Fast analysis method for fixed-frequency audio

Abstract

The invention relates to the technical field of audio analysis, in particular to a fixed-frequency audio quick analysis method, which comprises the following steps: extracting an audio to be analyzed; performing audio analysis on the extracted to-be-analyzed audio, preparing an audio sample, and preprocessing the audio sample; performing fast Fourier transform on the preprocessed audio sample and calculating frequency domain energy; calculating a frequency point interval when the energy is maximum according to the sample frequency; traversing the frequency domain energy, and judging whether a packet loss problem exists in the audio sample or not. And if the calculated frequency domain energy value before the frequency point interval exceeds a preset threshold value, determining that the audio sample has a packet loss problem. According to the invention, the method does not depend on manual intervention, and ensures high-precision and high-efficiency audio anomaly analysis.

Description

Technical Field

[0001] The present application relates to the technical field of audio analysis, and in particular to a method for fast analysis of fixed-frequency audio. Background Art

[0002] With the rapid development of audio technology and digital signal processing technology, audio data analysis has become an important part of speech recognition, audio monitoring, music analysis and other fields. Through in-depth analysis of audio signals, key information can be effectively extracted to achieve precise audio processing requirements in various application scenarios, which has broad application prospects.

[0003] Existing audio analysis technologies usually rely on professional tools or software to convert audio signals into spectrograms, and identify abnormal points or specific features in the audio through manual observation and analysis of the spectrograms. For example, a common method is to use specific tools to convert the format of audio files, generate spectrograms, and then locate the time-frequency relationship of the audio. Some tools can also extract audio features through basic algorithms to assist manual analysis.

[0004] However, existing technologies mainly rely on manual operations and cannot meet the needs of efficiently analyzing multiple audio segments and quickly locating audio anomalies. When the amount of audio data is large, manual detection is inefficient and it is difficult to quickly output complete analysis results. Summary of the invention

[0005] This application provides a method for fast analysis of fixed frequency audio, which can ensure high-precision and high-efficiency audio anomaly analysis without relying on manual intervention. This application provides the following technical solutions:

[0006] In a first aspect, the present application provides a method for fast parsing of fixed-frequency audio, the method comprising:

[0007] Extract the audio to be parsed;

[0008] Perform audio analysis on the extracted audio to be analyzed, prepare audio samples and pre-process them;

[0009] Perform fast Fourier transform on the preprocessed audio samples and calculate the frequency domain energy;

[0010] Calculate the frequency interval with maximum energy based on the sample frequency;

[0011] The frequency domain energy is traversed to determine whether there is a packet loss problem in the audio sample; if the frequency domain energy value before the calculated frequency point interval exceeds a preset threshold, it is considered that the audio sample has a packet loss problem.

[0012] In a specific implementation scheme, extracting the audio to be parsed includes:

[0013] Enter the directory path of the audio to be parsed and the audio file format to be parsed;

[0014] Recursively traverse the specified root directory to obtain all files and subdirectory paths under the directory;

[0015] Regular expression matching is performed on the obtained file list to filter out files that meet the predetermined audio file format, and a list of audio file paths that meet the conditions is output.

[0016] In a specific implementation scheme, the recursive traversal of the specified root directory includes:

[0017] Use the os.walk function in the Python standard library to recursively traverse the specified root directory.

[0018] In a specific implementation scheme, the preparing audio samples and preprocessing the audio samples includes:

[0019] Selecting an audio sample of length N from the extracted audio file, where N is a preset sample length;

[0020] Applying Hanning window function weighted processing to the audio sample, dividing the processed audio sample data into N equal parts;

[0021] Outputs the processed audio sample data.

[0022] In a specific implementation scheme, performing fast Fourier transform on the preprocessed audio samples and calculating frequency domain energy includes:

[0023] Convert the preprocessed audio samples from the time domain to the frequency domain;

[0024] Perform fast Fourier transform on the converted frequency domain data to obtain the amplitude information of the signal at different frequencies;

[0025] By calculating the absolute value of the fast Fourier transform result, the energy distribution of the audio signal at each frequency is obtained, and the frequency domain energy data of the audio signal is output.

[0026] In a specific implementation scheme, the frequency interval at which the energy is maximum is calculated according to the sample frequency, including:

[0027] Calculate the actual frequency interval corresponding to each frequency point according to the sampling frequency of the audio signal;

[0028] Based on the frequency domain energy distribution, find the frequency range with the maximum energy.

