WO2019127088A1

WO2019127088A1 - Snore recognition method and snore-stopping device

Info

Publication number: WO2019127088A1
Application number: PCT/CN2017/118950
Authority: WO
Inventors: 齐奇
Original assignee: 深圳和而泰数据资源与云技术有限公司
Priority date: 2017-12-27
Filing date: 2017-12-27
Publication date: 2019-07-04
Also published as: CN108697328A; CN108697328B

Abstract

A snore recognition method and a snore-stopping device. The method comprises: sampling a snore audio signal; carrying out analog-to-digital (ADC) conversion on the sampled signal; forming an audio frame based on a time domain and the sampled signal, and carrying out frame shifting on the audio frame to carry out snore analysis; carrying out window function framing, and carrying out a Fourier transform on an M1 sequence obtained by means of framing to transform same into a frequency domain, and summating the low-frequency energy of the frequency domain spectrum; forming, by means of the summated value of the low-frequency energy, an N sequence in the time domain and carrying out snore prediction on the N sequence, and forming an M2 sequence by means of an N sequence that is predicted to be qualified; and carrying out a Fourier transform on the M2 sequence to transform same into the frequency domain and matching pre-set snore parameters according to the spectral peak features of the M2 sequence, and then outputting a monitored snore signal.

Description

Beep recognition method and anti-snoring device

Technical field

The present application relates to a health monitoring device, and more particularly to a snoring recognition method and a snoring device.

Background technique

When a normal person sleeps, it will be deformed due to the relaxation of the soft tissue near the throat, which leads to the narrowing of the upper airway and hinders the smoothness of the breathing, which is a common snoring.

If snoring becomes a sleep habit, the snoring that affects breathing will accumulate health risks. For example, obstructive sleep apnea (OSA), soft tissue relaxation near the throat causes upper airway obstruction, and narrowing of the airway leads to apnea during sleep. Or OSAHS, obstructive sleep apnea hypopnea syndrome (obstructive sleep Apnea-hypopnea syndrome). OSAHS refers to apnea and hypopnea caused by airway collapse obstruction during sleep, accompanied by snoring, sleep structure disorder, frequent occurrence of oxygen saturation, and daytime sleepiness. Apnea refers to the cessation of nasal and nasal airflow during sleep ≥10s. Hypnea refers to the reduction of respiratory airflow during sleep more than 30% above the baseline level, with 3% oxygen saturation (SaO) or arousal. Sleep apnea syndrome, more than 30 apneas occurred during 7 hours of sleep, each time the airflow is stopped for more than 10s (including 10s), or the average number of sleep apnea hypopneas (respiratory disorder index) exceeds 5 times. Clinical syndrome that causes chronic hypoxemia and hypercapnia. Can be divided into central type, blocking type and mixed type.

Snoring is the early stage of sleep apnea syndrome (obstructive sleep apnea), which belongs to different periods of the same disease. If you do not promptly intervene, you will develop sleep apnea syndrome in a few decades.

Existing anti-snagging devices, such as anti-snag pads, include a controller, a click recognition module, a judging module, and a reminder mechanism. The humming recognition module monitors the snoring condition in real time while the user is asleep, collects ambient audio and recognizes snoring. After the buzz is recognized, the snoring information is sent to the controller, and the judging module of the controller drives the reminding mechanism to move when the snoring is determined to be snoring, and touches the user's body to stop the snoring.

However, the prior art snoring recognition generally uses Hidden Markov Model (HMM) speech recognition to perform snoring recognition. The model is more complex, computationally intensive, requires a powerful computing chip, and is not suitable for the development trend of miniaturization/portability of anti-snagging equipment.

Therefore, the prior art needs to be improved to solve the technical problems that arise.

Summary of the invention

The application provides a snoring recognition method and a snoring device, and the snoring recognition is accurate, and the calculation amount of the snoring device is greatly reduced, and can adapt to the demand for miniaturization of the stagnation device, including the shackle bracelet, the snoring earphone, the snoring pillow, and the like. .

