing audio signal spectrum analysis according to the real-time noise data stream, extracting features of each frequency band, and obtaining frequency domain distribution characteristics of noise signals; According to the frequency domain distribution characteristics and the real-time noise data stream, calling a pre-set noise type database to compare noise characteristics, and obtaining noise classification results; According to the noise classification result and the real-time noise data stream, corresponding suppression parameters are extracted from a preset suppression parameter set to obtain a preliminary suppression scheme; According to the real-time noise data stream and the preliminary suppression scheme, preliminary noise reduction is performed, and a preliminary noise reduction noise data stream is analyzed to obtain a noise intensity real-time fluctuation trend; According to the noise intensity real-time fluctuation trend and the preliminary suppression scheme, the suppression parameters are dynamically adjusted to generate a real-time updated final noise reduction control instruction to drive the terminal to perform noise reduction operation.

[0007] In an optional implementation, the real-time original audio signals of the noise collection terminals distributed in different positions are obtained, data synchronization transmission is realized by using a preset communication protocol, and a unified real-time noise data stream is obtained, including: The real-time original audio signals of the noise collection terminals distributed in different positions are obtained; According to the real-time original audio signals, time stamp marking is performed to obtain an initial audio stream with time marking; According to the initial audio stream, data synchronization is performed based on a preset communication protocol to obtain an intermediate audio stream; According to the intermediate audio stream, data format standardization conversion is performed in combination with the preset communication protocol to obtain a unified real-time noise data stream.

[0008] In an optional implementation, the real-time noise data stream is used to perform audio signal spectrum analysis, extract features of each frequency band, and obtain frequency domain distribution characteristics of noise signals, including: According to the real-time noise data stream, audio signal division is performed to extract frequency components of each frequency band to obtain a preliminary decomposition signal segment; According to the preliminary decomposition signal segment, frequency distribution analysis of each frequency band is performed to obtain characteristic identification frequency band data; According to the characteristic identification frequency band data, frequency band data refinement decomposition is performed to obtain a refined frequency band parameter set; According to the refined frequency band parameter set, frequency band feature correlation integration is performed to obtain frequency domain distribution characteristics of noise signals.

[0009] In an optional implementation, the frequency domain distribution characteristics and the real-time noise data stream are used to call a preset noise type database to perform noise feature comparison to obtain a noise classification result, including: According to the frequency domain distribution characteristics, the preset noise type database is called to compare each frequency band parameter one by one to obtain a preliminary deviation value set; The deviation values of each frequency band in the preliminary deviation value set are compared with a preset deviation value threshold, and when the deviation value of the frequency band data is less than the deviation value of the noise type characteristic data in the noise type database, it is determined that the corresponding signal belongs to the corresponding noise type, and a preliminary classification mark is obtained; According to the preliminary classification mark and the real-time noise data stream, a classification mark association is performed to obtain a noise classification result.

[0010] In an optional implementation, the preliminary suppression scheme is obtained according to the noise classification result and the real-time noise data stream from a preset suppression parameter set, including: According to the noise classification result, a corresponding suppression parameter is extracted from a preset suppression parameter set to obtain a preliminary suppression parameter set; According to the preliminary suppression parameter set and the real-time noise data stream, noise characteristic matching is performed to obtain an intermediate suppression parameter set; According to the intermediate suppression parameter set, combined with the noise classification result, an environmental noise adaptation process is performed to obtain a preliminary suppression scheme.

[0011] In an optional implementation, the noise intensity real-time fluctuation trend is obtained by performing preliminary noise reduction according to the real-time noise data stream and the preliminary suppression scheme, and analyzing the preliminary noise reduction noise data stream, including: According to the real-time noise data stream and the preliminary suppression scheme, preliminary noise reduction is performed to obtain a preliminary noise reduction noise data stream; According to the preliminary noise reduction noise data stream, noise intensity analysis is performed to obtain preliminary noise fluctuation data; According to the preliminary noise fluctuation data, frequency spectrum analysis is performed to extract dynamic fluctuation characteristics to obtain a noise intensity real-time fluctuation trend.

[0012] In an optional implementation, the noise intensity real-time fluctuation trend is obtained by performing preliminary noise reduction according to the real-time noise data stream and the preliminary suppression scheme, and analyzing the preliminary noise reduction noise data stream, including: The noise intensity real-time fluctuation trend is compared with a preset noise fluctuation trend threshold, and when the fluctuation trend is greater than a preset first noise fluctuation trend threshold, the suppression parameter in the preliminary suppression scheme is increased, and when the fluctuation trend is less than a preset second noise fluctuation trend threshold, the suppression parameter in the preliminary suppression scheme is decreased to obtain an optimized suppression scheme; The first noise fluctuation trend threshold is greater than the second noise fluctuation trend threshold; According to the optimized suppression scheme, the terminal performs an optimized noise reduction operation to obtain real-time execution feedback data containing an actual noise reduction effect; According to the real-time feedback data and the optimized suppression scheme, the actual noise reduction effect is compared with the preset target noise reduction effect, the suppression parameter is dynamically adjusted, a real-time updated final noise reduction control instruction is generated, and a terminal is driven to perform a noise reduction operation.