[0029] In a specific possible implementation scheme, the traversing frequency domain energy to determine whether there is a packet loss problem in the audio sample; if the frequency domain energy value before the calculated frequency point interval exceeds a preset threshold, it is considered that the audio sample has a packet loss problem, including:

[0030] Traverse the frequency domain energy data of the audio sample and check the energy value of each frequency point;

[0031] When the energy of a frequency point exceeds a preset threshold and the frequency point is located before the frequency point interval, it is determined that there is a packet loss problem in the audio sample;

[0032] Output the packet loss judgment result to inform whether there is a packet loss problem.

[0033] In a second aspect, the present application provides a fixed frequency audio fast analysis system, which adopts the following technical solution:

[0034] A fixed frequency audio fast analysis system, comprising:

[0035] An audio extraction module, used to extract the audio to be parsed;

[0036] An audio processing module, used for performing audio analysis on the extracted audio to be analyzed, preparing audio samples and preprocessing them;

[0037] An energy calculation module, used to perform fast Fourier transform on the preprocessed audio samples and calculate frequency domain energy;

[0038] An interval calculation module is used to calculate the frequency interval when the energy is maximum according to the sample frequency;

[0039] The audio analysis module is used to traverse the frequency domain energy and determine whether there is a packet loss problem in the audio sample; if the frequency domain energy value before the calculated frequency point interval exceeds a preset threshold, it is considered that the audio sample has a packet loss problem.

[0040] In a third aspect, the present application provides an electronic device, comprising a processor and a memory; the memory stores a program, and the program is loaded and executed by the processor to implement a fixed-frequency audio fast analysis method as described in the first aspect.

[0041] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the storage medium stores a program, and when the program is executed by a processor, the program is used to implement a fixed-frequency audio fast analysis method as described in the first aspect.

[0042] In summary, the beneficial effects of this application include at least:

[0043] 1) By combining signal preprocessing, Hanning window weighting, fast Fourier transform (FFT) and frequency domain energy analysis, this application can efficiently and accurately detect abnormal conditions in audio signals (such as packet loss, clipping, distortion, etc.). This process does not rely on human intervention, ensuring high-precision and efficient audio anomaly analysis.

[0044] 2) It can perform fully automatic frequency domain analysis on the extracted audio data without human intervention, which significantly improves the efficiency and accuracy of diagnosing audio anomalies and avoids errors and inefficiencies in the manual analysis process.

[0045] 3) Users can expand functions according to specific needs, such as adding detection of other types of audio anomalies (such as echo, distortion, etc.), or combining other algorithms to further improve detection accuracy to adapt to different industry applications.

[0046] By extracting the audio file to be parsed and filtering the format, the audio samples that meet the conditions are obtained. Then, the extracted audio is preprocessed using the Hanning window function to reduce spectrum leakage and ensure the accuracy of FFT processing. Then, the audio signal is converted from the time domain to the frequency domain by fast Fourier transform (FFT), and the energy distribution of the audio signal in the frequency domain is calculated, thereby locating the frequency components of the signal. Subsequently, by analyzing the frequency domain energy, the frequency point interval when the energy is maximum is calculated, and compared with the preset threshold to determine whether there is an abnormality, such as packet loss. If the energy peak is found to exceed the threshold before the normal interval of the frequency domain energy, the system will automatically identify it as a possible packet loss phenomenon and output the detection result. The above method solves the problem of insufficient accuracy and low detection efficiency of audio anomaly detection in the prior art. Through efficient audio sample processing, spectrum analysis and automated packet loss detection, the detection accuracy and processing speed of audio anomaly problems are greatly improved, avoiding the shortcomings and inefficiency of traditional manual analysis methods, and is particularly suitable for application scenarios that require real-time analysis of audio signals.

[0047] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application and implement it according to the contents of the specification, the following is a detailed description of the preferred embodiments of the present application in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a flowchart of extracting the audio to be analyzed in an embodiment of the present application.

[0049] Figure 2 It is a flowchart of audio analysis of the extracted audio to be analyzed in an embodiment of the present application.

[0050] Figure 3 It is a structural block diagram of a fixed frequency audio fast analysis system in an embodiment of the present application.

[0051] Figure 4 It is a block diagram of an electronic device for fast analysis of fixed frequency audio in an embodiment of the present application. DETAILED DESCRIPTION

[0052] The specific implementation methods of the present application are further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present application but are not intended to limit the scope of the present application.

[0053] Optionally, the present application uses the fixed-frequency audio fast analysis method provided in each embodiment as an example for explanation in an electronic device, where the electronic device is a terminal or a server. The terminal may be a computer, a tablet computer, etc. This embodiment does not limit the type of electronic device.

[0054] The present invention provides a method for fast parsing fixed-frequency audio, which includes at least the following steps:

[0055] Step S101: extract the audio to be analyzed.

