Abnormal sound detection method, device, test equipment and storage medium for loose particles

Abstract

The present invention discloses a method, device, test equipment and storage medium for detecting abnormal sound of loose particles, including sending a sweep frequency signal to a target audio playback device, collecting a sound signal emitted by the target audio playback device, and obtaining an audio sampling signal; filtering the audio sampling signal to filter out the sweep frequency signal, and obtaining a filtered audio signal; performing envelope extraction on the filtered audio signal, and obtaining a first audio curve; determining the target number of peaks in the first audio curve according to a preset loose particle abnormal sound threshold and a peak width threshold; determining whether the target audio playback device has abnormal sound caused by loose particles according to the target number of peaks and a preset peak number threshold. The present invention can accurately detect whether the target audio playback device has abnormal sound caused by loose particles.

Description

Technical Field

[0001] The present invention relates to the technical field of audio testing, and in particular to a method, device, testing equipment and storage medium for detecting abnormal sound of loose particles. Background Art

[0002] During the production process of audio playback devices (such as speakers), loose particles may fall on the diaphragm, causing abnormal sounds when the audio playback device is playing audio. Since the abnormal sounds caused by loose particles are randomly generated, conventional abnormal sound detection methods (such as measuring THD+N, total harmonic distortion plus noise) are only suitable for detecting periodic abnormal sounds, while the human ear listening method is more subjective and less efficient. Summary of the invention

[0003] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes a method, device, test equipment and storage medium for detecting abnormal sound of loose particles, which can accurately detect whether there is abnormal sound caused by loose particles in the target audio playback device.

[0004] In a first aspect, an embodiment of the present invention provides a method for detecting abnormal sound of loose particles, comprising:

[0005] Sending a sweep frequency signal to a target audio playback device, and collecting a sound signal emitted by the target audio playback device to obtain an audio sampling signal;

[0006] Performing filtering processing on the audio sampling signal to filter out the frequency sweeping signal, thereby obtaining a filtered audio signal;

[0007] Performing envelope extraction on the filtered audio signal to obtain a first audio curve;

[0008] Determining a target number of peaks in the first audio curve according to a preset loose particle noise threshold and a peak width threshold;

[0009] According to the target peak number and a preset peak number threshold, it is determined whether the target audio playback device has abnormal sound caused by loose particles.

[0010] According to some embodiments of the present invention, determining the target number of peaks in the first audio curve according to a preset loose particle noise threshold and a peak width threshold includes:

[0011] According to a preset loose particle abnormal sound threshold, determining all first peaks in the first audio curve;

[0012] Determining a target peak in the first audio curve according to a peak width of the first peak and a preset peak width threshold;

[0013] The number of the target peaks is counted to determine the target number of peaks in the first audio curve.

[0014] According to some embodiments of the present invention, determining the target number of peaks in the first audio curve according to a preset loose particle noise threshold and a peak width threshold includes:

[0015] According to a preset loose particle abnormal sound threshold, determining all first peaks in the first audio curve;

[0016] determining a second peak in the first audio curve according to a peak width of the first peak and a preset peak width threshold;

[0017] determining a target peak in the first audio curve according to a trough depth between two adjacent peaks in the second peaks;

[0018] The number of the target peaks is counted to determine the target number of peaks in the first audio curve.

[0019] According to some embodiments of the present invention, determining the target peak in the first audio curve according to the trough depth between two adjacent peaks in the second peak includes:

[0020] determining a target trough between a current spike and a previous spike in the second spike;

[0021] When the minimum value of the target trough is smaller than a preset hysteresis threshold, the current peak is determined as a target peak.

[0022] According to some embodiments of the present invention, determining all first peaks in the first audio curve according to a preset loose particle noise threshold includes:

[0023] Traversing all the peaks in the first audio curve to determine peak data sets of all the peaks;

[0024] All first peaks in the first audio curve are determined according to the peak data set and a preset loose particle noise threshold.

