Particle material identification method, device and electronic equipment

Abstract

The application is suitable for the technical field of material identification, and provides a particle material identification method, device and electronic equipment. The method comprises the following steps: obtaining an original sound signal generated by particle collision, frame dividing the original sound signal to obtain multiple frames of sound signals; calculating a frequency band variance value corresponding to each frame of sound signal in the multiple frames of sound signals, and determining a pulse decision value of each frame of sound signal according to the frequency band variance value corresponding to each frame of sound signal; determining multiple pulse signal segments of the particle according to the pulse decision value of each frame of sound signal and the frequency band variance value corresponding to each frame of sound signal, and constructing a reconstructed sound signal of the particle based on the multiple pulse signal segments; performing feature extraction on the reconstructed sound signal to obtain a feature vector of the particle, and inputting the feature vector into a preset material identification model to obtain the material of the particle. The application can construct the reconstructed sound signal in real time and accurately, and then accurately identify the material of the particle based on the reconstructed sound signal.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of material identification, in particular to a particle material identification method and device and electronic equipment. BACKGROUND

[0002] Excessive particles in electronic components refer to various metal or non-metal particles remaining or generated in the cavity of the electronic components during the processing and manufacturing and use of the electronic components. The excessive particles in the electronic components are one of the main factors affecting the reliability of electronic devices, and therefore the detection and identification of the excessive particles in the electronic components are very important.

[0003] In the related art, the detection device can only detect whether there are excessive particles, but cannot accurately identify the material of the excessive particles, which further causes the staff to be unable to accurately trace the source of the excessive particles according to the material of the excessive particles and unable to improve the manufacturing process. SUMMARY

[0004] Therefore, the embodiments of the present application provide a particle material identification method and device and electronic equipment to solve the technical problem that the material of the particle cannot be accurately identified in the related art, which further causes the staff to be unable to accurately trace the source of the excessive particles according to the material of the particle and unable to improve the manufacturing process.

[0005] In a first aspect, the embodiments of the present application provide a particle material identification method, which includes: obtaining an original sound signal generated by particle collision, frame dividing the original sound signal to obtain a plurality of sound signals; calculating a frequency band variance value corresponding to each sound signal in the plurality of sound signals, and determining a pulse decision value of each sound signal according to the frequency band variance value corresponding to each sound signal; determining a plurality of pulse signal segments of the particle according to the pulse decision value of each sound signal and the frequency band variance value corresponding to each sound signal, and constructing a reconstructed sound signal of the particle based on the plurality of pulse signal segments; performing feature extraction on the reconstructed sound signal to obtain a feature vector of the particle, and inputting the feature vector into a preset material identification model to obtain the material of the particle, wherein the material identification model inputs the feature vector of the particle and outputs the material of the particle.

[0006] In a possible implementation manner of the first aspect, the pulse decision value of each sound signal is determined according to the frequency band variance value corresponding to each sound signal, which includes: determining a short-time signal-to-noise ratio corresponding to each sound signal according to the frequency band variance value of each sound signal and the frequency band variance value of the previous sound signal of each sound signal; and determining the pulse decision value corresponding to each sound signal according to the short-time signal-to-noise ratio corresponding to each sound signal.

[0007] In one possible implementation of the first aspect, the pulse decision value includes a pulse signal start-point decision value and a pulse signal end-point decision value; determining the pulse decision value corresponding to each frame of the acoustic signal based on the short-time signal-to-noise ratio corresponding to each frame of the acoustic signal includes: according to the expression:

[0008] Determine the pulse signal start-point decision value G corresponding to each frame of the sound signal. i1 and pulse signal endpoint decision value G i2 Where j = 1, 2, 1 represents the start point of the pulse signal, 2 represents the end point of the pulse signal, i is the frame number corresponding to the sound signal, i = 1, 2, 3, ..., I, I is determined according to the total number of frames of the sound signal, SNR i Let D be the short-time signal-to-noise ratio corresponding to the i-th frame of the audio signal. i-1 The frequency band variance value corresponding to the sound signal in the previous frame of the i-th frame and the (i-1)-th frame is α1, which is the first preset coefficient and α2, which is the second preset coefficient, and α1 is greater than α2. Y is the first preset threshold and -Y is the second preset threshold. dB represents decibels.

[0009] In one possible implementation of the first aspect, determining the short-time signal-to-noise ratio (SNR) of each frame of sound signal based on the frequency band variance of each frame of sound signal and the frequency band variance of the previous frame of sound signal includes: according to the expression:

[0010] Determine the short-time signal-to-noise ratio (SNR) for each frame of the audio signal. i In the formula, i is the frame number corresponding to the sound signal, i = 1, 2, 3, ..., I, where I is determined based on the total number of frames of the sound signal, and D i Let D be the frequency band variance value corresponding to the i-th frame of the audio signal. i-1 The frequency band variance value corresponding to the sound signal in the previous frame and the (i-1)th frame is given.

[0011] In one possible implementation of the first aspect, the pulse decision value includes a pulse signal start-up decision value and a pulse signal end-up decision value; based on the pulse decision value of each frame of acoustic signal and the corresponding frequency band variance value of each frame of acoustic signal, multiple pulse signal segments of the particle are determined, and based on the multiple pulse signal segments, a reconstructed acoustic signal of the particle is constructed, including: determining the start-up of the pulse signal segment based on the frequency band variance value corresponding to each frame of acoustic signal and the corresponding pulse signal start-up decision value, and determining the end-up of the pulse signal segment based on the frequency band variance value corresponding to each frame of acoustic signal and the corresponding pulse signal end-up decision value; based on the start-up and end-up of the pulse signal segments, multiple pulse signal segments of the particle are determined, and the multiple pulse signal segments are sequentially connected to obtain the reconstructed acoustic signal of the particle.