[0013] In a second aspect, the present application provides an audio device intelligent noise reduction device based on environmental noise, comprising: A data acquisition module is configured to acquire real-time original audio signals of noise collection terminals distributed at different positions, implement data synchronous transmission by using a preset communication protocol, and obtain unified real-time noise data streams. A spectrum analysis module is configured to perform audio signal spectrum analysis according to the real-time noise data streams, extract frequency band features, and obtain frequency domain distribution characteristics of noise signals. A noise classification module is configured to call a preset noise type database to perform noise feature comparison according to the frequency domain distribution characteristics and the real-time noise data streams, and obtain a noise classification result. A parameter extraction module is configured to extract corresponding suppression parameters from a preset suppression parameter set according to the noise classification result and the real-time noise data streams, and obtain a preliminary suppression scheme. A preliminary noise reduction module is configured to perform preliminary noise reduction according to the real-time noise data streams and the preliminary suppression scheme, analyze preliminary noise reduction noise data streams, and obtain a noise intensity real-time fluctuation trend. A result output module is configured to dynamically adjust suppression parameters according to the noise intensity real-time fluctuation trend and the preliminary suppression scheme, generate a real-time updated final noise reduction control instruction, and drive a terminal to perform a noise reduction operation.

[0014] In a third aspect, the present application further provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the audio device intelligent noise reduction method based on environmental noise according to any one of the above-mentioned embodiments when executing the computer program.

[0015] In a fourth aspect, the present application further provides a computer readable storage medium, comprising a stored computer program, wherein the computer readable storage medium controls a device where the computer readable storage medium is located to execute the audio device intelligent noise reduction method based on environmental noise according to any one of the above-mentioned embodiments when the computer program runs.

[0016] Compared with the prior art, the present application has the following beneficial effects: (1) The application obtains real-time original audio signals of noise collection terminals distributed in different positions, realizes data synchronous transmission by using a preset communication protocol, and obtains unified real-time noise data stream. By configuring a high-precision clock module in each noise collection terminal to add a timestamp mark to the real-time original audio signal, it is ensured that the audio signals collected at multiple points can be aligned according to the collection time. This time alignment mechanism effectively avoids the signal misalignment problem caused by the difference in terminal position, so that the data collected by all terminals remains consistent in the time dimension, laying a foundation for subsequent data processing, thereby significantly improving the accuracy of data synchronization. In addition, by combining the preset communication protocol for data format standardization conversion, the intermediate audio stream is integrated into a unified real-time noise data stream. This process eliminates the processing complexity caused by inconsistent formats by checking and converting the data format, ensures the uniformity and operability of the data stream, reduces the additional computational burden caused by format adjustment in the subsequent steps, and further optimizes the calculation process.

[0017] (2) According to the real-time noise data stream, the application performs audio signal spectrum analysis, extracts the characteristics of each frequency band, and obtains the frequency domain distribution characteristics of the noise signal. By dividing the audio signal, the real-time noise data stream is decomposed into multiple frequency bands and the frequency components of each frequency band are extracted to generate a preliminary decomposition signal segment. This decomposition process block processes complex audio data, reduces the data amount of a single analysis, and simplifies the complexity of subsequent feature extraction. In addition, by refining the feature identification frequency band data, a refined frequency band parameter set is obtained, and the frequency band features are associated and integrated, further enhancing the recognition accuracy of noise features. This refinement and integration process ensures the comprehensiveness and accuracy of the frequency domain distribution characteristics by analyzing the relationship between the frequency bands, providing high-quality input data for subsequent noise classification, and further improving the accuracy of the overall analysis.

[0018] (3) According to the frequency domain distribution characteristics and the real-time noise data stream, the application calls a preset noise type database to compare noise features and obtains a noise classification result. The preset noise type database is called to compare each frequency band parameter in the frequency domain distribution characteristics one by one to generate a preliminary deviation value set. This comparison process can quickly filter out possible noise types by directly matching the feature values in the database, reducing the number of repeated calculations and invalid comparisons, thereby significantly improving the calculation efficiency of the comparison process. In addition, by associating the preliminary classification mark with the real-time noise data stream, the final noise classification result is generated. This association process integrates the mark and data stream information to ensure the accuracy and consistency of the classification result, reducing the possibility of classification errors, thereby enhancing the accuracy of classification.

[0019] (4) According to the noise classification result and the real-time noise data stream, corresponding suppression parameters are extracted from a preset suppression parameter set to obtain a preliminary suppression scheme. According to the noise classification result, corresponding suppression parameters are extracted from a preset suppression parameter set to generate a preliminary suppression parameter set. This process can quickly lock the parameter range related to the noise type by directly matching the classification result and the parameter set, thereby reducing the time for blind search in a large parameter set, and significantly improving the calculation efficiency of parameter extraction. In addition, according to the intermediate suppression parameter set and the noise classification result, environmental noise adaptation processing is performed to generate a preliminary suppression scheme, and the matching degree of the scheme is ensured through structured recording and demand judgment. This adaptation and judgment process refines the relevance of parameters and classification results to ensure that the suppression scheme can accurately adapt to the current noise environment, thereby improving the accuracy and effectiveness of the scheme.

[0020] (5) According to the real-time noise data stream and the preliminary suppression scheme, preliminary noise reduction is performed, and preliminary noise data stream is analyzed to obtain a noise intensity real-time fluctuation trend. The real-time noise data stream is subjected to preliminary noise reduction in combination with the preliminary suppression scheme to generate a preliminary noise reduction noise data stream. This process directly executes the noise reduction operation using the generated suppression scheme, avoiding repeated design of complex noise reduction algorithms, simplifying the step-by-step operation of noise processing, and significantly improving the calculation efficiency. In addition, by performing frequency spectrum analysis on the preliminary noise fluctuation data, dynamic fluctuation characteristics are extracted to generate a noise intensity real-time fluctuation trend. This extraction process ensures the accuracy and comprehensiveness of the fluctuation characteristics through frequency spectrum analysis, which can truly reflect the noise change trend after noise reduction, thereby improving the accuracy of the fluctuation trend analysis.