[0025] According to some embodiments of the present invention, determining all first peaks in the first audio curve according to a preset loose particle noise threshold includes:

[0026] Traversing all the peaks in the first audio curve to determine peak value data sets and peak width data sets of all the peaks;

[0027] All first peaks in the first audio curve are determined based on the peak data set and a preset loose particle noise threshold, and the peak widths of the first peaks are determined in the peak width data set.

[0028] According to some embodiments of the present invention, determining the target peak in the first audio curve according to the peak width of the first peak and a preset peak width threshold includes:

[0029] According to the loose particle abnormal sound threshold, determine a first moment and a second moment of the first peak, the first moment is used to represent the moment when the rising edge of the first peak intersects with the loose particle abnormal sound threshold, and the second moment is used to represent the moment when the falling edge of the first peak intersects with the loose particle abnormal sound threshold;

[0030] Determining a peak width of the first peak according to the first moment and the second moment;

[0031] When the peak width of the first peak is greater than or equal to a preset peak width threshold, the first peak is determined as a target peak.

[0032] In a second aspect, an embodiment of the present invention provides a loose particle abnormal sound detection device, comprising:

[0033] A frequency sweep sampling module, used to send a frequency sweep signal to a target audio playback device, and collect a sound signal emitted by the target audio playback device to obtain an audio sampling signal;

[0034] A filtering module, used for filtering the audio sampling signal to filter out the frequency sweeping signal and obtain a filtered audio signal;

[0035] An envelope extraction module, used to perform envelope extraction on the filtered audio signal to obtain a first audio curve;

[0036] A first detection module, used to determine a target number of peaks in the first audio curve according to a preset loose particle abnormal sound threshold and a peak width threshold;

[0037] The second detection module is used to determine whether the target audio playback device has abnormal sound caused by loose particles based on the target peak number and a preset peak number threshold.

[0038] In a third aspect, an embodiment of the present invention provides a testing device, including a processor and a memory, wherein the memory stores a computer program, and when the processor runs the computer program, it is used to implement the above-mentioned method for detecting abnormal sound of loose particles.

[0039] In a fourth aspect, an embodiment of the present invention provides a storage medium, in which a computer program is stored, and when the computer program is executed, the above-mentioned method for detecting abnormal sound of loose particles is implemented.

[0040] The embodiments of the present invention have at least the following beneficial effects:

[0041] By sending a sweep signal to the target audio playback device and collecting the sound signal emitted by the target audio playback device, and then obtaining the first audio curve through filtering and envelope extraction, the target number of peaks in the first audio curve is analyzed according to the preset loose particle abnormal sound threshold, and then a judgment is made based on the target peak number and the preset peak number threshold, it is possible to accurately detect whether the target audio playback device has abnormal sounds caused by loose particles.

[0042] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0044] Figure 1 A flowchart of the steps of the method for detecting abnormal sound of loose particles according to an embodiment of the present invention;

[0045] Figure 2 A first audio curve waveform diagram of a good audio playback device according to an embodiment of the present invention;

[0046] Figure 3 A first audio curve waveform diagram of a defective audio playback device according to an embodiment of the present invention;

[0047] Figure 4 A schematic diagram of a partial waveform of a first audio curve of a defective audio playback device according to an embodiment of the present invention;

[0048] Figure 5 This is a principle block diagram of a device for detecting abnormal sound of loose particles according to an embodiment of the present invention;

[0049] Figure 6 The figure is a functional block diagram of a testing device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0050] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0051] In the description of the present invention, "several" means one or more, "multiple" means more than two, greater than, less than, and exceeding are understood as not including the number itself, and "above", "below", and "within" are understood as including the number itself. If there is a description of "first", "second", etc., it is only used for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0052] In the description of the present invention, unless otherwise clearly defined, words such as “setting”, “installation” and “connection” should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above words in the present invention in combination with the specific content of the technical solution.