[0012] In a possible implementation of the first aspect, the start point of the pulse signal segment is determined according to the frequency band variance value corresponding to each frame of the sound signals and the corresponding pulse signal start point decision value, and the end point of the pulse signal segment is determined according to the frequency band variance value corresponding to each frame of the sound signals and the corresponding pulse signal end point decision value, including: starting from the second frame of the sound signals in the sound signals, if the frequency band variance values corresponding to a continuous preset number of frames of the sound signals are all greater than the pulse signal start point decision value corresponding to the corresponding frame of the sound signals, the next frame of the sound signals of the continuous preset number of frames of the sound signals is taken as the start point of the pulse signal segment; if the frequency band variance values corresponding to the continuous preset number of frames of the sound signals are all less than the pulse signal end point decision value corresponding to the corresponding frame of the sound signals, the next frame of the sound signals of the continuous preset number of frames of the sound signals is taken as the end point of the pulse signal segment.

[0013] In a possible implementation of the first aspect, after the plurality of pulse signal segments of the particle are determined according to the pulse decision value of each frame of the sound signals and the frequency band variance value corresponding to each frame of the sound signals, the method further includes: determining whether there is a pulse signal segment in the plurality of pulse signal segments, the pulse duration corresponding to the pulse signal segment being less than a preset duration; if there is a pulse signal segment in the plurality of pulse signal segments, the pulse duration corresponding to the pulse signal segment being less than the preset duration, deleting the pulse signal segment with the pulse duration less than the preset duration from the plurality of pulse signal segments, and taking the remaining pulse signal segments of the plurality of pulse signal segments as the plurality of pulse signal segments.

[0014] In a possible implementation of the first aspect, the feature extraction is performed on the reconstructed sound signal to obtain a feature vector of the particle, and the feature vector is input into a preset material identification model to obtain the material of the particle, including: performing a plurality of feature extractions on the reconstructed sound signal to obtain a plurality of feature quantities respectively, and constructing the feature vector of the particle according to the plurality of feature quantities; performing dimension reduction processing on the feature vector to obtain a dimension-reduced feature vector; and inputting the dimension-reduced feature vector into the material identification model to obtain the material of the particle.

[0015] In the second aspect, an embodiment of the present application provides a particle material identification device, including:

[0016] The acquisition module is configured to acquire an original sound signal generated by particle collision, and frame the original sound signal to obtain a plurality of frames of sound signals.

[0017] The calculation module is configured to calculate a frequency band variance value corresponding to each frame of the sound signals in the plurality of frames of sound signals, and determine a pulse decision value of each frame of the sound signals according to the frequency band variance value corresponding to each frame of the sound signals.

[0018] The construction module is configured to determine a plurality of pulse signal segments of the particle according to the pulse decision value of each frame of the sound signals and the frequency band variance value corresponding to each frame of the sound signals, and construct a reconstructed sound signal of the particle based on the plurality of pulse signal segments.

[0019] The identification module is configured to perform feature extraction on the reconstructed acoustic signal to obtain a feature vector of the particle, and input the feature vector into a preset material identification model to obtain a material of the particle.

[0020] In a third aspect, an electronic device is provided, which includes a memory and a processor. The memory stores a computer program capable of running on the processor. The processor implements the particle material identification method according to any one of the first aspect when executing the computer program.

[0021] In a fourth aspect, a computer readable storage medium is provided, which stores a computer program. The computer program is executed by a processor to implement the particle material identification method according to any one of the first aspect.

[0022] In a fifth aspect, a computer program product is provided. When the computer program product is run on an electronic device, the electronic device executes the particle material identification method according to any one of the first aspect.

[0023] It can be understood that the beneficial effects of the second aspect to the fifth aspect can be referred to the related description of the first aspect, which will not be repeated here.

[0024] The particle material identification method, device and electronic device provided by the embodiments of the present application can reduce the influence of various signals such as noise signals, interference signals and component signals on the reconstructed acoustic signal, improve the accuracy and real-time performance of the reconstructed acoustic signal, and then accurately identify the material of the particle based on the reconstructed acoustic signal, so that the source of the redundant particle can be accurately traced based on the material of the particle, and the manufacturing process and use procedure of the electronic component can be improved.

[0025] It should be understood that the general description above and the detailed description below are only exemplary and explanatory, and cannot limit the present specification. BRIEF DESCRIPTION OF DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0027] Figure 1 is a flowchart of a particle material identification method provided by an embodiment of the present application;

[0028] Figure 2 is a flowchart of a particle material identification method provided by another embodiment of the present application;

[0029] Figure 3 is a structural diagram of a particle material identification device provided by an embodiment of the present application;

[0030] Figure 4 is a structural diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0031] The present application will be described in more detail by the following specific embodiments. The following embodiments will help those skilled in the art to further understand the role of the present application, but do not limit the present application in any form. It should be noted that for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made. These all belong to the protection scope of the present application.

[0032] It should be understood that when used in the specification and appended claims of the present application, the term "comprising" indicates the presence of the described features, whole, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, whole, steps, operations, elements, components and / or sets thereof.

[0033] It should also be understood that the term "and / or" used in the specification and appended claims of the present application means any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0034] In the description of the present application and the appended claims, the terms "first", "second", "third" and the like are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0035] In the present application, the reference "one embodiment" or "some embodiments" and the like means that the specific features, structures or characteristics described in connection with the embodiment are included in one or more embodiments of the present application. Therefore, the statements "in one embodiment", "in some embodiments", "in other some embodiments", "in other some embodiments" and the like appearing in different places in the specification are not necessarily all referring to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized. The terms "include", "contain", "have" and their variants mean "include but not limited to", unless otherwise specifically emphasized.

[0036] In addition, the "multiple" mentioned in the embodiments of the present application should be interpreted as two or more than two.

[0037] Excess particles in electronic components refer to various metal or non-metal particles remaining or generated inside the cavity of electronic components during the manufacturing and use of the electronic components. The excess particles inside the electronic components are one of the main factors affecting the reliability of electronic devices, and therefore the detection and identification of the excess particles inside the electronic components are very important. In the related art, the detection device can only detect whether there are excess particles, but cannot accurately identify the material of the excess particles, thereby causing the staff to be unable to accurately trace the source of the excess particles according to the material of the excess particles and unable to improve the manufacturing process.