[0056] In step S101, Figure 1 A schematic flow chart of extracting audio to be parsed in an embodiment of the present application is shown. First, the directory path where the audio to be parsed is located and the audio file format to be parsed are input through the tool interface or script configuration file. Then, the os.walk function in the Python standard library is used to recursively traverse the root directory specified by the user, and all files and subdirectory paths in the directory are obtained layer by layer. Each item returned by the os.walk function includes the current directory path, the subdirectory list, and the file list under the current directory. Finally, during the traversal process, the obtained file list is matched using a regular expression to filter out audio files that meet the input file suffix condition. The final program outputs a list containing all qualified audio file paths, which will be used as input data for subsequent audio processing and analysis.

[0057] Step S102: perform audio analysis on the extracted audio to be analyzed, prepare audio samples and pre-process them.

[0058] In step S102, Figure 2 The schematic diagram of the process of performing audio analysis on the extracted audio to be analyzed in the embodiment of the present application is shown, the extracted audio file to be analyzed is processed, an audio sample is prepared, and the audio sample is preprocessed and converted into a data format suitable for spectrum analysis. An audio sample of length N is taken out, and after being processed by the Hanning window function, the sample data is divided into N equal parts.

[0059] Specifically, an audio sample of length N is selected from the extracted audio file, where N is a preset sample length. The selection of the audio sample length will affect the accuracy and efficiency of the frequency domain analysis. Each time the analysis is performed, a time domain signal of the corresponding length is extracted from the audio file to prepare for the subsequent frequency domain processing. The audio sample is then weighted using the Hanning window function to reduce the spectrum leakage caused by signal truncation. FFT can only process time domain signals of a finite length, so the signal needs to be truncated. Since the signal is not an integer multiple of the periodicity when truncated, spectrum leakage may occur, and the Hanning window function can reduce this leakage phenomenon by smoothing and weighting the signal, thereby ensuring the accuracy of the frequency domain analysis. Finally, the processed audio sample data is divided equally to ensure that the data length of each segment is suitable for the subsequent fast Fourier transform (FFT) execution. The equally divided data will be used as the input for FFT analysis, which will help improve the accuracy and efficiency of the analysis.

[0060] Step S103: Perform fast Fourier transform on the preprocessed audio samples and calculate frequency domain energy.

[0061] In step S103, the preprocessed audio samples are converted from the time domain to the frequency domain, and the energy distribution of the audio signal in the frequency domain is calculated to provide basic data for subsequent analysis of audio anomalies.

[0062] Specifically, the audio sample data is divided into multiple sub-segments after weighting by the Hanning window function. Next, the fast Fourier transform (FFT) is applied to these sample data. FFT is an efficient algorithm that can convert a time domain signal into a frequency domain signal and convert the signal from the time coordinate system to the frequency coordinate system. The audio signal is composed of the superposition of multiple sound waves of different frequencies. FFT can separate these frequency components through mathematical algorithms to obtain the amplitude of the signal at each frequency.

[0063] In practice, since the output of FFT is usually a complex number, it represents the information of the amplitude and phase of the signal at different frequencies. In order to obtain the energy in the frequency domain, it is necessary to take the absolute value of the FFT result. At this time, the spectrum obtained represents the energy of the audio signal at each frequency, that is, the amplitude information in the frequency domain. These frequency domain data can effectively reflect the frequency characteristics of the audio signal and are used for subsequent analysis, such as packet loss or other anomaly detection. The frequency domain data after taking the absolute value can be further statistically analyzed to obtain the energy distribution of each frequency point. This energy data will serve as the basis for subsequent detection of audio anomalies (such as packet loss, signal distortion, etc.). If the energy exceeds a certain threshold within a specific frequency range, it can be determined that the audio signal in this frequency band is abnormal. Energy distribution can also help identify changes in different frequency components in the audio and further evaluate the quality of the audio.

[0064] Through step S103, the audio signal is converted from the time domain to the frequency domain, and the energy information in the frequency domain is obtained. These data will become the key basis for subsequent judgment of audio anomalies (such as packet loss, distortion, etc.). This step is one of the core steps for audio analysis tools to achieve efficient and accurate analysis.

[0065] Step S104: Calculate the frequency interval with maximum energy according to the sample frequency.

[0066] In step S104, the actual frequency interval corresponding to each frequency point is first calculated according to the sampling frequency of the audio signal. Then, based on the distribution of frequency domain energy, the frequency band where the frequency point with the largest energy is located is found. The interval where these frequency points are located reflects the main frequency components of the audio signal. By analyzing these frequency point intervals, possible audio anomalies, such as packet loss, can be further located.