In a first aspect, the embodiment of the present application provides a method for identifying a click, including:

Beep audio signal sampling;

Performing analog-to-digital (ADC) conversion on the sampled signal;

Forming an audio frame based on the time domain and the sampled signal, and performing a click analysis on the audio frame shift frame;

The window function is framed, and the M1 sequence obtained by the frame is Fourier transformed into the frequency domain, and the low frequency energy of the frequency domain spectrum is summed;

The low-frequency energy and value form an N-sequence in the time domain, and the N-sequence is pre-judged, and the qualified N-sequence forms a M2 sequence;

The M2 sequence is Fourier transformed to the frequency domain, and the preset chirp parameters are matched according to the spectral peak characteristics of the M2 sequence, and the monitoring chirp signal is output.

The method for identifying the click sound further includes: the audio frame is moved by the first-in first-out frame at a 1/X rate, and X is an arbitrary number; wherein the method further includes an audio signal pre-emphasis process.

Preferably, the beating pre-judging of the pair of N sequences comprises: calculating a maximum minimum value and a maximum difference value of the N-sequence data; wherein the parameter predicted by the click sound comprises a zero-crossing rate, a waveform amplitude, and a zero-crossing gap.

In order to improve the accuracy of the click recognition, the method further comprises: normalizing the M2 sequence and truncating the average processing.

Specifically, the window function is a Hanning window function (Haning) or a Hamming window function (Hamming) or a Blackman window function (Blackman).

In a second aspect, the embodiment of the present application further provides a stagnation device, including a microcontroller, an analog-to-digital conversion circuit (ADC), and a sound pickup circuit, the sound pickup circuit is configured to sample a click audio signal, and the analog-to-digital conversion circuit And performing analog-to-digital (ADC) conversion on the sampling signal, further comprising: a click recognition module, configured to form an audio frame based on the time domain and the sampled signal, perform a click analysis on the audio frame shift frame; Function framing, transforming the M1 sequence obtained by the framing into the frequency domain, and summing the low-frequency energy of the frequency domain spectrum; for the low-frequency energy and value to form the N-sequence in the time domain, and the frequency-domain N-sequence By pre-judging, the pre-determined N sequence forms an M2 sequence; it is also used to perform Fourier transform on the M2 sequence to the frequency domain, and matches the preset chirp parameters according to the spectral peak characteristics of the M2 sequence, and outputs a monitoring chirp signal; The stagnation mechanism, wherein the microcontroller receives the monitoring squeak signal, and then activates the stagnation mechanism to touch the monitored user.

The click recognition module is further configured to move the frame to the audio frame at a 1/X rate, where X is an arbitrary number; the click recognition module is further configured to pre-emphasize the audio signal.

Preferably, the click recognition module is further configured to: when the sound is predicted, calculate a maximum minimum value and a maximum difference value of the N sequence data; wherein the parameters of the click sound prediction include a zero crossing rate, a waveform amplitude, and a zero crossing gap. .

In order to improve the accuracy of the click recognition, the click recognition module is further configured to perform normalization processing and truncation averaging processing on the M2 sequence.

In a third aspect, the embodiment of the present application further provides an electronic device, including:

At least one processor; and,

a memory communicatively coupled to the at least one processor; wherein

The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method as described above.

In a fourth aspect, the embodiment of the present application further provides a non-transitory computer readable storage medium, where the computer readable storage medium stores computer executable instructions for causing a computer to execute the above The method described.

In a fifth aspect, the embodiment of the present application further provides a computer program product, where the computer program product includes a computer program stored on a non-transitory computer readable storage medium, the computer program includes program instructions, when When the program instructions are executed by the computer, the computer is caused to perform the method as described above.

The method and device for detecting the click sound provided by the embodiment of the present application divides the click recognition after the window function is divided into the energy differentiation stage and the click characteristic comparison stage. In the quiet sleep environment, the recognition model is simplified, and the calculation amount is greatly reduced, which can be realized. Miniaturization and portable design of anti-snoring products; at the same time, in a quiet sleep environment, 100% accurate beep monitoring can be achieved for buzzing above 60 decibels.

DRAWINGS

The one or more embodiments are exemplified by the accompanying drawings in the accompanying drawings, and FIG. The figures in the drawings do not constitute a scale limitation unless otherwise stated.