[0038] Based on the above problems, the inventors have found that, by frame dividing the original sound signal generated by particle collision, calculating the corresponding adaptive pulse decision value for each frame of sound signal in the obtained multiple frames of sound signal, and then determining multiple pulse signal segments for constructing a reconstructed sound signal according to the adaptive pulse decision value, the influence of various signals such as noise signals, interference signals and component signals on the reconstructed sound signal can be reduced, and then the material of the particle can be accurately determined based on the reconstructed sound signal and a preset material identification model.

[0039] Figure 1 is a flowchart of a particle material identification method provided by an embodiment of the present application. As shown in Figure 1 The method in the embodiments of the present application can include the following steps.

[0040] Step 101: Obtain the original sound signal generated by particle collision, frame divide the original sound signal, and obtain multiple frames of sound signals.

[0041] For example, the embodiment can obtain the original sound signal generated by particle collision detected by a particle collision noise automatic detection system, and perform discrete Fourier transform on the original sound signal to obtain a frequency domain sound signal corresponding to the original sound signal. The original sound signal includes pulse signals, noise signals, interference signals and component signals, etc., and only the pulse signals are useful signals for identifying the material of the particle. The embodiment frame divides the above frequency domain sound signal to obtain multiple frames of sound signals. After the discrete Fourier transform and frame division of the original sound signal, it can be represented as:

[0042]

[0043] In the formula, X i is the i-th frame of sound signal in the frequency domain sound signal, i is the frame number corresponding to the sound signal, i = 1, 2, 3, …, I, I is determined according to the total number of frames of sound signals, x i() is the i-th frame of the original sound signal, m is the window function variable, N is the frame length, and k is between 0 and N-1.

[0044] Step 102, calculate the frequency band variance value corresponding to each frame of sound signal in the multi-frame sound signal, and determine the pulse decision value of each frame of sound signal according to the frequency band variance value corresponding to each frame of sound signal.

[0045] For example, the embodiment can calculate the frequency band variance value corresponding to each frame of sound signal according to the following formula:

[0046]

[0047] In the formula, D i is the frequency band variance value corresponding to the i-th frame of sound signal.

[0048] Optionally, the embodiment can determine the pulse decision value corresponding to each frame of sound signal according to the frequency band variance value of each frame of sound signal and the frequency band variance value of the previous frame of sound signal of each frame of sound signal. The pulse decision value can include a pulse signal start point decision value and a pulse signal end point decision value. The pulse signal start point decision value is used to determine the start point of the pulse signal segment, and the pulse signal end point decision value is used to determine the end point of the pulse signal segment.

[0049] Step 103, determine a plurality of pulse signal segments of the particle according to the pulse decision value of each frame of sound signal and the frequency band variance value corresponding to each frame of sound signal, and construct a reconstructed sound signal of the particle based on the plurality of pulse signal segments.

[0050] For example, as known from the foregoing, the pulse decision value includes a pulse signal start point decision value and a pulse signal end point decision value. The embodiment can start from a certain frame of sound signal, such as the second frame of sound signal, of the multi-frame sound signal, and sequentially determine the pulse start point or the pulse end point of each frame of sound signal in order of frame number from small to large. For example, the embodiment can determine the start points of a plurality of pulse signal segments according to the pulse signal start point decision value and the frequency band variance value corresponding to each frame of sound signal, and determine the end points of the plurality of pulse signal segments according to the pulse signal end point decision value and the frequency band variance value corresponding to each frame of sound signal, and then determine the plurality of pulse signal segments according to the start points of the plurality of pulse signal segments and the end points of the plurality of pulse signal segments, so as to construct a reconstructed sound signal of the particle based on the plurality of pulse signal segments.

[0051] In the embodiment, the pulse decision value corresponding to each frame of sound signal is determined according to the frequency band variance value corresponding to each frame of sound signal, so the pulse decision value is not a fixed value, and when the pulse start point or the pulse end point of each frame of sound signal is determined according to the pulse decision value, the start point or the end point of the pulse signal segment of the frame of sound signal can be more accurately determined, and the situation that other signals such as noise signals, interference signals and component signals are misjudged as pulse signals is reduced, so that only pulse signals are included in the reconstructed sound signal, and the accuracy of the reconstructed sound signal is improved.

[0052] It should be noted that the determination of the pulse start point or the pulse end point starts from the second frame of sound signal in the multiple frames of sound signals, because in the initial stage of sound signal detection of the particle collision noise automatic detection system, the original sound signal of the detected particle collision basically does not include pulse signals and other signals such as noise signals, interference signals and component signals, that is, the first frame of sound signal in the multiple frames of sound signals generally does not include pulse signals and other signals, so the determination is performed from the second frame of sound signal in the multiple frames of sound signals.

[0053] In addition, the reconstructed sound signal obtained in the embodiment only includes pulse signal segments, and other signals such as noise signals, interference signals and component signals are removed, so that the data calculation amount can be reduced in the subsequent feature extraction process of the reconstructed sound signal. The pulse signals in the original sound signal appear randomly, and the accuracy of particle material identification based on the original sound signal is low, while the reconstructed sound signal is a smooth signal, and the accuracy of particle material identification based on the reconstructed sound signal is higher.

[0054] In step 104, the reconstructed sound signal is subjected to feature extraction to obtain a feature vector of the particle, and the feature vector is input into a preset material identification model to obtain the material of the particle.

[0055] The material identification model inputs the feature vector of the particle and outputs the material of the particle.

[0056] In a possible implementation, the reconstructed sound signal can be subjected to feature extraction to obtain a plurality of feature quantities, and a feature vector of the particle is constructed according to the plurality of feature quantities, the feature vector is subjected to dimension reduction processing to obtain a reduced feature vector, the overlap and redundancy between the features are reduced, the subsequent calculation is simplified, the reduced feature vector is input into the material identification model, and the material of the particle is quickly identified and obtained.

[0057] For example, the plurality of feature quantities can include a pulse peak-peak value F, a pulse time domain energy E T , a pulse duration T, a pulse zero-crossing rate Zcr and a pulse Teager energy E Tgand a pulse energy-entropy ratio EF. The embodiment can include C1 to C6 when a plurality of feature quantities are obtained by performing a plurality of feature extractions on the reconstructed acoustic signal.