[0021] (6) According to the noise intensity real-time fluctuation trend and the preliminary suppression scheme, the suppression parameters are dynamically adjusted to generate a real-time updated final noise reduction control instruction to drive the terminal to execute the noise reduction operation. First, the noise intensity real-time fluctuation trend is compared with a preset noise fluctuation trend threshold, and when the fluctuation trend is greater than a preset first noise fluctuation trend threshold, the suppression parameters are increased, and when the fluctuation trend is less than a preset second noise fluctuation trend threshold, the suppression parameters are decreased to generate an optimized suppression scheme. This comparison and adjustment process quickly determines the adjustment direction through the preset threshold, avoiding complex parameter optimization iteration and reducing unnecessary calculation steps, thereby significantly improving the calculation efficiency of the adjustment process. In addition, by comparing the actual noise reduction effect in the real-time execution feedback data with the preset target noise reduction effect, the suppression parameters are dynamically adjusted to generate a real-time updated final noise reduction control instruction to drive the terminal to execute the noise reduction operation. This comparison and adjustment process analyzes the deviation between the actual and expected values to ensure that the control instruction can accurately adapt to changes in the current noise environment, thereby enhancing the overall noise reduction effect. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 is a flow chart of an audio device intelligent noise reduction method based on environmental noise provided by the first embodiment of the present application; Figure 2 is a structural schematic diagram of an audio device intelligent noise reduction system / device based on environmental noise provided by the second embodiment of the present application. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0024] With reference to Figure 1 The first embodiment of the present application provides an audio device intelligent noise reduction method based on environmental noise, comprising the following steps: S11, obtaining real-time original audio signals of noise collection terminals distributed at different positions, realizing data synchronous transmission by using a preset communication protocol, and obtaining a unified real-time noise data stream; S12, performing audio signal spectrum analysis according to the real-time noise data stream, extracting features of each frequency band, and obtaining frequency domain distribution characteristics of noise signals; S13, calling a preset noise type database to perform noise feature comparison according to the frequency domain distribution characteristics and the real-time noise data stream, and obtaining a noise classification result; S14, extracting corresponding suppression parameters from a preset suppression parameter set according to the noise classification result and the real-time noise data stream, and obtaining a preliminary suppression scheme; S15, performing preliminary noise reduction according to the real-time noise data stream and the preliminary suppression scheme, analyzing a preliminary noise reduction noise data stream, and obtaining a real-time fluctuation trend of noise intensity; S16, dynamically adjusting suppression parameters according to the real-time fluctuation trend of noise intensity and the preliminary suppression scheme, generating a real-time updated final noise reduction control instruction, and driving a terminal to perform a noise reduction operation.

[0025] In step S11, real-time original audio signals of noise collection terminals distributed at different positions are obtained, data synchronous transmission is realized by using a preset communication protocol, and a unified real-time noise data stream is obtained.

[0026] In one implementation mode, obtaining real-time original audio signals of noise collection terminals distributed at different positions, realizing data synchronous transmission by using a preset communication protocol, and obtaining a unified real-time noise data stream, comprises: The real-time original audio signals of the noise collection terminals distributed at different positions are acquired; time stamp labels are added to the real-time original audio signals to obtain initial audio streams with time labels; data synchronization is performed on the initial audio streams based on a preset communication protocol to obtain intermediate audio streams; and data format standardization conversion is performed on the intermediate audio streams based on the preset communication protocol to obtain unified real-time noise data streams.

[0027] It should be noted that, first, a plurality of noise collection terminals are deployed in different areas in a city or other application scenarios, and high-precision microphones and sensors on the terminals are used to collect sound signals in the environment in real time to acquire real-time original audio signals of the noise collection terminals distributed at different positions. These signals record the sound wave changes of the environment in the form of digital waveforms, such as traffic noise or construction noise. Then, according to the real-time original audio signals collected in real time, a time stamp label is added to each frame of audio data on each terminal through a built-in high-precision clock module to generate initial audio streams with time labels. This time stamp label ensures that the clocks of all terminals are synchronized through the network time protocol, thereby ensuring the consistency of audio data collected at different positions in the time dimension. Then, according to the initial audio streams, a preset communication protocol based on TCP / IP is used to perform data synchronization through a time stamp comparison and delay compensation mechanism. The specific method is to compare the time stamps of each terminal. If the signal of a certain terminal does not arrive in time due to network delay, the data is temporarily stored in a buffer. After all terminal data is aligned, an intermediate audio stream is generated. This process ensures the time consistency of the data. The preset communication protocol based on TCP / IP specifically uses an enhanced TCP protocol combined with the real-time transport protocol (RTP) to ensure reliable transmission of data packets and timing integrity. The TCP provides a reliable data delivery mechanism, and the RTP supports real-time synchronization processing of audio streams through time stamps and sequence numbers. Then, according to the intermediate audio stream, the data stream is processed through a format standardization conversion tool combined with the enhanced TCP protocol. Specifically, the audio data is uniformly converted to a standard sampling rate and bit depth format, such as through a resampling algorithm to uniformly convert audio data of different sampling rates to a standard format, while the data is segmented and recombined to ensure continuity, and finally a unified real-time noise data stream is generated. The format standardization conversion tool specifically uses the FFmpeg tool, which adjusts the audio data collected by different terminals to a standard WAV format with a sampling rate of 44.1 kHz and a bit depth of 16 bits through built-in resampling and format conversion functions, and integrates the data stream in chronological order through a segmentation and recombination algorithm. The real-time noise data stream refers to the audio data stream integrated through the above steps, which has time consistency and format uniformity, and contains environmental noise information collected at different positions, which can be used for subsequent steps such as frequency spectrum analysis, noise classification, and noise reduction processing. Through real-time transmission and processing, it provides a reliable data basis for intelligent noise reduction.