[0053] Please refer to Figure 1 This embodiment discloses a method for detecting abnormal sound of loose particles, including steps S100 to S500. It should be noted that the steps in this embodiment are numbered only for the convenience of review and understanding, and do not limit the execution order of the steps. The contents of each step are described in detail below:

[0054] S100, sending a sweep frequency signal to a target audio playback device, and collecting a sound signal emitted by the target audio playback device to obtain an audio sampling signal;

[0055] Exemplarily, the target audio playback device can be an audio playback device such as headphones, speakers, or a speaker in an audio playback device. In order to reduce the interference of environmental noise on the detection, it is necessary to create a good test environment, such as placing the target audio playback device in a closed soundproof cavity, and setting a microphone in the soundproof cavity, and collecting the sound signal emitted by the speaker through the microphone. Among them, in order to improve the convenience of analysis, the sweep signal usually adopts a pure tone signal, such as a sine wave signal. Of course, according to the needs of actual applications, the sweep signal can also adopt a voice clip or a music clip, etc.

[0056] S200, filtering the audio sampling signal to filter out the frequency sweeping signal, and obtaining a filtered audio signal;

[0057] Exemplarily, for audio playback devices, there are many types of abnormal sounds, and the causes of the abnormal sounds may be different, such as the abnormal sounds caused by loose particles that need to be detected in this embodiment, or, for example, the abnormal sounds caused by the collision between the diaphragm and the housing. Therefore, the audio sampling signal contains more sound components, which mainly include sweep frequency signals, abnormal sound signals and environmental noise signals. In audio signal processing technology, the usual practice is to filter out abnormal sound signals and environmental noise signals to restore the audio sampling signal to the original signal (swept frequency signal) as much as possible. Different from conventional technology, this embodiment filters the swept frequency signal from the audio sampling signal to obtain a filtered audio signal, wherein the filtered audio signal includes the abnormal sound signal and the environmental noise signal.

[0058] S300, performing envelope extraction on the filtered audio signal to obtain a first audio curve;

[0059] Exemplarily, in order to facilitate quantitative analysis of the filtered audio signal, it is necessary to perform envelope extraction on the filtered audio signal to obtain the envelope of the filtered audio signal, that is, the first audio curve.

[0060] S400, determining a target number of peaks in a first audio curve according to a preset loose particle abnormal sound threshold and a peak width threshold;

[0061] Exemplarily, as described above, there may be abnormal sound signals and environmental noise signals in the filtered audio signal, and in the abnormal sound signals, in addition to the abnormal sound signals that may be caused by loose particles, there may also be other types of abnormal sound signals. However, in the study, it was found that the abnormal sound signals caused by loose particles are usually manifested as obvious peaks in the audio curve. By setting a suitable threshold, the abnormal sound signals caused by loose particles can be distinguished from most other abnormal sound signals or environmental noise signals, that is, the abnormal sound signals that may be caused by loose particles can be extracted. Among them, the loose particle abnormal sound threshold can be obtained by comparing and analyzing good audio playback devices and defective audio playback devices that only have loose particle abnormal sound signals.

[0062] S500: Determine whether a target audio playback device has abnormal sound caused by loose particles according to a target peak number and a preset peak number threshold.

[0063] Exemplarily, the peaks in the first audio curve may be caused by other reasons besides the abnormal sounds caused by loose particles, such as abnormalities in the sound signal acquisition process. Through research on multiple defective playback devices with abnormal sounds caused by loose particles, it is found that the abnormal sounds caused by loose particles usually show multiple peaks on the audio curve. Therefore, by setting an appropriate peak number threshold, it is possible to more accurately determine whether the target audio playback device has abnormal sounds caused by loose particles. When the target peak number exceeds the peak number threshold, it is determined that the target audio playback device has abnormal sounds caused by loose particles. For example, please refer to Figure 2 and Figure 3 , Figure 2 and Figure 3 The mark L1 indicates the first audio curve, the mark L2 indicates the loose particle noise threshold, and the mark P1 indicates the part of the first audio curve that exceeds the loose particle noise threshold. Figure 2 It can be seen that the target peak number of the good audio playback device is small (only 1). Figure 3 It can be seen that for a defective audio playback device with more obvious abnormal sound, the number of target peaks is relatively large (more than 100).