[0058] C1, calculating the sum of absolute values of maximum and minimum values of pulse signal amplitudes in each pulse signal segment in the reconstructed acoustic signal, and performing mean value calculation on the sum of absolute values to obtain a first mean value as a pulse peak-to-peak value F.

[0059] C2, calculating a sum of squares of pulse signal amplitudes in each pulse signal segment in the reconstructed acoustic signal, and taking the sum of squares as a pulse time-domain energy E. T .

[0060] C3, calculating a duration from a start time to an end time of each pulse signal segment in the reconstructed acoustic signal, and performing mean value calculation on the durations to obtain a second mean value as a pulse duration T.

[0061] C4, calculating a number of times of zero-crossing of the pulse signal per unit time in the reconstructed acoustic signal to obtain a pulse zero-crossing rate Zcr.

[0062] C5, calculating a transient energy corresponding to each pulse signal segment in the reconstructed acoustic signal to obtain a pulse Teager energy E. Tg .

[0063] C6, performing Fourier transform on the reconstructed acoustic signal to obtain a pulse frequency-domain energy and a spectral entropy value, and performing ratio calculation on the pulse frequency-domain energy and the spectral entropy value to obtain a ratio as a pulse energy-entropy ratio EF.

[0064] Optionally, the six feature quantities are fused in a serial manner to construct a feature vector E of the particle. E = [F, E T , T, Zcr, E Tg , EF].

[0065] For example, the embodiment can use Fisher discriminant method to perform dimension reduction processing on the feature vector E, map the six-dimensional feature vector to a low-dimensional feature parameter space, and obtain a reduced feature vector E' = [T1, T2, T3], wherein T1, T2, and T3 are vectors in the reduced feature vector, so as to reduce the overlap and redundancy between the features and simplify the calculation.

[0066] Here, after obtaining the reduced dimension feature vector, the embodiment can input the reduced dimension feature vector into a material recognition model, the material recognition model inputs the feature vector of the particle and outputs the material of the particle. Optionally, the material recognition model outputs the weight of the classification result, and different classification results correspond to different materials of the particle. If the material recognition model outputs multiple weights, the embodiment can take the material of the particle corresponding to the classification result with the highest weight as the final material of the particle. Wherein, the embodiment can use a genetic algorithm back propagation neural network (GA-BP neural network) to construct the material recognition model, for example, the embodiment sets the input layer of the above material recognition model to have 3 nodes, the hidden layer to have 8 nodes, and the output layer to have 4 nodes. Wherein, each node of the output layer corresponds to a classification result, for example, [1 0 0 0] represents that the material of the particle is rubber, [0 1 0 0] represents that the material of the particle is aluminum particle, [0 0 1 0] represents that the material of the particle is copper particle, and [0 0 0 1] represents that the material of the particle is solder particle.

[0067] In addition, the weight of the material recognition model can also be optimized by combining a genetic algorithm in the embodiment, for example, in the process of identifying the material of the particle, the genetic algorithm is combined to continuously screen individuals by selection, crossover and mutation, and the weight corresponding to the fitness of the optimal individual is obtained, and then the material of the particle is more accurately determined according to the weight.

[0068] For example, after determining the material of the particle, the embodiment can also determine the source of the redundant particle in the electronic component according to the material of the particle and the pre-stored corresponding relationship between the material of the particle and the related steps, so that the staff can improve the manufacturing process or use process of the electronic component according to the obtained source of the redundant particle, thereby reducing the number of redundant particles in the electronic component. Wherein, the above pre-stored corresponding relationship between the material of the particle and the related steps can be set according to each step in the production and use process of the electronic component, and the pre-set corresponding relationship between the material of the particle and the related steps corresponding to different electronic components is different.

[0069] The particle material recognition method provided by the embodiment can reduce the influence of various signals such as noise signals, interference signals and component signals on the reconstructed sound signal, improve the accuracy and real-time performance of the reconstructed sound signal, and then accurately identify the material of the particle based on the reconstructed sound signal, so that the source of the redundant particle can be accurately traced and the manufacturing process and use process of the electronic component can be improved based on the material of the particle.

[0070] To more accurately determine the pulse decision value of each frame of sound signal, and more accurately determine the plurality of pulse signal segments, the embodiment can further calculate the short-time signal-to-noise ratio corresponding to each frame of sound signal on the basis of the above-mentioned embodiment, and determine the pulse decision value of each frame of sound signal based on the short-time signal-to-noise ratio corresponding to each frame of sound signal, and determine the start point of the pulse signal segment according to the frequency band variance value corresponding to each frame of sound signal and the corresponding pulse signal start point decision value, determine the end point of the pulse signal segment according to the frequency band variance value corresponding to each frame of sound signal and the corresponding pulse signal segment end point decision value, thereby determining the plurality of pulse signal segments based on the start point and the end point of the pulse signal segment. Figure 2 is a flowchart of a particle material identification method provided by another embodiment of the present application. As shown in Figure 2 , the method in the embodiment of the present application can include:

[0071] Step 201, acquiring the original sound signal generated by particle collision, and performing frame division on the original sound signal to obtain a plurality of frames of sound signal.

[0072] The specific implementation process and principles of step 201 can refer to step 101, which will not be described here.

[0073] Step 202, determining the short-time signal-to-noise ratio corresponding to each frame of sound signal according to the frequency band variance value of each frame of sound signal and the frequency band variance value of the previous frame of sound signal of each frame of sound signal.

[0074] For example, the embodiment can determine the short-time signal-to-noise ratio SNR corresponding to each frame of sound signal according to the expression:

[0075] i .

[0076] In the formula, i is the frame number corresponding to the sound signal, i = 1, 2, 3, …, I, I is determined according to the total frame number of the sound signal, D i is the frequency band variance value corresponding to the i-th frame of sound signal, D i-1 is the frequency band variance value corresponding to the i-1-th frame of sound signal.