[0028] In step S12, audio signal spectrum analysis is performed according to the real-time noise data stream, and features of each frequency band are extracted to obtain frequency domain distribution characteristics of the noise signal.

[0029] In an implementation, audio signal spectrum analysis is performed according to the real-time noise data stream, and features of each frequency band are extracted to obtain frequency domain distribution characteristics of the noise signal, including: According to the real-time noise data stream, audio signal division is performed to extract frequency components of each frequency band to obtain a preliminary decomposed signal segment; according to the preliminary decomposed signal segment, frequency distribution analysis of each frequency band is performed to obtain characteristic identification frequency band data; according to the characteristic identification frequency band data, frequency band data refinement decomposition is performed to obtain a refined frequency band parameter set; and according to the refined frequency band parameter set, frequency band feature correlation integration is performed to obtain frequency domain distribution characteristics of the noise signal.

[0030] It should be noted that according to the real-time noise data stream, the audio signal is analyzed by using the fast Fourier transform algorithm to divide it into multiple frequency bands, such as low, medium and high frequency ranges, and the time domain signal is converted into the frequency domain signal by Fourier transform, the frequency components of each frequency band are extracted, and the preliminary decomposition signal segment is generated, which reflects the distribution of noise in different frequency ranges. Then, according to the preliminary decomposition signal segment, the power spectral density analysis method is used to calculate the frequency distribution and intensity of each frequency band, and the characteristic parameters such as the peak frequency and average intensity of each frequency band are determined by calculating the energy distribution of each frequency band, and the characteristic identification frequency band data is generated, which adds characteristic identification to each frequency band for subsequent classification and processing. The power spectral density analysis method is implemented by using the Welch method, which first divides the preliminary decomposition signal segment into several overlapping subsegments, such as 1024 sampling points per segment, and applies a Hanning window function to each subsegment for windowing processing to reduce spectral leakage. Then, the frequency domain representation of each subsegment is calculated by fast Fourier transform, the power of each frequency component is calculated, and then the average value of all subsegment power values is taken to generate the power spectral density curve of each frequency band. Then, the curve is analyzed to find the maximum value of the energy distribution as the peak frequency, and the average value of all power values in the frequency range is calculated as the average intensity, and finally the characteristic identification frequency band data is generated, such as labeling the peak frequency and average intensity value of a certain frequency band, for subsequent noise classification and processing. Then, according to the characteristic identification frequency band data, the frequency band is further decomposed by using the discrete Fourier transform method to further split it into smaller subbands, and more detailed frequency characteristics are extracted to generate a set of refined frequency band parameters, which improves the analysis accuracy of noise characteristics. Then, according to the set of refined frequency band parameters, the correlation analysis method is used to integrate the feature parameters of each frequency band by calculating the correlation coefficient between the frequency bands to analyze their correlation, such as determining whether the low frequency and medium frequency are affected by the same noise source, to generate the frequency domain distribution characteristics of the noise signal. The correlation analysis method specifically uses the Pearson correlation coefficient method, which first calculates the feature parameters of each pair of frequency bands one by one, specifically by comparing the covariance between the two sets of data divided by the product of the two standard deviations to obtain the Pearson correlation coefficient, the coefficient value ranges from -1 to 1, where close to 1 indicates positive correlation, close to -1 indicates negative correlation, and close to 0 indicates no correlation. For example, if the intensity change trend of the low frequency band and the medium frequency band is consistent, the correlation coefficient is close to 1, indicating that both may be affected by the same noise source, such as traffic noise. Then, according to these correlation coefficients, the correlation between the frequency bands is sorted and integrated to generate the frequency domain distribution characteristics of the noise signal, reflecting the mutual influence and overall distribution characteristics of each frequency band.The frequency domain distribution characteristic refers to the characteristic distribution of the noise signal obtained through the above steps in the frequency domain, contains the frequency, intensity and correlation information of each frequency band, is obtained through spectrum analysis and characteristic integration, can be used for subsequent calling of a noise type database for classification, and further provides a basis for formulating a noise reduction scheme.

[0031] In step S13, a preset noise type database needs to be called according to the frequency domain distribution characteristic and the real-time noise data stream for noise characteristic comparison, so as to obtain a noise classification result.

[0032] In an implementation manner, according to the frequency domain distribution characteristic and the real-time noise data stream, a preset noise type database is called for noise characteristic comparison, so as to obtain a noise classification result, which includes: According to the frequency domain distribution characteristic, a preset noise type database is called to compare each frequency band parameter one by one, so as to obtain a preliminary deviation value set; the deviation value of each frequency band in the preliminary deviation value set is compared with a preset deviation value threshold; when the deviation value of a certain frequency band data and the noise characteristic data of a certain noise type in the noise type database is less than the preset deviation value threshold, it is determined that the corresponding signal belongs to the corresponding noise type, and a preliminary classification mark is obtained; according to the preliminary classification mark and the real-time noise data stream, classification mark correlation is performed, so as to obtain a noise classification result.