[0064] In this way, by sending a sweep signal to the target audio playback device and collecting the sound signal emitted by the target audio playback device, and then obtaining the first audio curve through filtering and envelope extraction, the target number of peaks in the first audio curve is analyzed according to the preset loose particle abnormal sound threshold, and then a judgment is made based on the target peak number and the preset peak number threshold, so as to accurately detect whether the target audio playback device has abnormal sounds caused by loose particles.

[0065] In some application examples, step S400, determining a target number of peaks in the first audio curve according to a preset loose particle noise threshold and a peak width threshold, includes:

[0066] S410, determining all first peaks in the first audio curve according to a preset loose particle abnormal sound threshold;

[0067] S420, determining a target peak in the first audio curve according to the peak width of the first peak and a preset peak width threshold;

[0068] S430: Count the number of target peaks to determine the number of target peaks in the first audio curve.

[0069] For example, the study also found that there may be many peaks in the audio curve, but not all abnormal sound signals corresponding to the peaks will affect the sound playback quality of the target audio playback device. For example, when the width of the peak is narrow (i.e., the duration is short, such as Figure 4In other words, when the peak width is small enough, even if the sound signal corresponding to the peak is an abnormal sound caused by loose particles, it will not make the user feel uncomfortable, and the abnormal sound is an acceptable abnormal sound signal. Therefore, in order to more accurately detect abnormal sounds, all the first peaks in the first audio curve are determined, and the peak width threshold is used to determine whether the corresponding first peak is a target peak. If it is a target peak, it is counted, otherwise it is not included in the counting range.

[0070] In some other application examples, step S400, determining a target number of peaks in the first audio curve according to a preset loose particle noise threshold and a peak width threshold, includes:

[0071] S410, determining all first peaks in the first audio curve according to a preset loose particle abnormal sound threshold;

[0072] S420, determining a second peak in the first audio curve according to the peak width of the first peak and a preset peak width threshold;

[0073] S440, determining a target peak in the first audio curve according to a trough depth between two adjacent peaks in the second peaks;

[0074] S430: Count the number of target peaks to determine the number of target peaks in the first audio curve.

[0075] For example, it is found in the study that in addition to the peak width, the fluctuation between two adjacent peaks will also affect the human ear's acceptance of abnormal sounds. Therefore, in this embodiment, the trough depth between two peaks is used to characterize the fluctuation between two adjacent peaks. When the trough depth meets the requirement, the corresponding peak is determined as the target peak.

[0076] Wherein, step S440, determining a target peak in the first audio curve according to a trough depth between two adjacent peaks in the second peak, comprises:

[0077] S441, determining a target trough between a current peak and a previous peak in the second peak;

[0078] S442: When the minimum value of the target trough is smaller than a preset hysteresis threshold, the current peak is determined as the target peak.

[0079] For example, the hysteresis threshold is usually smaller than the loose particle noise threshold in value, and during the analysis, the hysteresis threshold is detected after the first spike (e.g. Figure 4 When the target peak is the peak F1, the trough depth needs to be determined starting from the second peak. Figure 4When analyzing the peaks (as shown by the mark F2 in the figure), there must be a trough between the two adjacent peaks (the first peak and the second peak). Figure 4 As shown by the mark B1 in the figure), the trough must have a minimum value. If the minimum value of the target trough is greater than the hysteresis threshold (such as Figure 4 If the minimum value of the target trough (such as the trough shown by mark B2 between mark F1 and mark F3) is less than the hysteresis threshold, it can be considered that the fluctuation degree between the two adjacent peaks (such as the peaks shown by mark F1 and mark F3) is large enough, and the abnormal sound signal corresponding to the peak will affect the sound playback quality of the audio playback device, and the corresponding peak (such as the peak shown by mark F3) is determined as the target peak, otherwise the corresponding peak is ignored and the next peak is judged.