[0077] Optionally, in the embodiment, the short-time signal-to-noise ratio corresponding to each frame of sound signal is determined according to the frequency band variance value of each frame of sound signal and the previous frame of sound signal, that is, the short-time signal-to-noise ratio corresponding to each frame of sound signal is variable. The embodiment can more accurately determine the pulse decision value corresponding to each frame of sound signal by adaptively updating the short-time signal-to-noise ratio corresponding to each frame.

[0078] ​Step 203, determining the pulse decision value corresponding to each frame of sound signal according to the short-time signal-to-noise ratio corresponding to each frame of sound signal.

[0079] Exemplarily, the pulse decision value includes a pulse signal start point decision value and a pulse signal end point decision value. As known from the foregoing, the pulse signal start point decision value is used to determine the start point of the pulse signal segment, and the pulse signal end point decision value is used to determine the end point of the pulse signal segment. In this embodiment, the pulse signal start point decision value and the pulse signal end point decision value can be determined according to the following expressions:

[0080] determining the pulse signal start point decision value G i1 and the pulse signal end point decision value G i2 .

[0081] wherein j = 1, 2, 1 represents the pulse signal start point, 2 represents the pulse signal end point, i is the frame number corresponding to the sound signal, i = 1, 2, 3, …, I, I is determined according to the total frame number of the sound signal, SNR i is the short-time signal-to-noise ratio corresponding to the i-th frame of sound signal, D i-1 is the frequency band variance value corresponding to the i-1-th frame of sound signal of the previous frame of the i-th frame, a1 is the first preset coefficient, a2 is the second preset coefficient, and a1 is greater than a2, Y is the first preset threshold value, -Y is the second preset threshold value, and dB represents decibel.

[0082] Optionally, in order to ensure that the pulse signal start point decision value is greater than the pulse signal end point decision value, the first preset coefficient is set to be greater than the second preset coefficient, for example, the first preset coefficient can be set to 3, and the second preset coefficient can be set to 1.5. Of course, the first preset coefficient and the second preset coefficient can also be set to other numerical values, which are not limited here. The first preset threshold value and the second preset threshold value can be obtained through a large amount of data test fitting, for example, the first preset threshold value can be set to 40, and the second preset threshold value can be set to -40.

[0083] Exemplarily, in this embodiment, the pulse decision value corresponding to each frame of sound signal is determined according to the short-time signal-to-noise ratio corresponding to each frame of sound signal and the preset threshold value, that is, the pulse signal start point decision value and the pulse signal end point decision value corresponding to different frames of sound signal are variable. In this embodiment, the pulse decision value corresponding to each frame is updated adaptively, so that the pulse signal segment of the particle can be more accurately determined in the subsequent process of determining the pulse signal segment according to the pulse decision value.

[0084] Step 204, determining the start point of the pulse signal segment according to the frequency band variance value corresponding to each frame of sound signal and the corresponding pulse signal start point decision value, and determining the end point of the pulse signal segment according to the frequency band variance value corresponding to each frame of sound signal and the corresponding pulse signal end point decision value.

[0085] Wherein, from the foregoing, the pulse decision value includes the pulse signal start point decision value and the pulse signal end point decision value.

[0086] In one possible implementation, the embodiment can include A1 to A2 when determining the start point of the pulse signal segment and determining the end point of the pulse signal segment.

[0087] A1, starting from the second frame of sound signals in each frame of sound signals, if the frequency band variance values corresponding to the continuous preset number of frames of sound signals are all greater than the pulse signal start point decision value corresponding to the corresponding frame of sound signals, the next frame of sound signals of the continuous preset number of frames of sound signals is taken as the start point of the pulse signal segment.

[0088] A2, if the frequency band variance values corresponding to the continuous preset number of frames of sound signals are all less than the pulse signal end point decision value corresponding to the corresponding frame of sound signals, the next frame of sound signals of the continuous preset number of frames of sound signals is taken as the end point of the pulse signal segment.

[0089] For example, in the embodiment, the preset number of frames can be three frames, and of course the preset number of frames can also be set to other values as needed. The embodiment starts from the second frame of sound signals in multiple frames of sound signals, and determines each frame of sound signals in order from small to large frame number. If the frequency band variance values corresponding to the continuous three frames of sound signals are all greater than the pulse signal start point decision value corresponding to the continuous three frames of sound signals, the next frame of sound signals of the continuous three frames of sound signals is taken as the start point of a pulse signal segment. Wherein, the determination of the continuous three frames of sound signals can avoid the situation of misjudging noise signals, interference signals and component signals as pulse signals, so as to accurately determine the start point of the pulse signal segment.

[0090] For example, starting from the second frame of sound signals in multiple frames of sound signals, if the frequency band variance value corresponding to the third frame of sound signals is greater than the pulse signal start point decision value corresponding to the third frame of sound signals, the frequency band variance value corresponding to the fourth frame of sound signals is greater than the pulse signal start point decision value corresponding to the fourth frame of sound signals, and the frequency band variance value corresponding to the fifth frame of sound signals is greater than the pulse signal start point decision value corresponding to the fifth frame of sound signals, the sixth frame of sound signals is determined as the start point of a pulse signal segment.

[0091] For example, the embodiment starts from the second frame of sound signals in multiple frames of sound signals, and determines each frame of sound signals in order from small to large frame number. If the frequency band variance values corresponding to the continuous three frames of sound signals are all less than the pulse signal end point decision value corresponding to the continuous three frames of sound signals, the next frame of sound signals of the continuous three frames of sound signals is taken as the end point of a pulse signal segment. Wherein, the determination of the continuous three frames of sound signals can avoid the situation of misjudging noise signals, interference signals and component signals as pulse signals, so as to accurately determine the end point of the pulse signal segment.

[0092] For example, starting from the second frame of the multi-frame acoustic signal, if the frequency band variance value corresponding to the thirteenth frame of acoustic signal is less than the pulse signal end point decision value corresponding to the thirteenth frame of acoustic signal, the frequency band variance value corresponding to the fourteenth frame of acoustic signal is less than the pulse signal end point decision value corresponding to the fourteenth frame of acoustic signal, and the frequency band variance value corresponding to the fifteenth frame of acoustic signal is less than the pulse signal end point decision value corresponding to the fifteenth frame of acoustic signal, it is determined that the sixteenth frame of acoustic signal is the end point of a pulse signal segment.