1 is a flowchart of an embodiment of a method for identifying a click sound provided by an embodiment of the present application;

2 is a product module diagram of a stagnation device provided by an embodiment of the present application;

3 is a schematic diagram of the action of band pass filtering of the click recognition method provided by the embodiment of the present application;

4 is a frequency energy relationship diagram of pre-emphasis processing of the click recognition method provided by the embodiment of the present application;

FIG. 5 is a general flowchart of a method for identifying a click sound provided by an embodiment of the present application; FIG.

6 is a schematic diagram of the energy summation of the M1 sequence in the click recognition method provided by the embodiment of the present application;

7 is a schematic diagram of a click recognition of an M2 sequence in a click recognition method provided by an embodiment of the present application;

8 is a flowchart of a click characteristic comparison of the click recognition method provided by the embodiment of the present application;

FIG. 9 is a schematic diagram showing the hardware structure of an electronic device for performing a click recognition method according to an embodiment of the present application.

Detailed ways

The embodiment will be described in detail below with reference to the accompanying drawings and embodiments.

The present application relates to a click recognition method and a stop device using the same. The snoring recognition method and device divides the snoring recognition into an energy accumulation phase and a snoring feature comparison phase, and the snoring recognition model is simplified, the calculation amount is greatly reduced, and the miniaturization and portable design of the stagnation product can be realized; meanwhile, in a quiet sleep environment For a beep of more than 60 decibels, 100% accurate bark monitoring can be achieved.

The energy accumulation phase and the click characteristic comparison phase are all completed by the click recognition module.

Referring to FIG. 2, a product module diagram of the device is shown. The stagnation device of the present embodiment includes a microcontroller 10, an analog-to-digital conversion circuit (ADC) 30, a click recognition module 40, a sound pickup circuit 20, and a stop mechanism. 50. Among them, ADC stands for Analog to Digital Converter.

The microcontroller 10 controls the operation of the entire stagnation device. The sound pickup circuit 20 samples the click audio signal. In the present embodiment, the sound pickup circuit 20 acquires a waveform signal (analog signal) of the click sound through the microphone and its associated circuit. The analog to digital conversion circuit performs analog to digital (ADC) conversion on the sampled signal.

The anti-snoring mechanism 50 can be implemented in various ways. For example, in the sleep monitoring wristband embodiment, the anti-snoring mechanism should be designed to be small and the wake-up power is moderate. In the sleep monitoring mattress embodiment, the size and the pressure awakening force of the stagnation mechanism are relatively larger than the design. The connection to the microcontroller 10 can include a motor, a gear set, and a touch-pressure structure that transmits power from the motor and the gear set under the command of the microcontroller 10, through the touch-pressure structure to the detected sleep user. Perform a wake-up action.

The click recognition module forms an audio frame based on the time domain and the sampled signal. The click recognition module performs a click analysis on the audio frame shift frame. The click recognition module also advances the first-in-first-out frame of the audio frame at a rate of 1/X (where X is an arbitrary number). In order to clarify the technical solution, in the embodiment, the value of X is 2.

The click recognition module selects the window function to divide the frame, and then transforms the M1 sequence obtained by the frame into the frequency domain, and sums the low frequency energy of the frequency domain spectrum to complete the low frequency energy accumulation phase of the click recognition.

Then, the click recognition module combines the low frequency energy and the value into an N sequence in the time domain, and then completes the click prediction of the N sequence, and predicts that the qualified back N sequence forms the M2 sequence. The click recognition module further performs Fourier transform on the M2 sequence to the frequency domain, and matches the preset click parameters according to the spectral peak characteristics of the M2 sequence, and outputs a monitoring click signal of whether the sleep user is snoring.

After receiving the monitoring click signal, the microcontroller 10 activates the stop mechanism 50 to touch the monitored user and wake up the sleep user to stop snoring.

The click recognition module of the embodiment of the present application is simple and fast to process audio signals, and only band pass filter calculation, windowing calculation, and fast Fourier transform (FFT) occupy relatively more computing resources, but the above calculation can be implemented in a common MCU. .

The click recognition module starts the click analysis after the frame is moved at the 1/X rate. In order to clarify the technical solution, in the embodiment, the value of X is 2. That is, the data M1 sequence is 1/2 frame shifted, assuming that the length of the M1 sequence is m, that is, the M1 sequence of length m is sequentially shifted by m/2 length.