[0080] In some application examples, step S410, determining all first peaks in the first audio curve according to a preset loose particle noise threshold, includes:

[0081] S411, traversing all the peaks in the first audio curve to determine peak data sets of all the peaks;

[0082] S412: Determine all first peaks in the first audio curve according to the peak data set and a preset loose particle noise threshold.

[0083] For example, in signal curve analysis, peak detection is a relatively mature technology. The peak value corresponding to the peak can be determined by the peak detection technology. The peak values ​​of all the peaks can form a peak data set. According to the peak data set and the loose particle noise threshold, all the first peaks can be screened out from all the peaks. For example, when the peak value of a certain peak is greater than the loose particle noise threshold, the corresponding peak is determined as the first peak, and so on, all the first peaks can be screened out.

[0084] In some other application examples, step S410, determining all first peaks in the first audio curve according to a preset loose particle abnormal sound threshold, includes:

[0085] S413, traversing all the peaks in the first audio curve to determine peak value data sets and peak width data sets of all the peaks;

[0086] S414, determining all first peaks in the first audio curve according to the peak data set and the preset loose particle noise threshold, and determining the peak widths of the first peaks in the peak width data set.

[0087] Exemplarily, in this application example, when traversing all peaks, the peak data set and the peak width data set of all peaks are synchronously determined. By comparing the peak data set with the loose particle abnormal sound threshold, it can be determined which peaks in the first audio curve are the first peaks whose peak values ​​exceed the loose particle abnormal sound threshold. At this time, the peak width corresponding to the first peak can be determined in the peak width data set by means of the peak sequence number index. In this way, the detection efficiency can be improved while ensuring the detection accuracy.

[0088] Step S420, determining a target peak in the first audio curve according to the peak width of the first peak and a preset peak width threshold, includes:

[0089] S421, determining a first moment and a second moment of the first peak according to the loose particle abnormal sound threshold, the first moment being used to characterize the moment when the rising edge of the first peak intersects with the loose particle abnormal sound threshold, and the second moment being used to characterize the moment when the falling edge of the first peak intersects with the loose particle abnormal sound threshold;

[0090] S422, determining a peak width of a first peak according to the first moment and the second moment;

[0091] S423: When the peak width of the first peak is greater than or equal to a preset peak width threshold, determine the first peak as a target peak.

[0092] For example, the loose particle noise threshold is usually set to a single value, which is represented by a straight line parallel to the time axis in the first audio curve. If the peak value of the first peak exceeds the loose particle noise threshold, then the curve segment corresponding to the first peak in the first audio curve must intersect with the straight line corresponding to the loose particle noise threshold, and the first moment (e.g. Figure 4 and the second moment (as shown in the figure t1) Figure 4 t2 in the figure), thereby determining the peak width of the first peak (as shown in Figure 4 The width between the marker t1 and the marker t2 in the figure is used to determine whether the first peak is a target peak.

[0093] In this way, this embodiment analyzes the first audio curve by comprehensively analyzing multiple dimensions, thereby accurately and stably detecting whether the target audio playback device has abnormal sounds caused by loose particles, thereby avoiding the subjectivity and uncertainty of manual listening detection.

[0094] Please refer to Figure 5 This embodiment also provides a loose particle abnormal sound detection device, including a sweep frequency sampling module 110, a filtering module 120, an envelope extraction module 130, a first detection module 140 and a second detection module 150.