[0033] It should be noted that first, according to the frequency domain distribution characteristics, the characteristic value records corresponding to the frequency band parameters are extracted from the preset noise type database, and the Euclidean distance calculation method is used to compare the frequency and intensity parameters of each frequency band with the noise type characteristic values stored in the database. Each frequency band parameter is compared with the corresponding reference characteristic value, the sum of the squares of the differences in frequency and intensity dimensions is calculated, and then the square root of the sum is taken to obtain the Euclidean distance value. This value represents the geometric distance between two characteristic points. For each noise type, repeat this calculation process to generate a set of distance values corresponding to each frequency band and all noise types in the database, and sort these distance values by size and record them as a preliminary deviation value set. This set contains the matching degree between each frequency band and each noise type. Then, compare the deviation values of each frequency band in the preliminary deviation value set with the preset deviation value threshold. When the deviation value of a certain frequency band data and the noise type characteristic data in the noise type database is less than the preset deviation value threshold, it indicates that the characteristics of the frequency band are highly similar to the noise type characteristic data in the database, and then the signal belongs to the corresponding noise type, such as traffic noise or construction noise, and a preliminary classification label is generated. Then, according to the preliminary classification label and the real-time noise data stream, the label is time series associated with the timestamp and frequency band information in the data stream, and the label is checked for consistency with the time period of the data stream. If consistent, integrate into the final result, if not, adjust the label to ensure the accuracy of the classification, and obtain the noise classification result. The noise classification result refers to the structured data obtained by the above steps, indicating the type of noise signal, containing the noise type label corresponding to each frequency band, which can be used to extract the adaptive suppression parameter from the suppression parameter set, and provide specific guidance for the development of the noise reduction scheme.

[0034] In step S14, the corresponding suppression parameter is extracted from the preset suppression parameter set according to the noise classification result and the real-time noise data stream to obtain a preliminary suppression scheme.

[0035] In one implementation, the corresponding suppression parameter is extracted from the preset suppression parameter set according to the noise classification result and the real-time noise data stream to obtain a preliminary suppression scheme, including: According to the noise classification result, the corresponding suppression parameter is extracted from the preset suppression parameter set to obtain a preliminary suppression parameter set. According to the preliminary suppression parameter set and the real-time noise data stream, the noise characteristics are matched to obtain an intermediate suppression parameter set. According to the intermediate suppression parameter set, the noise classification result is combined to perform environmental noise adaptation processing to obtain a preliminary suppression scheme.

[0036] It should be noted that according to the noise classification result, the suppression parameters corresponding to each noise type are extracted from the preset suppression parameter set, for example, a specific noise reduction coefficient is extracted for traffic noise, to generate a preliminary suppression parameter set, which contains multiple possible parameter values. Then, according to the preliminary suppression parameter set and the real-time noise data stream, a correlation analysis method is used for noise characteristic matching, specifically, each suppression parameter in the parameter set is compared with the frequency and intensity characteristics in the data stream one by one, and the parameter closest to the current noise characteristic is selected to generate an intermediate suppression parameter set. This process ensures the preliminary adaptability of the parameters by calculating the similarity between the characteristics, and the correlation analysis method specifically uses the Spearman correlation coefficient method. First, the frequency and intensity characteristic data are extracted from the real-time noise data stream, and the corresponding parameter characteristics are obtained from the preliminary suppression parameter set, for example, the mapping value of the noise reduction coefficient and the frequency range. Then, sort each group of characteristic data, convert the original data to rank data, and then calculate the Spearman correlation coefficient. Specifically, by comparing the sum of the squares of the differences between the two groups of ranks with the number of data points, the coefficient value is obtained, which ranges from -1 to 1, where 1 indicates strong positive correlation, -1 indicates strong negative correlation, and 0 indicates no correlation. For example, if the frequency range rank of a certain suppression parameter is highly consistent with the frequency characteristic rank of the data stream, and the coefficient value is close to 1, it is considered that the parameter matches the current noise characteristic. Then, according to these correlation coefficients, the parameters are sorted, and the parameter with the highest correlation is selected to form the intermediate suppression parameter set, ensuring the preliminary adaptability of the parameters to the noise characteristics. Then, according to the intermediate suppression parameter set, combined with the noise classification result, the weighted average method is used for environmental noise adaptation processing, specifically, the parameter value is adjusted according to the weight of different noise types, for example, the corresponding parameter is appropriately enhanced when the weight of traffic noise is high, and finally the preliminary suppression scheme is generated. The preliminary suppression scheme refers to the initial noise reduction strategy obtained through the above steps, which contains an optimized parameter set for the current noise classification result and real-time data stream, and can be used for subsequent preliminary noise reduction operation to provide basic data support for dynamic adjustment.

[0037] In step S15, preliminary noise reduction is performed according to the real-time noise data stream and the preliminary suppression scheme, and the preliminary noise reduction noise data stream is analyzed to obtain the real-time fluctuation trend of noise intensity.

[0038] In one implementation, preliminary noise reduction is performed according to the real-time noise data stream and the preliminary suppression scheme, and the preliminary noise reduction noise data stream is analyzed to obtain the real-time fluctuation trend of noise intensity, including: According to the real-time noise data stream and the preliminary suppression scheme, preliminary noise reduction is performed to obtain a preliminary noise reduction data stream; according to the preliminary noise reduction data stream, noise intensity analysis is performed to obtain preliminary noise fluctuation data; and according to the preliminary noise fluctuation data, frequency spectrum analysis is performed to extract dynamic fluctuation characteristics and obtain a noise intensity real-time fluctuation trend.