[0093] For example, starting from the second frame of the multi-frame acoustic signal, in the order of frame number from small to large, the start point of the pulse signal segment and the end point of the pulse signal segment are determined for each frame of acoustic signal in turn until the determination of the above multi-frame acoustic signal is completed, and the start point of the plurality of pulse signal segments and the end point of the plurality of pulse signal segments can be obtained.

[0094] Step 205, based on the start point of the pulse signal segment and the end point of the pulse signal segment, determining a plurality of pulse signal segments of the particle, and connecting the plurality of pulse signal segments in turn to obtain a reconstructed acoustic signal of the particle.

[0095] For example, starting from the second frame of the multi-frame acoustic signal, in the order of frame number from small to large, the start point of the pulse signal segment and the end point of the pulse signal segment are determined for each frame of acoustic signal in turn until the determination of the above multi-frame acoustic signal is completed, and the start point of the plurality of pulse signal segments and the end point of the plurality of pulse signal segments can be obtained.

[0096] Optionally, after obtaining the plurality of pulse signal segments of the particle, the first pulse signal segment, the second pulse signal segment, …, the Mth pulse signal segment are connected in turn to obtain the reconstructed acoustic signal of the particle. The Mth pulse signal segment is the last obtained pulse signal segment.

[0097] In a possible implementation, after determining the plurality of pulse signal segments of the particle according to the pulse decision value of each frame of acoustic signal and the frequency band variance value corresponding to each frame of acoustic signal, the embodiment can further include B1 to B2.

[0098] B1, judging whether the pulse duration corresponding to the pulse signal segment in the plurality of pulse signal segments is less than a preset duration.

[0099] B2, if the pulse duration corresponding to the pulse signal segment in the plurality of pulse signal segments is less than the preset duration, deleting the pulse signal segment with the pulse duration less than the preset duration in the plurality of pulse signal segments, and taking the remaining pulse signal segments in the plurality of pulse signal segments as the plurality of pulse signal segments.

[0100] For example, the embodiment determines whether the pulse duration corresponding to each pulse signal segment obtained is less than a preset duration, and determines the pulse signal segment with the pulse duration less than the preset duration as an interference signal, a component signal or a noise signal, deletes the pulse signal segment with the pulse duration less than the preset duration, and takes the remaining pulse signal segments as the plurality of pulse signal segments to construct the reconstructed acoustic signal of the particle according to the plurality of pulse signal segments. The preset duration is set according to the duration corresponding to the interference signal, the component signal or the noise signal.

[0101] In the embodiment, the plurality of pulse signal segments of the obtained particle are judged again, that is, it is further judged whether the plurality of pulse signal segments of the obtained particle are pulse signals, and the pulse signal segments that are not pulse signals in the plurality of pulse signal segments are deleted, and the remaining pulse signal segments are taken as the plurality of pulse signal segments, so as to accurately construct the reconstructed acoustic signal of the particle.

[0102] In step 206, the feature of the reconstructed acoustic signal is extracted to obtain a feature vector of the particle, and the feature vector is input into a preset material recognition model to obtain the material of the particle.

[0103] The specific implementation process and principle of step 206 can be referred to step 104, which will not be described here.

[0104] The particle material recognition method provided by the embodiment can frame the original acoustic signal, calculate the corresponding adaptive short-time signal-to-noise ratio of each frame of acoustic signal in the obtained plurality of frames of acoustic signal, calculate the corresponding adaptive pulse decision value, determine the plurality of pulse signal segments for constructing the reconstructed acoustic signal according to the pulse decision value and the preset rule, and improve the accuracy and real-time performance of the reconstructed acoustic signal, so as to accurately recognize the material of the particle based on the reconstructed acoustic signal, so as to accurately trace the source of the particle and improve the manufacturing process and use procedure of the electronic component.

[0105] It should be understood that the size of the serial number of each step in the above embodiment does not mean the execution order, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiment.

[0106] Figure 3 is a structural schematic diagram of a particle material identification device provided by an embodiment of the present application. As shown in the figure, Figure 3 the particle material identification device provided by the embodiment can include an acquisition module 301, a calculation module 302, a construction module 303, and an identification module 304.

[0107] The acquisition module 301 is configured to acquire original sound signals generated by particle collision, frame the original sound signals, and obtain multiple frames of sound signals.

[0108] The calculation module 302 is configured to calculate frequency band variance values corresponding to each frame of sound signal in the multiple frames of sound signals, and determine pulse decision values of each frame of sound signal according to the frequency band variance values corresponding to each frame of sound signal.

[0109] The construction module 303 is configured to determine multiple pulse signal segments of the particle according to the pulse decision values of each frame of sound signal and the frequency band variance values corresponding to each frame of sound signal, and construct a reconstructed sound signal of the particle based on the multiple pulse signal segments.

[0110] The identification module 304 is configured to extract features of the reconstructed sound signal to obtain a feature vector of the particle, and input the feature vector into a pre-set material identification model to obtain a material of the particle, wherein the material identification model inputs the feature vector of the particle and outputs the material of the particle.

[0111] Optionally, the calculation module 302 is specifically configured to determine a short-time signal-to-noise ratio corresponding to each frame of sound signal according to the frequency band variance value of each frame of sound signal and the frequency band variance value of a previous frame of sound signal of each frame of sound signal, and determine a pulse decision value corresponding to each frame of sound signal according to the short-time signal-to-noise ratio corresponding to each frame of sound signal.

[0112] Optionally, the pulse decision value includes a pulse signal start point decision value and a pulse signal end point decision value; the construction module 303 is specifically configured to determine a start point of a pulse signal segment according to the frequency band variance value corresponding to each frame of sound signal and the corresponding pulse signal start point decision value, determine an end point of the pulse signal segment according to the frequency band variance value corresponding to each frame of sound signal and the corresponding pulse signal end point decision value, determine the multiple pulse signal segments of the particle based on the start point of the pulse signal segment and the end point of the pulse signal segment, and connect the multiple pulse signal segments in sequence to obtain the reconstructed sound signal of the particle.