[0095] The frequency sweep sampling module 110 is used to send a frequency sweep signal to a target audio playback device and collect a sound signal emitted by the target audio playback device to obtain an audio sampling signal;

[0096] The filtering module 120 is used to filter the audio sampling signal to filter out the swept frequency signal to obtain a filtered audio signal;

[0097] The envelope extraction module 130 is used to perform envelope extraction on the filtered audio signal to obtain a first audio curve;

[0098] The first detection module 140 is used to determine the target number of peaks in the first audio curve according to the preset loose particle abnormal sound threshold and the peak width threshold;

[0099] The second detection module 150 is used to determine whether the target audio playback device has abnormal sound caused by loose particles according to the target peak number and a preset peak number threshold.

[0100] The inventive concept of the embodiment of the device for detecting abnormal sound of loose particles is the same as the inventive concept of the embodiment of the method for detecting abnormal sound of loose particles described above. The contents not involved in the embodiment of the device for detecting abnormal sound of loose particles can be referred to the embodiment of the method for detecting abnormal sound of loose particles described above, and will not be repeated here. By sending a sweep signal to the target audio playback device and collecting the sound signal emitted by the target audio playback device, and then filtering and envelope extraction to obtain the first audio curve, the target number of peaks in the first audio curve is analyzed according to the preset loose particle abnormal sound threshold, and then the target number of peaks and the preset peak number threshold are used to make a judgment, so as to accurately detect whether the target audio playback device has abnormal sound caused by loose particles.

[0101] Please refer to Figure 6 , this embodiment also provides a test device, including a processor 210 and a memory 220, wherein a computer program is stored in the memory 220, and when the processor 210 runs the computer program, it is used to implement the above-mentioned method for detecting abnormal sound of loose particles. The details of the method for detecting abnormal sound of loose particles can be found above and will not be repeated here. By sending a sweep signal to the target audio playback device and collecting the sound signal emitted by the target audio playback device, and then obtaining a first audio curve through filtering and envelope extraction, the target number of peaks in the first audio curve is analyzed according to the preset loose particle abnormal sound threshold, and then the target number of peaks and the preset peak number threshold are used to make a judgment, so as to accurately detect whether the target audio playback device has abnormal sound caused by loose particles.

[0102] The present embodiment also provides a storage medium, in which a computer program is stored, and when the computer program is executed, the above-mentioned method for detecting abnormal sound of loose particles is implemented. The details of the method for detecting abnormal sound of loose particles can be found above and will not be repeated here. By sending a sweep signal to the target audio playback device and collecting the sound signal emitted by the target audio playback device, and then obtaining the first audio curve through filtering and envelope extraction, the target number of peaks in the first audio curve is analyzed according to the preset loose particle abnormal sound threshold, and then the target number of peaks and the preset peak number threshold are used to make a judgment, so as to accurately detect whether the target audio playback device has abnormal sound caused by loose particles.

[0103] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge scope of ordinary technicians in the relevant technical field without departing from the purpose of the present invention.

Claims

1. A method for detecting abnormal sound of loose particles, characterized in that: include: Sending a sweep frequency signal to a target audio playback device, and collecting a sound signal emitted by the target audio playback device to obtain an audio sampling signal; Performing filtering processing on the audio sampling signal to filter out the frequency sweeping signal, thereby obtaining a filtered audio signal; Performing envelope extraction on the filtered audio signal to obtain a first audio curve; Determining a target number of peaks in the first audio curve according to a preset loose particle noise threshold and a peak width threshold; Determining whether the target audio playback device has abnormal sound caused by loose particles according to the target peak number and a preset peak number threshold; The step of determining the target number of peaks in the first audio curve according to the preset loose particle abnormal sound threshold and peak width threshold includes: According to a preset loose particle abnormal sound threshold, determining all first peaks in the first audio curve; According to the loose particle abnormal sound threshold, determine a first moment and a second moment of the first peak, the first moment is used to represent the moment when the rising edge of the first peak intersects with the loose particle abnormal sound threshold, and the second moment is used to represent the moment when the falling edge of the first peak intersects with the loose particle abnormal sound threshold; Determining a peak width of the first peak according to the first moment and the second moment; When the peak width of the first peak is greater than or equal to a preset peak width threshold, determining the first peak as a target peak; The number of the target peaks is counted to determine the target number of peaks in the first audio curve.