[0039] It should be noted that, according to the real-time noise data stream and the preliminary suppression scheme, an adaptive filter algorithm is used to perform preliminary noise reduction on the audio signal, specifically by applying the parameters in the preliminary suppression scheme to attenuate the high-intensity noise frequency band in the data stream to generate a preliminary noise reduction data stream, which reflects the changes in the ambient sound after preliminary noise reduction. Then, according to the preliminary noise reduction data stream, a time domain analysis method is used to calculate the sound pressure level change of each time period to identify the fluctuations in noise intensity and generate preliminary noise fluctuation data. The time domain analysis method specifically uses a sliding window average method to achieve this step. First, the preliminary noise reduction data stream is divided into several fixed time periods, for example, each lasting 1 second, and a sliding window is used to cover each time period, with a window length of 0.1 seconds and a step length of 0.01 seconds. Then, the instantaneous sound pressure level of the audio samples in each window is calculated, specifically by taking the square average of the sample amplitude, calculating the logarithmic value and converting it to decibel units to obtain the sound pressure level value of the window. After that, the sound pressure level values of adjacent windows are compared to calculate the difference between the front and back windows, and the absolute value of the difference is the sound pressure level change, which reflects the fluctuations in noise intensity, for example, from 50 decibels to 55 decibels. Finally, all the sound pressure level change data of the time periods are integrated to generate preliminary noise fluctuation data, which records the fluctuation range of the noise intensity after noise reduction and is used for subsequent frequency spectrum analysis. Then, according to the preliminary noise fluctuation data, a fast Fourier transform algorithm is used to perform frequency spectrum analysis to convert the fluctuation data into frequency domain information, extract dynamic fluctuation characteristics such as peak frequency and change trend, and generate a noise intensity real-time fluctuation trend. The noise intensity real-time fluctuation trend reflects the dynamic characteristics of the noise intensity after noise reduction with respect to time and frequency, and can be used for subsequent dynamic adjustment of the suppression parameters to provide real-time basis for generating the final noise reduction control instruction.

[0040] In step S16, the noise intensity real-time fluctuation trend and the preliminary suppression scheme are used to dynamically adjust the suppression parameters to generate real-time updated final noise reduction control instructions to drive the terminal to perform noise reduction operations.

[0041] In one implementation, the noise intensity real-time fluctuation trend and the preliminary suppression scheme are used to dynamically adjust the suppression parameters to generate real-time updated final noise reduction control instructions to drive the terminal to perform noise reduction operations, including: The noise intensity real-time fluctuation trend is compared with a preset noise fluctuation trend threshold, when the fluctuation trend is greater than a preset first noise fluctuation trend threshold, the suppression parameter in the preliminary suppression scheme is increased, when the fluctuation trend is less than a preset second noise fluctuation trend threshold, the suppression parameter in the preliminary suppression scheme is decreased, to obtain an optimized suppression scheme; the first noise fluctuation trend threshold is greater than the second noise fluctuation trend threshold; according to the optimized suppression scheme, a terminal is driven to perform an optimized noise reduction operation to obtain real-time execution feedback data containing an actual noise reduction effect; according to the real-time execution feedback data and the optimized suppression scheme, the actual noise reduction effect is compared with a preset target noise reduction effect, the suppression parameter is dynamically adjusted, a real-time updated final noise reduction control instruction is generated, and the terminal is driven to perform a noise reduction operation.

[0042] It should be noted that first, the noise intensity real-time fluctuation trend is compared with a preset noise fluctuation trend threshold, specifically by extracting the intensity change slope of the fluctuation trend, if the slope is greater than a preset first noise fluctuation trend threshold, the suppression parameter in the preliminary suppression scheme is increased by a linear adjustment method, for example, the noise reduction coefficient is increased, when the fluctuation trend is less than the preset first noise fluctuation trend threshold and greater than a preset second noise fluctuation trend threshold, the suppression parameter remains unchanged, maintaining the current preliminary suppression scheme settings, to avoid unnecessary adjustments and ensure smooth operation of the system in a stable noise environment, if the slope is less than the preset second noise fluctuation trend threshold, the suppression parameter is decreased by the same method, to generate an optimized suppression scheme. Next, it is clear that the first noise fluctuation trend threshold is greater than the second noise fluctuation trend threshold, a dynamic adjustment interval is formed by setting two different levels of thresholds to ensure the rationality and sensitivity of the adjustment, the first threshold corresponds to high-intensity fluctuations, and the second threshold corresponds to low-intensity fluctuations. Then, according to the optimized suppression scheme, the adjusted parameters are sent to the execution unit of each terminal using a signal transmission protocol, the terminal is driven to perform an optimized noise reduction operation through an adaptive filtering algorithm, and the environment data after noise reduction is collected in real time through a sensor to generate real-time execution feedback data containing the actual noise reduction effect. Next, according to the real-time execution feedback data and the optimized suppression scheme, the actual noise reduction effect is compared with a preset target noise reduction effect, specifically, the difference between the intensity value after actual noise reduction and the target value is calculated, if the difference exceeds a preset range, the suppression parameter is dynamically adjusted by a gradient descent algorithm, a real-time updated final noise reduction control instruction is generated, and the terminal is driven to perform a noise reduction operation again. The final noise reduction control instruction refers to the noise reduction operation instruction generated by the above steps, which is optimized in real time, contains the suppression parameter adapted to the current noise environment, and is obtained through fluctuation trend comparison, parameter adjustment and feedback analysis. It can be used to drive the terminal to perform a final noise reduction operation to ensure that the noise reduction effect reaches the expected target.

[0043] In summary, the present application ensures the consistency and real-time nature of the data by collecting real-time raw audio signals using a multi-device cooperative network and performing synchronization and standardization processing, providing a reliable foundation for subsequent analysis. Next, by extracting features of each frequency band through spectral analysis and generating frequency domain distribution characteristics, the frequency distribution of noise is accurately described. Then, by calling a noise type database for feature comparison and combining real-time data streams to generate classification results, different types of noise such as traffic noise or construction noise can be accurately distinguished. Based on the classification results, adaptive suppression parameters are extracted and a preliminary suppression scheme is generated, providing targeted strategies for noise reduction. Further, by preliminary noise reduction and analysis of noise intensity fluctuation trends, the dynamic variation of noise is mastered. Finally, by dynamically adjusting the suppression parameters and generating real-time updated control instructions, the noise reduction operation can adapt to noise changes, achieving accurate classification and dynamic suppression of noise.