[0067] Specifically, the frequency range corresponding to each frequency point is determined through the audio sampling frequency and FFT results. The frequency band where the peak of the frequency domain energy is located is found to be the frequency area with the strongest energy in the signal. By comparing these frequency point intervals with the preset threshold, it is detected whether there is an abnormality in the audio signal (such as packet loss, clipping, etc.).

[0068] Step S105: traverse the frequency domain energy to determine whether there is a packet loss problem in the audio sample. If the frequency domain energy value before the calculated frequency point interval exceeds a preset threshold, it is considered that the audio sample has a packet loss problem.

[0069] In step S105, first, the frequency domain energy of the audio sample is traversed to check whether the energy value exceeds a preset threshold. The threshold is set based on experience or demand, and is used to distinguish normal signals from possible abnormal signals. If it is found that the energy of a certain frequency point exceeds the threshold, and the frequency point is located before the frequency point interval, it means that there is an abnormal energy peak near the frequency band, which may be a manifestation of packet loss or other abnormalities. If it is found that the frequency domain energy of the audio signal before the frequency point interval exceeds the threshold, it is determined that there is a packet loss problem in the audio sample. The packet loss problem usually causes abnormal energy distribution in the spectrum, especially outside the expected frequency range, and the signal strength is abnormal. Finally, the result of the packet loss judgment is output and provided to the user to inform whether there is a packet loss problem and to help subsequent analysis and processing.

[0070] During this process, the frequency domain energy data of the audio samples are checked one by one, with special attention paid to whether prominent energy peaks are found outside the predetermined frequency range, which can be used as an important indicator of packet loss. Through this analysis, the system can automatically identify potential packet loss problems in the audio, thereby achieving rapid location and processing, avoiding the inefficiency and errors of manual analysis.

[0071] In summary, by extracting the audio file to be parsed and performing format screening, qualified audio samples are obtained. Next, the extracted audio is preprocessed using the Hanning window function to reduce spectrum leakage and ensure the accuracy of FFT processing. Then, the audio signal is converted from the time domain to the frequency domain by fast Fourier transform (FFT), and the energy distribution of the audio signal in the frequency domain is calculated, thereby locating the frequency components of the signal. Subsequently, by analyzing the frequency domain energy, the frequency point interval when the energy is maximum is calculated, and compared with the preset threshold to determine whether there is an abnormality, such as packet loss. If the energy peak is found to exceed the threshold before the normal interval of the frequency domain energy, the system will automatically identify it as a possible packet loss phenomenon and output the detection result. The above method solves the problem of insufficient accuracy and low detection efficiency of audio anomaly detection in the prior art. Through efficient audio sample processing, spectrum analysis and automated packet loss detection, the detection accuracy and processing speed of audio anomaly problems are greatly improved, avoiding the shortcomings and inefficiency of traditional manual analysis methods, and is particularly suitable for application scenarios that require real-time analysis of audio signals.

[0072] In addition, on this basis, users can expand functions according to specific needs, such as adding detection of other types of audio anomalies (such as echo, distortion, etc.), or combining other algorithms to further improve detection accuracy and adapt to different industry applications.

[0073] Figure 3 This is a structural block diagram of a fixed frequency audio fast analysis system provided by an embodiment of the present application. The system includes at least the following modules:

[0074] An audio extraction module, used to extract the audio to be parsed;

[0075] An audio processing module, used for performing audio analysis on the extracted audio to be analyzed, preparing audio samples and preprocessing them;

[0076] An energy calculation module, used to perform fast Fourier transform on the preprocessed audio samples and calculate frequency domain energy;

[0077] An interval calculation module is used to calculate the frequency interval when the energy is maximum according to the sample frequency;

[0078] The audio analysis module is used to traverse the frequency domain energy and determine whether there is a packet loss problem in the audio sample; if the frequency domain energy value before the calculated frequency point interval exceeds a preset threshold, it is considered that the audio sample has a packet loss problem.

[0079] For relevant details, refer to the above method embodiment.

[0080] Figure 4 4 is a block diagram of an electronic device provided by an embodiment of the present application. The device at least includes a processor 401 and a memory 402.

[0081] The processor 401 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 401 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 401 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 401 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 401 may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0082] The memory 402 may include one or more computer-readable storage media, which may be non-transitory. The memory 402 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 402 is used to store at least one instruction, which is used to be executed by the processor 401 to implement the fixed frequency audio fast parsing method provided in the method embodiment of the present application.