The microcontroller 10 converts the sound waveform signal (analog signal) acquired by the sound pickup circuit 20 into a digital signal through an analog-to-digital conversion circuit 30 (ADC), and then performs the click recognition calculation and processing by the click recognition module 40. The discrimination result is output to the microcontroller 10.

The click recognition module 40 calculates the maximum and minimum values of the N-series data during the pre-judgment in order to facilitate the high-frequency determination. In addition, the click prediction is performed. The parameters used in the steps include the zero-crossing rate, the amplitude of the waveform, and the zero-crossing gap, eliminating the signal that is initially judged to be non-squeaky.

In the click-to-speech feature comparison stage, in order to facilitate module judgment and calculation, the click recognition module 40 is further configured to perform normalization processing and truncation averaging processing on the M2 sequence.

The window function used in this application may be a Hanning window function or a Hamming window function or a Blackman window function (Blackman).

For specific description, in the embodiment of the present application, the framing signal is processed by a Hanning window to obtain a short-time audio M1 sequence, the length of the short-time audio M1 sequence is m, multiplied by the element m by the Hanning window. The coefficient of the filter. This window function can be considered to be replaced with other window functions. When there is only a Fast Fourier Transform (FFT) calculation, there is no spectrum leakage.

Referring to FIG. 1 and FIG. 5 together, a main flowchart and a detailed flowchart of an embodiment of a method for identifying a click sound according to an embodiment of the present application are shown.

In order to ensure the recognition accuracy, the click recognition method is divided into an energy discrimination phase 100 and a click feature comparison phase 300, wherein a click prediction step 200 is set in the middle of the energy discrimination phase 100 and the click feature comparison phase 300. Both the energy discrimination phase 100 and the click feature comparison phase 300 are performed in the frequency domain. The specific method is as follows:

Step S101: sampling the chirp audio signal. In this embodiment, the chirp signal sampling uses a hardware microphone and a related circuit;

Step S103: Perform analog-to-digital (ADC) conversion on the sampling signal, and set the analog-to-digital conversion circuit 30 to perform analog-to-digital (ADC) conversion on the sampling signal;

Step S106: forming an audio frame based on the time domain and the sampling signal, shifting the frame of the audio frame, and starting the energy discrimination phase 100 of the click analysis.

First, the window function is framed, wherein the frame of the short-time audio data M1 is obtained by moving the frame of the audio frame and adopting a window function, and the framed signal is processed by Hanning window. The length of the short-term audio M1 sequence is m, multiplied by the element m by the coefficient of the Hanning window filter. This window function can be replaced by other window functions, and there is no spectrum leakage when fast Fourier transform (FFT) calculation is required.

After the audio frame is formed, the click recognition method includes performing band pass filtering on the sampled audio frame. A schematic diagram of the action of the bandpass filtering of the click recognition method shown in FIG. 3, the bandpass filtering is performed on the digital bandpass filtering of the audio signal to preserve the signal of the chirp voice frequency component, and the DC signal low frequency signal A (non-speech) Frequency range) and high frequency signal B (non-speech frequency range) are filtered out.

The frame shift in the step S106 is for an audio frame, and the audio frame is FIFO framed at a 1/X rate; the short-time audio M1 sequence can be 1/2 framed or 1/x framed (where x For any number), the data needs to be met continuously for the first-in, first-out move operation.

Wherein, for the short-term audio M1 sequence, the method further comprises pre-emphasizing the audio signal of the sequence. As shown in curve C of Figure 4, in order to increase the amplitude of the spectrum in the high frequency portion, the click recognition module is also used to pre-emphasize the audio signal. The pre-emphasis processing of the audio signal uses a digital high-pass filter for boosting the amplitude of the high-frequency portion of the speech signal to flatten the spectrum.

Step S108: Fourier transforming the M1 sequence subjected to pre-emphasis processing to the frequency domain, and summing the low-frequency energy of the frequency domain spectrum; refer to the energy summation diagram of the M1 sequence shown in FIG. The energy is mainly concentrated in the low frequency part, and the long-term energy signal sequence is composed by acquiring the short-time low-frequency signal energy. Perform a spectrum analysis on the long-term energy sequence to determine whether the long-term audio is a click. Therefore, the low-frequency partial energy of the spectrum is summed here, and T represents the sum of the low-frequency energy.