[0044] Reference Figure 2 The second embodiment of the present application provides an audio device intelligent noise reduction device based on environmental noise, comprising: A data acquisition module is configured to acquire real-time raw audio signals of noise collection terminals distributed at different positions, and to realize synchronous data transmission using a preset communication protocol to obtain unified real-time noise data streams. A spectral analysis module is configured to perform audio signal spectral analysis based on the real-time noise data streams, extract features of each frequency band, and obtain frequency domain distribution characteristics of noise signals. A noise classification module is configured to call a preset noise type database for noise feature comparison based on the frequency domain distribution characteristics and the real-time noise data streams, and obtain noise classification results. A parameter extraction module is configured to extract corresponding suppression parameters from a preset suppression parameter set based on the noise classification results and the real-time noise data streams, and obtain a preliminary suppression scheme. A preliminary noise reduction module is configured to perform preliminary noise reduction based on the real-time noise data streams and the preliminary suppression scheme, analyze the preliminary noise reduction noise data streams, and obtain a real-time fluctuation trend of noise intensity. A result output module is configured to dynamically adjust the suppression parameters based on the real-time fluctuation trend of noise intensity and the preliminary suppression scheme, generate real-time updated final noise reduction control instructions, and drive the terminal to perform noise reduction operations.

[0045] It should be noted that the audio device intelligent noise reduction device based on environmental noise provided by the embodiments of the present application is used to perform all process steps of the audio device intelligent noise reduction method based on environmental noise described above. The working principles and beneficial effects of the two are one-to-one correspondence, and thus will not be repeated.

[0046] The embodiment of the present application further provides an electronic device. The electronic device comprises a processor, a memory, and a computer program stored in the memory and executable on the processor, for example, an ambient noise based audio device intelligent noise reduction program. The processor implements the steps in each of the above ambient noise based audio device intelligent noise reduction method embodiments when executing the computer program, for example Figure 1 The processor implements the functions of each module / unit in each of the above device embodiments when executing the computer program, for example, a noise classification module.

[0047] For example, the computer program can be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present application. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the electronic device.

[0048] The electronic device can be a desktop computer, a notebook computer, a palm computer, a smart tablet and the like. The electronic device can comprise, but is not limited to, a processor, a memory. Those skilled in the art can understand that the above components are only examples of the electronic device and do not constitute a limitation on the electronic device, and the electronic device can comprise more or fewer components than the above, or combine certain components, or different components, for example, the electronic device can further comprise an input / output device, a network access device, a bus and the like.

[0049] The processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor and the like. The processor is the control center of the electronic device, and connects all parts of the electronic device through various interfaces and lines.

[0050] The memory can be used to store the computer program and / or modules, and the processor realizes various functions of the electronic device by running or executing the computer program and / or modules stored in the memory, and calling data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required by a function (such as a sound playing function, an image playing function, etc.), and the like; and the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), and the like. In addition, the memory can include a high-speed random access memory, and can also include a nonvolatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory devices.

[0051] The modules / units integrated in the electronic device, if realized in the form of software function units and sold or used as independent products, can be stored in a computer readable storage medium. Based on this understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. When the processor executes the computer program, the steps of the above-mentioned various method embodiments can be realized. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms, etc. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the contents included in the computer readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0052] It should be noted that the apparatus embodiments described above are merely illustrative, and the units described as separate units can or can not be physically separate, and the units displayed as units can or can not be physical units, i.e. can be located in one place, or can be distributed to multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment scheme according to actual needs. In addition, the connection relationship between the modules in the apparatus embodiment provided by the present application indicates that there is a communication connection between them, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0053] The above specific embodiments further illustrate the purpose, technical scheme and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not intended to limit the protection scope of the present application. It is particularly pointed out that any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. An ambient noise based intelligent noise reduction method for an audio device, the method comprising: The method comprises the following steps: acquiring real-time original audio signals of noise collection terminals distributed at different positions, realizing synchronous transmission of data by using a preset communication protocol, and obtaining a unified real-time noise data stream; performing audio signal spectrum analysis according to the real-time noise data stream, extracting features of each frequency band, and obtaining frequency domain distribution characteristics of noise signals; calling a preset noise type database to perform noise feature comparison according to the frequency domain distribution characteristics and the real-time noise data stream, and obtaining a noise classification result; extracting corresponding suppression parameters from a preset suppression parameter set according to the noise classification result and the real-time noise data stream, and obtaining a preliminary suppression scheme; performing preliminary noise reduction according to the real-time noise data stream and the preliminary suppression scheme, analyzing a preliminary noise reduction noise data stream, and obtaining a real-time fluctuation trend of noise intensity; dynamically adjusting suppression parameters according to the real-time fluctuation trend of noise intensity and the preliminary suppression scheme, generating a real-time updated final noise reduction control instruction, and driving a terminal to perform a noise reduction operation.

2. The environmental noise-based audio device intelligent noise reduction method of claim 1, wherein, The method of acquiring real-time original audio signals of noise collection terminals distributed at different positions, realizing synchronous transmission of data by using a preset communication protocol, and obtaining a unified real-time noise data stream comprises the following steps: acquiring real-time original audio signals of noise collection terminals distributed at different positions; performing time stamp marking according to the real-time original audio signals, and obtaining an initial audio stream with time marking; performing data synchronization based on a preset communication protocol according to the initial audio stream, and obtaining an intermediate audio stream; performing data format standardization conversion in combination with a preset communication protocol according to the intermediate audio stream, and obtaining a unified real-time noise data stream.