2. The method for detecting abnormal sound of loose particles according to claim 1, characterized in that: The step of determining the target number of peaks in the first audio curve according to the preset loose particle abnormal sound threshold and the peak width threshold comprises: According to a preset loose particle abnormal sound threshold, determining all first peaks in the first audio curve; determining a second peak in the first audio curve according to a peak width of the first peak and a preset peak width threshold; determining a target peak in the first audio curve according to a trough depth between two adjacent peaks in the second peaks; The number of the target peaks is counted to determine the target number of peaks in the first audio curve.

3. The method for detecting abnormal sound of loose particles according to claim 2, characterized in that: The step of determining a target peak in the first audio curve according to a trough depth between two adjacent peaks in the second peaks includes: determining a target trough between a current spike and a previous spike in the second spike; When the minimum value of the target trough is smaller than a preset hysteresis threshold, the current peak is determined as a target peak.

4. The method for detecting abnormal sound of loose particles according to claim 1, 2 or 3, characterized in that: The step of determining all first peaks in the first audio curve according to a preset loose particle abnormal sound threshold comprises: Traversing all the peaks in the first audio curve to determine peak data sets of all the peaks; All first peaks in the first audio curve are determined according to the peak data set and a preset loose particle noise threshold.

5. The method for detecting abnormal sound of loose particles according to claim 1, 2 or 3, characterized in that: The step of determining all first peaks in the first audio curve according to a preset loose particle abnormal sound threshold comprises: Traversing all the peaks in the first audio curve to determine peak value data sets and peak width data sets of all the peaks; All first peaks in the first audio curve are determined based on the peak data set and a preset loose particle noise threshold, and the peak widths of the first peaks are determined in the peak width data set.

6. A loose particle noise detection device, characterized in that: include: A frequency sweep sampling module, used to send a frequency sweep signal to a target audio playback device, and collect a sound signal emitted by the target audio playback device to obtain an audio sampling signal; A filtering module, used for filtering the audio sampling signal to filter out the frequency sweeping signal and obtain a filtered audio signal; An envelope extraction module, used to perform envelope extraction on the filtered audio signal to obtain a first audio curve; A first detection module, used to determine a target number of peaks in the first audio curve according to a preset loose particle abnormal sound threshold and a peak width threshold; A second detection module is used to determine whether the target audio playback device has abnormal sound caused by loose particles according to the target peak number and a preset peak number threshold; The step of determining the target number of peaks in the first audio curve according to the preset loose particle abnormal sound threshold and peak width threshold includes: According to a preset loose particle abnormal sound threshold, determining all first peaks in the first audio curve; According to the loose particle abnormal sound threshold, determine a first moment and a second moment of the first peak, the first moment is used to represent the moment when the rising edge of the first peak intersects with the loose particle abnormal sound threshold, and the second moment is used to represent the moment when the falling edge of the first peak intersects with the loose particle abnormal sound threshold; Determining a peak width of the first peak according to the first moment and the second moment; When the peak width of the first peak is greater than or equal to a preset peak width threshold, determining the first peak as a target peak; The number of the target peaks is counted to determine the target number of peaks in the first audio curve.

7. A test device, comprising a processor and a memory, wherein a computer program is stored in the memory, characterized in that: When the processor runs the computer program, it is used to implement the method for detecting abnormal sound of loose particles as described in any one of claims 1 to 5.

8. A storage medium storing a computer program, characterized in that: When the computer program is executed, the method for detecting abnormal sound of loose particles as described in any one of claims 1 to 5 is implemented.

Images