As shown in FIG. 6, the M1 sequence is subjected to Fast Fourier Transform (FFT), and a short-time audio waveform sequence (a sequence of m points) is subjected to FFT conversion of m points to obtain a spectrum distribution result of short-time audio. Wherein, calculating the low frequency energy sum value T is summing the low frequency energy of the spectrum distribution of the short time audio M1 sequence.

Step S110: The low frequency energy and the value T form an N sequence in the time domain. Specifically, the calculated low frequency energy T and the value data are stored in the N sequence of the N length to obtain a long time N sequence of the audio information. Before the click characteristic phase, the N sequence is pre-determined 200, and the qualified N sequence forms an M2 sequence that enters the click characteristic phase to make the final click judgment.

Referring to FIG. 5 together, the click prediction step 200 includes:

The long-time N-sequence is filtered to remove DC processing to remove the apparently non-squeaky energy signal, wherein the N-sequence filtering DC processing is used to filter the DC component in the N-sequence data to obtain an alternating sequence.

In order to find the high frequency, the maximum and minimum calculation of the long-time N-sequence data and the maximum difference calculation, wherein the N-sequence maximum, minimum calculation, and maximum difference calculation can be performed by first obtaining the maximum value of the N-sequence. The minimum value is then calculated as the difference between the maximum and minimum values;

In addition, the squeak pre-judgment also includes the elimination of irregular audio data, high-noise noise such as car whistle, etc., and the culling pre-judgment parameters include zero-crossing rate, waveform amplitude, and zero-crossing gap.

Step S112: Predetermine the qualified N sequence to form the M2 sequence, and enter the click feature comparison phase 300.

In a preferred embodiment, in the click feature comparison stage 300, in order to make the recognition calculation more accurate to obtain hardware support, the click recognition method further includes: performing normalization processing and truncation averaging processing on the M2 sequence. Among them, the N-sequence normalization process, the truncated averaging process, normalizes the element amplitude values of the sequence N to a uniform data value size interval by the maximum value and the minimum value. And the data is subjected to averaging processing of the short sequence n (n < N / 4). The normalized and averaged sequence is the M2 sequence. The normalization process and the truncated averaging process have little effect on the recognition calculation result, and the normalization process and the truncation averaging process step may not be set. Among them, the window function of the normalization processing and the truncated averaging processing also adopts the Hanning window function (Hanning) as described above.

Step S114: Fourier transforming the M2 sequence to the frequency domain, and matching the preset chirp parameter according to the spectral peak characteristic of the M2 sequence, and outputting the monitoring chirp signal to the anti-snoring device.

For details, please refer to FIG. 7. The M1 sequence is subjected to Fast Fourier Transform (FFT), and the M1 point fast Fourier transform is performed based on the M1 sequence or the N sequence (when the pre-judging step is set), and the spectrum distribution result of the long-term audio is obtained. In the spectrum analysis process, the information of the three largest spectral peaks (P1, P2, P3) in the spectral distribution is recorded. If the three spectral peaks (P1, P2, P3) satisfy the determination condition, it is determined that the long-term audio is Beep, and output a beep signal; otherwise, it is recognized as a non-beep, and then a non-beep signal is output.

Please refer to FIG. 8 , which is a flow chart of the click characteristic comparison of the click recognition method. The final screening of the click is different according to the characteristics of the click. In this embodiment, the subsequent judgment process of the three points (P1, P2, P3) with the largest peak of the spectrum found from the M2 sequence is as follows:

Obtain three frequency points (P1, P2, P3) with the largest spectrum energy;

Calculating a multiple of the maximum spectral energy and the second spectral energy;

If the multiple is greater than 2, the squeaking characteristic is further determined, and it is further determined whether the frequency greater than the multiple 2 meets the breathing rate, and the squeak is determined when the breathing rate is met; when the breathing rate is not met, the sound is determined to be non-squeaking;

If the single frequency point multiple is less than 2, it does not meet the snoring characteristics, and it can be further determined whether the three frequency points all correspond to the breathing rate, and the squeak is determined when the breathing rate is met; when the breathing rate is not met, the sound is determined to be non-squeaking.

The above embodiment is judged by the maximum three spectral peaks, and the determination rule can be specifically determined based on single or multiple spectral peaks.