[0083] In some embodiments, the electronic device may further optionally include: a peripheral device interface and at least one peripheral device. The processor 401, the memory 402 and the peripheral device interface may be connected via a bus or a signal line. Each peripheral device may be connected to the peripheral device interface via a bus, a signal line or a circuit board. Schematically, the peripheral devices include but are not limited to: a radio frequency circuit, a touch display screen, an audio circuit, and a power supply.

[0084] Of course, the electronic device may also include fewer or more components, which is not limited in this embodiment.

[0085] Optionally, the present application also provides a computer-readable storage medium, in which a program is stored, and the program is loaded and executed by a processor to implement the fixed-frequency audio fast parsing method of the above method embodiment.

[0086] Optionally, the present application also provides a computer product, which includes a computer-readable storage medium, in which a program is stored, and the program is loaded and executed by a processor to implement the fixed-frequency audio fast analysis method of the above method embodiment.

[0087] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0088] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for fast analysis of fixed frequency audio, characterized in that: The method comprises: Extract the audio to be parsed; Perform audio analysis on the extracted audio to be analyzed, prepare audio samples and pre-process them; Perform fast Fourier transform on the preprocessed audio samples and calculate the frequency domain energy; Calculate the frequency interval with maximum energy based on the sample frequency; The frequency domain energy is traversed to determine whether there is a packet loss problem in the audio sample; if the frequency domain energy value before the calculated frequency point interval exceeds a preset threshold, it is considered that the audio sample has a packet loss problem.

2. The fixed frequency audio fast analysis method according to claim 1, characterized in that: The extracting of the audio to be parsed comprises: Enter the directory path of the audio to be parsed and the audio file format to be parsed; Recursively traverse the specified root directory to obtain all files and subdirectory paths under the directory; Regular expression matching is performed on the obtained file list to filter out files that meet the predetermined audio file format, and a list of audio file paths that meet the conditions is output.

3. The fixed frequency audio fast parsing method according to claim 2, characterized in that: The recursive traversal of the specified root directory includes: Use the os.walk function in the Python standard library to recursively traverse the specified root directory.

4. The fixed frequency audio fast analysis method according to claim 1, characterized in that: The preparing audio samples and preprocessing them includes: Selecting an audio sample of length N from the extracted audio file, where N is a preset sample length; Applying Hanning window function weighted processing to the audio sample, dividing the processed audio sample data into N equal parts; Outputs the processed audio sample data.

5. The method for fast analysis of fixed frequency audio according to claim 1, characterized in that: The performing fast Fourier transform on the preprocessed audio samples and calculating the frequency domain energy comprises: Convert the preprocessed audio samples from the time domain to the frequency domain; Perform fast Fourier transform on the converted frequency domain data to obtain the amplitude information of the signal at different frequencies; By calculating the absolute value of the fast Fourier transform result, the energy distribution of the audio signal at each frequency is obtained, and the frequency domain energy data of the audio signal is output.

6. The method for fast analysis of fixed frequency audio according to claim 1, characterized in that: The frequency interval when the energy is maximum according to the sample frequency calculation includes: Calculate the actual frequency interval corresponding to each frequency point according to the sampling frequency of the audio signal; Based on the frequency domain energy distribution, find the frequency range with the maximum energy.

7. The method for fast analysis of fixed frequency audio according to claim 1, characterized in that: The frequency domain energy is traversed to determine whether there is a packet loss problem in the audio sample; If the frequency domain energy value before the calculated frequency point interval exceeds the preset threshold, it is considered that the audio sample has packet loss problems including: Traverse the frequency domain energy data of the audio sample and check the energy value of each frequency point; When the energy of a frequency point exceeds a preset threshold and the frequency point is located before the frequency point interval, it is determined that there is a packet loss problem in the audio sample; Output the packet loss judgment result to inform whether there is a packet loss problem.

8. A fixed frequency audio fast analysis system, characterized in that: include: An audio extraction module, used to extract the audio to be parsed; An audio processing module, used for performing audio analysis on the extracted audio to be analyzed, preparing audio samples and preprocessing them; An energy calculation module, used to perform fast Fourier transform on the preprocessed audio samples and calculate frequency domain energy; An interval calculation module is used to calculate the frequency interval when the energy is maximum according to the sample frequency; The audio analysis module is used to traverse the frequency domain energy and determine whether there is a packet loss problem in the audio sample; if the frequency domain energy value before the calculated frequency point interval exceeds a preset threshold, it is considered that the audio sample has a packet loss problem.

9. An electronic device, characterized in that: The device includes a processor and a memory; a program is stored in the memory, and the program is loaded and executed by the processor to implement a fixed-frequency audio fast analysis method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The storage medium stores a program, which, when executed by a processor, is used to implement a fixed-frequency audio fast analysis method as described in any one of claims 1 to 7.

Images