[0113] Optionally, the constructing module 303 is further configured to: starting from a second frame of the sound signals, if the frequency band variance values corresponding to a continuous preset number of frames of the sound signals are all greater than the start point decision value of the impulse signal corresponding to the frame of sound signals, then taking a next frame of sound signals of the continuous preset number of frames of the sound signals as the start point of the impulse signal segment; if the frequency band variance values corresponding to the continuous preset number of frames of the sound signals are all less than the end point decision value of the impulse signal corresponding to the frame of sound signals, then taking a next frame of sound signals of the continuous preset number of frames of the sound signals as the end point of the impulse signal segment.

[0114] Optionally, the constructing module 303 is further configured to: determining whether there is an impulse signal segment corresponding to a pulse duration less than a preset duration in the plurality of impulse signal segments; if there is an impulse signal segment corresponding to a pulse duration less than a preset duration in the plurality of impulse signal segments, then deleting the impulse signal segment corresponding to the pulse duration less than the preset duration in the plurality of impulse signal segments, and taking the remaining impulse signal segments in the plurality of impulse signal segments as the plurality of impulse signal segments.

[0115] Optionally, the identifying module 304 is configured to: performing a plurality of feature extractions on the reconstructed sound signal to obtain a plurality of feature quantities, and constructing a feature vector of the particle according to the plurality of feature quantities; performing dimension reduction processing on the feature vector to obtain a dimension-reduced feature vector; and inputting the dimension-reduced feature vector into the material identification model to obtain the material of the particle.

[0116] It should be noted that the information interaction between the above apparatuses / units, the execution process, and the like, are based on the same concept as the method embodiments of the present application, and specific functions and technical effects brought by the same can be referred to the method embodiments part, which will not be repeated here.

[0117] Figure 4 is a structural schematic diagram of an electronic device provided by an embodiment of the present application. As shown in Figure 4 the electronic device 400 of this embodiment includes a processor 410, a memory 420, and the memory 420 stores a computer program 421 executable on the processor 410. The processor 410 implements the steps in any of the above method embodiments when executing the computer program 421, such as Figure 1 steps 101 to 104 shown in the figure. Alternatively, the processor 410 implements the functions of the modules / units in the above apparatus embodiments when executing the computer program 421, such as Figure 3 the functions of the modules 301 to 304 shown in the figure.

[0118] For example, the computer program 421 can be divided into one or more modules / units, one or more modules / units are stored in the memory 420 and executed by the processor 410 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 421 in the electronic device 400.

[0119] Those skilled in the art can understand that, Figure 4 The electronic device is only an example and does not constitute a limitation on the electronic device, and can include more or fewer components than the illustration, or combine certain components, or different components, such as input / output devices, network access devices, buses, etc.

[0120] The processor 410 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 gate 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.

[0121] The memory 420 can be an internal storage unit of the electronic device, such as a hard disk or memory of the electronic device, and can also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. The above-mentioned memory 420 can include both the internal storage unit and the external storage device of the electronic device. The above-mentioned memory 420 is used to store computer programs and other programs and data required by the electronic device. The memory 420 can also be used to temporarily store data that has been output or will be output.

[0122] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be realized in the form of hardware or software. In addition, the specific names of each functional unit and module are only for easy distinction, and do not limit the protection scope of the present application. The specific working process of the units and modules in the above system can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0123] In the above embodiments, the description of each embodiment has its own emphasis, and the parts not described or recorded in detail in a certain embodiment can be referred to the related description of other embodiments.

[0124] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized in electronic hardware or a combination of computer software and electronic hardware. Whether the functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0125] In the embodiments provided by the present application, it should be understood that the disclosed apparatus / equipment and method can be implemented by other ways. For example, the above-mentioned apparatus / equipment embodiments are only schematic, and the division of the modules or units is only a logical function division, and there can be another division way in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0126] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0127] In addition, each function unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software function unit.

[0128] The integrated module / unit, if realized in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such 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 computer program is executed by a processor, the steps of each method embodiment can be implemented. The computer program includes computer program code, which can be in the form of source code, object code, executable files or some intermediate forms. 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.

[0129] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A particle material identification method, characterized by, The method comprises: obtaining an original sound signal generated by particle collision, frame dividing the original sound signal to obtain a plurality of frames of sound signals; calculating a frequency band variance value corresponding to each frame of sound signal in the plurality of frames of sound signals, and determining a pulse decision value of each frame of sound signal according to the frequency band variance value corresponding to the frame of sound signal; determining a plurality of pulse signal segments of the particle according to the pulse decision value of each frame of sound signal and the frequency band variance value corresponding to each frame of sound signal, and constructing a reconstructed sound signal of the particle based on the plurality of pulse signal segments; extracting a feature vector of the particle by performing feature extraction on the reconstructed sound signal, and inputting the feature vector into a pre-set material recognition model to obtain a material of the particle, wherein the material recognition model inputs a feature vector of a particle and outputs a material of the particle; wherein the determining of the pulse decision value of each frame of sound signal according to the frequency band variance value corresponding to the frame of sound signal comprises: determining a short-time signal-to-noise ratio corresponding to each frame of sound signal according to a frequency band variance value of each frame of sound signal and a frequency band variance value of a previous frame of sound signal of the frame of sound signal; determining a pulse decision value corresponding to each frame of sound signal according to the short-time signal-to-noise ratio corresponding to the frame of sound signal; the pulse decision value comprises a pulse signal start point decision value and a pulse signal end point decision value; the determining of the pulse decision value corresponding to each frame of sound signal according to the short-time signal-to-noise ratio corresponding to the frame of sound signal comprises: according to the expression: determining a start decision value of the pulse signal corresponding to each frame of sound signal and a terminal decision value of the pulse signal ; in, , Indicates the start point of the pulse signal. Indicates the end point of the pulse signal. The number of frames corresponding to the sound signal. i =1,2,3,…,I, where I is determined based on the total number of frames in the audio signal. For the first The short-time signal-to-noise ratio corresponding to the frame audio signal. For the first The frame before the frame, the first The frequency band variance value corresponding to the frame audio signal. The first preset coefficient, It is the second preset coefficient, and Greater than Y is the first preset threshold. Y is the second preset threshold. It represents decibels.