The snoring recognition method and the snoring device of the present application divide the snoring recognition into an energy accumulation phase and a snoring feature comparison phase, and the snoring recognition model is simplified, the calculation amount is greatly reduced, and the miniaturization and portable design of the stagnation product can be realized while ensuring high The detection accuracy of the click sound; the method and device for detecting the click sound accurately recognizes the click sound with high signal to noise, and the spectrum energy of the click sound is mainly concentrated in the low frequency part, and the long-term energy signal sequence is composed by acquiring the short-time low-frequency signal energy and then the long-term energy is generated. The sequence is subjected to spectrum analysis to obtain whether the long-term audio is a click sound, and the click recognition accuracy is high.

FIG. 9 is a schematic diagram of the hardware structure of the electronic device 600 according to the method for identifying the click sound provided by the embodiment of the present application. As shown in FIG. 9, the electronic device 600 includes:

One or more of the microcontroller 610 and the memory 620, a microcontroller 610 is exemplified in FIG.

The microcontroller 610 and the memory 620 can be connected by a bus or other means, as exemplified by a bus connection in FIG.

The memory 620 is used as a non-volatile computer readable storage medium, and can be used for storing a non-volatile software program, a non-volatile computer-executable program, and a module, such as a program instruction corresponding to the click recognition method in the embodiment of the present application. / Module (for example, the click recognition module 40 shown in Figure 2). The microcontroller 610 performs various function applications of the terminal device or the server and completes data processing by executing non-volatile software programs, instructions, and modules stored in the memory 620, that is, implementing the click recognition method of the above method embodiments.

The memory 620 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application required for at least one function; the storage data area may store data created according to the use of the click recognition device, and the like. Moreover, memory 620 can include high speed random access memory, and can also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some embodiments, memory 620 can optionally include memory remotely located relative to microcontroller 610, which can be connected to the click recognition device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The one or more modules are stored in the memory 620, and when executed by the one or more microcontrollers 610, perform a click recognition method in any of the above method embodiments, for example, performing the above described FIG. The method steps 101 to S114 in FIG. 5, the method steps 100 to 300 in FIG. 5, implement the functions of the modules 31-34 in FIG. 4, and implement the functions of the click recognition module 40 in FIG.

The above products can perform the methods provided by the embodiments of the present application, and have the corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in this embodiment, reference may be made to the method provided by the embodiments of the present application.

The electronic device of the embodiment of the present application exists in various forms, including but not limited to:

(1) Mobile communication devices: These devices are characterized by mobile communication functions and are mainly aimed at providing voice and data communication. Such terminals include: smart phones (such as iPhone), multimedia phones, functional phones, and low-end phones.

(2) Ultra-mobile personal computer equipment: This type of equipment belongs to the category of personal computers, has computing and processing functions, and generally has mobile Internet access. Such terminals include: PDAs, MIDs, and UMPC devices, such as the iPad.

(3) Server: A device that provides computing services. The server consists of a microcontroller, a hard disk, a memory, a system bus, etc. The server is similar to a general-purpose computer architecture, but because of the need to provide highly reliable services, it is capable of processing and stable. Sex, reliability, security, scalability, manageability, etc. are highly demanding.

(4) Other electronic devices with data interaction functions.

Embodiments of the present application provide a non-transitory computer readable storage medium storing computer-executable instructions that are executed by one or more microcontrollers, such as FIG. One of the microcontrollers 610 can cause the one or more microcontrollers to perform the click recognition method in any of the above method embodiments, for example, to perform the method steps 101 to S114 in FIG. 1 described above, FIG. In steps 100 through 300 of the method, the functions of the modules 31-34 in FIG. 4 are implemented, and the functions of the click recognition module 40 in FIG. 2 are implemented.

The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, ie may be located A place, or it can be distributed to multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the embodiment.

Through the description of the above embodiments, those skilled in the art can clearly understand that the various embodiments can be implemented by means of software plus a general hardware platform, and of course, by hardware. A person skilled in the art can understand that all or part of the process of implementing the above embodiments can be completed by a computer program to instruct related hardware, and the program can be stored in a computer readable storage medium. When executed, the flow of an embodiment of the methods as described above may be included. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (Random Access). Memory, RAM), etc.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and are not limited thereto; in the idea of the present application, the technical features in the above embodiments or different embodiments may also be combined. The steps may be carried out in any order, and there are many other variations of the various aspects of the present application as described above, which are not provided in the details for the sake of brevity; although the present application has been described in detail with reference to the foregoing embodiments, The skilled person should understand that the technical solutions described in the foregoing embodiments may be modified, or some of the technical features may be equivalently replaced; and the modifications or substitutions do not deviate from the embodiments of the present application. The scope of the technical solution.