3. The environmental noise-based audio device intelligent noise reduction method of claim 1, wherein, The method of performing audio signal spectrum analysis according to the real-time noise data stream, extracting features of each frequency band, and obtaining frequency domain distribution characteristics of noise signals comprises the following steps: performing audio signal division according to the real-time noise data stream, extracting frequency components of each frequency band, and obtaining a preliminary decomposition signal segment; performing frequency distribution analysis of each frequency band according to the preliminary decomposition signal segment, and obtaining characteristic identification frequency band data; performing frequency band data refinement decomposition according to the characteristic identification frequency band data, and obtaining a refined frequency band parameter set; performing frequency band feature correlation integration according to the refined frequency band parameter set, and obtaining frequency domain distribution characteristics of noise signals.

4. The environmental noise-based audio device intelligent noise reduction method of claim 1, wherein, The method of calling a preset noise type database to perform noise feature comparison according to the frequency domain distribution characteristics and the real-time noise data stream, and obtaining a noise classification result comprises the following steps: calling a preset noise type database to perform one-by-one comparison of each frequency band parameter according to the frequency domain distribution characteristics, and obtaining a preliminary deviation value set; comparing deviation values of each frequency band in the preliminary deviation value set with a preset deviation value threshold, and when a deviation value of certain frequency band data is less than a deviation value of certain noise type feature data in the noise type database, it is determined that the corresponding signal belongs to the corresponding noise type, and a preliminary classification mark is obtained; performing classification mark correlation according to the preliminary classification mark and the real-time noise data stream, and obtaining a noise classification result.

5. The environmental noise-based audio device intelligent noise reduction method of claim 1, wherein, The preliminary suppression scheme is obtained by extracting corresponding suppression parameters from a preset suppression parameter set according to the noise classification result and the real-time noise data stream. The preliminary suppression parameter set is obtained by extracting corresponding suppression parameters from a preset suppression parameter set according to the noise classification result. The intermediate suppression parameter set is obtained by performing noise characteristic matching according to the preliminary suppression parameter set and the real-time noise data stream. The preliminary suppression scheme is obtained by performing environmental noise adaptation processing according to the intermediate suppression parameter set and the noise classification result.

6. The environmental noise-based audio device intelligent noise reduction method of claim 1, wherein, The noise intensity real-time fluctuation trend is obtained by performing preliminary noise reduction according to the real-time noise data stream and the preliminary suppression scheme, and analyzing a preliminary noise reduction noise data stream. The preliminary noise reduction noise data stream is obtained by performing preliminary noise reduction according to the real-time noise data stream and the preliminary suppression scheme. The preliminary noise fluctuation data is obtained by performing noise intensity analysis according to the preliminary noise reduction noise data stream. The noise intensity real-time fluctuation trend is obtained by performing frequency spectrum analysis according to the preliminary noise fluctuation data, and extracting dynamic fluctuation characteristics.

7. The environmental noise-based audio device intelligent noise reduction method of claim 1, wherein, The noise intensity real-time fluctuation trend is obtained by performing preliminary noise reduction according to the real-time noise data stream and the preliminary suppression scheme, and analyzing a preliminary noise reduction noise data stream. The noise intensity real-time fluctuation trend is compared with a preset noise fluctuation trend threshold value, when the fluctuation trend is greater than a preset first noise fluctuation trend threshold value, the suppression parameters in the preliminary suppression scheme are increased, when the fluctuation trend is less than a preset second noise fluctuation trend threshold value, the suppression parameters in the preliminary suppression scheme are reduced, and an optimized suppression scheme is obtained. The first noise fluctuation trend threshold value is greater than the second noise fluctuation trend threshold value. The terminal is driven to perform an optimized noise reduction operation according to the optimized suppression scheme, and real-time execution feedback data containing an actual noise reduction effect is obtained. The actual noise reduction effect is compared with a preset target noise reduction effect according to the real-time execution feedback data and the optimized suppression scheme, the suppression parameters are dynamically adjusted, the real-time updated final noise reduction control instruction is generated, and the terminal is driven to perform the noise reduction operation.

8. An ambient noise based intelligent noise reduction system for an audio device, the system comprising: The data acquisition module is configured to acquire real-time original audio signals of noise collection terminals distributed at different positions, and to realize data synchronous transmission by using a preset communication protocol, thereby obtaining a unified real-time noise data stream. The spectrum analysis module is configured to perform audio signal frequency spectrum analysis according to the real-time noise data stream, extract frequency band characteristics, and obtain frequency domain distribution characteristics of noise signals. The noise classification module is configured to call a preset noise type database to perform noise characteristic comparison according to the frequency domain distribution characteristics and the real-time noise data stream, and obtain a noise classification result. The parameter extraction module is configured to extract corresponding suppression parameters from a preset suppression parameter set according to the noise classification result and the real-time noise data stream, and obtain a preliminary suppression scheme. The preliminary noise reduction module is configured to perform preliminary noise reduction according to the real-time noise data stream and the preliminary suppression scheme, analyze a preliminary noise reduction noise data stream, and obtain a noise intensity real-time fluctuation trend. ​ The result output module is configured to dynamically adjust the suppression parameter according to the real-time fluctuation trend of the noise intensity and the preliminary suppression scheme, generate a real-time updated final noise reduction control instruction, and drive the terminal to perform noise reduction operation.

9. An electronic device, comprising: The computer readable storage medium comprises a stored computer program, wherein the computer program, when executed, controls a device in which the computer readable storage medium is located to perform the environment noise based intelligent noise reduction method of the audio device according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium comprises a stored computer program, wherein the computer program, when executed, controls a device in which the computer readable storage medium is located to perform the environment noise based intelligent noise reduction method of the audio device according to any one of claims 1 to 7.

Images