2. The particle material identification method according to claim 1, characterized by, the determining of the short-time signal-to-noise ratio corresponding to each frame of sound signal according to the frequency band variance value of each frame of sound signal and the frequency band variance value of the previous frame of sound signal of the frame of sound signal comprises: according to the expression: determining a short-time signal-to-noise ratio corresponding to each frame of the acoustic signal ; In the formula, is the frame number corresponding to the sound signal, i = 1, 2, 3, …, I, I is determined according to the total frame number of the sound signal, is the frame number corresponding to the sound signal, is the frequency band variance value corresponding to the sound signal of the i th frame, is the frequency band variance value corresponding to the sound signal of the i th frame, is the previous frame of the i th frame, and is the frequency band variance value corresponding to the sound signal of the i th frame.

3. The particle material identification method according to claim 1, characterized by, the pulse decision value comprises a pulse signal start point decision value and a pulse signal end point decision value; the determining of the plurality of pulse signal segments of the particle according to the pulse decision value of each frame of sound signal and the frequency band variance value corresponding to each frame of sound signal, and the construction of the reconstructed sound signal of the particle based on the plurality of pulse signal segments, comprises: determining a start point of a pulse signal segment according to the frequency band variance value corresponding to each frame of sound signal and a corresponding pulse signal start point decision value, and determining an end point of the pulse signal segment according to the frequency band variance value corresponding to the frame of sound signal and a corresponding pulse signal end point decision value; determining the plurality of pulse signal segments of the particle based on the start point of the pulse signal segment and the end point of the pulse signal segment, and connecting the plurality of pulse signal segments in sequence to obtain the reconstructed sound signal of the particle.

4. The particle material identification method according to claim 3, characterized by, the determining of the start point of the pulse signal segment according to the frequency band variance value corresponding to each frame of sound signal and the corresponding pulse signal start point decision value, and the determining of the end point of the pulse signal segment according to the frequency band variance value corresponding to the frame of sound signal and the corresponding pulse signal end point decision value, comprises: If the frequency band variance values corresponding to the continuous preset number of frames of sound signals are all greater than the pulse signal start point decision value corresponding to the respective frames of sound signals, the next frame of sound signal of the continuous preset number of frames of sound signals is taken as the start point of the pulse signal segment; If the frequency band variance values corresponding to the continuous preset number of frames of sound signals are all less than the pulse signal end point decision value corresponding to the respective frames of sound signals, the next frame of sound signal of the continuous preset number of frames of sound signals is taken as the end point of the pulse signal segment.

5. The granular material identification method according to any one of claims 1 to 4, characterized in that, After the plurality of pulse signal segments of the particle are determined according to the pulse decision values of the respective frames of sound signals and the frequency band variance values corresponding to the respective frames of sound signals, the method further comprises: determining whether the pulse duration corresponding to any pulse signal segment in the plurality of pulse signal segments is less than a preset duration; If the pulse duration corresponding to any pulse signal segment in the plurality of pulse signal segments is less than the preset duration, the pulse signal segment with the pulse duration less than the preset duration is deleted from the plurality of pulse signal segments, and the remaining pulse signal segments of the plurality of pulse signal segments are taken as the plurality of pulse signal segments.

6. The granular material identification method according to any one of claims 1 to 4, characterized in that, The feature extraction on the reconstructed sound signal obtains a feature vector of the particle, and the feature vector is input into a preset material recognition model to obtain the material of the particle, comprising: a plurality of feature quantities are obtained by performing a plurality of feature extractions on the reconstructed sound signal, and a feature vector of the particle is constructed according to the plurality of feature quantities; dimensionality reduction processing is performed on the feature vector to obtain a dimensionally reduced feature vector; the dimensionally reduced feature vector is input into the material recognition model to obtain the material of the particle.

7. A particle material identification device, characterized in that, comprising: an acquisition module configured to acquire an original sound signal generated by particle collision, frame the original sound signal to obtain a plurality of frames of sound signals; a calculation module configured to calculate a frequency band variance value corresponding to each frame of sound signal in the plurality of frames of sound signals, and determine a pulse decision value of each frame of sound signal according to the frequency band variance value corresponding to the frame of sound signal; a construction module configured to determine a plurality of pulse signal segments of the particle according to the pulse decision values of the respective frames of sound signals and the frequency band variance values corresponding to the respective frames of sound signals, and construct a reconstructed sound signal of the particle based on the plurality of pulse signal segments; an identification module configured to perform feature extraction on the reconstructed sound signal to obtain a feature vector of the particle, and input the feature vector into a preset material recognition model to obtain the material of the particle, wherein the material recognition model inputs the feature vector of the particle and outputs the material of the particle; wherein the calculation module is further configured to determine a short-time signal-to-noise ratio corresponding to each frame of sound signal according to the frequency band variance value of the frame of sound signal and the frequency band variance value of a previous frame of sound signal of the frame of sound signal, and determine a pulse decision value corresponding to the frame of sound signal according to the short-time signal-to-noise ratio corresponding to the frame of sound signal; the pulse decision value comprises a pulse signal start point decision value and a pulse signal end point decision value; and is further configured to determine the pulse signal start point decision value according to the expression: determining a pulse signal start decision value corresponding to the each frame of sound signal and a pulse signal end decision value ; in, , Indicates the start point of the pulse signal. Indicates the end point of the pulse signal. The number of frames corresponding to the sound signal. i =1,2,3,…,I, where I is determined based on the total number of frames in the audio signal. For the first The short-time signal-to-noise ratio corresponding to the frame audio signal. For the first The frame before the frame, the first The frequency band variance value corresponding to the frame audio signal. The first preset coefficient, It is the second preset coefficient, and Greater than Y is the first preset threshold. Y is the second preset threshold. It represents decibels.

8. An electronic device comprising a memory and a processor, said memory having stored therein a computer program operable on said processor, characterized in that, The processor implements the particle material identification method according to any one of claims 1-6 when executing the computer program.

Images