Claims

A method for identifying a click, comprising:

Beep audio signal sampling;

Performing analog-to-digital (ADC) conversion on the sampled signal;

Forming an audio frame based on the time domain and the sampled signal, and performing a click analysis on the audio frame shift frame;

The window function is framed, and the M1 sequence obtained by the frame is Fourier transformed into the frequency domain, and the low frequency energy of the frequency domain spectrum is summed;

The low frequency energy and the value form an N sequence in the time domain, and the N sequence is predicted by the click, and the qualified N sequence is predicted to form the M2 sequence;

The M2 sequence is Fourier transformed to the frequency domain, and the preset chirp parameters are matched according to the spectral peak characteristics of the M2 sequence, and the monitoring chirp signal is output.
The method of claim 1 wherein the method further comprises:

The audio frame is moved first and last out at a 1/X rate, where X is an arbitrary number;

Wherein, the method further comprises an audio signal pre-emphasis process.
The method of claim 2 wherein said predicting the N sequence comprises:

Calculating a maximum minimum value and a maximum difference value of the N sequence data;

The parameters predicted by the click include a zero-crossing rate, a waveform amplitude, and a zero-crossing gap.
The method of any of claims 1-3, wherein the method further comprises:

The M2 sequence is subjected to normalization processing and truncated averaging processing.
The method of claim 4, wherein

The window function is a Hanning window function (Haning) or a Hamming window function (Hamming) or a Blackman window function (Blackman).
A stagnation device comprising a microcontroller, an analog to digital conversion circuit (ADC) and a sound pickup circuit, the sound pickup circuit for sampling a click audio signal, the analog to digital conversion circuit for modulating the sample signal Number (ADC) conversion, also includes:

a click recognition module, wherein the click recognition module is configured to form an audio frame based on the time domain and the sampled signal, perform a click analysis on the audio frame shift frame; and use the window function to divide the frame, and transform the M1 sequence obtained by the frame to the frequency. Domain, and summing the low-frequency energy of the frequency domain spectrum; for the low-frequency energy and value to form the N-sequence in the time domain, and pre-judging the frequency-domain N-sequence, pre-determining the qualified N-sequence to form the M2 sequence; Performing a Fourier transform on the M2 sequence to the frequency domain, and matching the preset chirp parameters according to the spectral peak characteristics of the M2 sequence, and outputting the monitoring chirp signal;

The stopping mechanism, wherein the microcontroller initiates the stop mechanism to touch the monitored user after receiving the monitoring click signal.
The anti-snoring device according to claim 6, wherein the click recognition module is further configured to move the frame at a 1/X rate on the audio frame (where X is an arbitrary number); the click recognition module further uses Pre-emphasis processing of the audio signal.
The anti-snoring device according to claim 7, wherein the click recognition module is further configured to:

Calculating a maximum minimum value and a maximum difference value of the N sequence data;

The parameters predicted by the click include a zero-crossing rate, a waveform amplitude, and a zero-crossing gap.
The anti-snoring device according to any one of claims 6-8, wherein the click recognition module is further configured to perform normalization processing and truncation averaging processing on the M2 sequence.
The anti-snoring device according to claim 9, wherein

The window function is a Hanning window function (Haning) or a Hamming window function (Hamming) or a Blackman window function (Blackman).
An electronic device, comprising:

At least one microcontroller; and,

a memory communicatively coupled to the at least one microcontroller; wherein

The memory stores instructions executable by the at least one microcontroller, the instructions being executed by the at least one microcontroller to enable the at least one microcontroller to perform any of claims 1-5 Said method.
A non-transitory computer readable storage medium, wherein the computer readable storage medium stores computer executable instructions for causing a computer to perform the method of any of claims 1-5 method.
A computer program product, comprising: a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions, when the program instructions are executed by a computer, The computer performs the method of any of claims 1-5.