阅读说明：本技术 一种通过声音采集检测睡眠质量的方法 (Method for detecting sleep quality through sound collection ) 是由 陶伟胜 阳纯正 于 2019-09-20 设计创作，主要内容包括：本发明实施例公开了一种通过声音采集检测睡眠质量的方法,包括以下步骤,采集睡眠状态下的声音,生成关于睡眠状态下的声音信号时域图；声音信号预处理,剪除声音信号时域图中的首末误差区信号；根据呼吸信号的频率和幅值特性,从声音信号时域图内分离呼吸信号和干扰信号；根据呼吸信号的时域图计算生成呼吸信号的频谱图,将干扰信号细化为环境噪音干扰信号和睡眠做梦干扰信号；统计呼吸信号的波形变化,总结睡眠质量；根据睡眠总时长,以及深睡阶段的占比推断睡眠质量,实现方式简单,操作方便,集成在手机中即可使用,用户能随时检测自己的睡眠情况。(The embodiment of the invention discloses a method for detecting sleep quality through sound collection, which comprises the following steps of collecting sound in a sleep state and generating a sound signal time domain diagram in the sleep state; preprocessing a sound signal, and cutting out first and last error region signals in a time domain diagram of the sound signal; according to the frequency and amplitude characteristics of the respiratory signal, separating the respiratory signal and the interference signal from a sound signal time domain diagram; calculating and generating a spectrogram of the respiratory signal according to a time domain diagram of the respiratory signal, and refining the interference signal into an environmental noise interference signal and a sleep dream interference signal; counting the waveform change of the respiratory signal and summarizing the sleep quality; the sleep quality is inferred according to the total sleep duration and the proportion of the deep sleep stage, the implementation mode is simple, the operation is convenient, the sleep quality can be integrated in a mobile phone for use, and a user can detect the sleep condition of the user at any time.)

1. A method for detecting sleep quality through sound collection is characterized by comprising the following steps:

step 100, collecting sound in a sleep state, and generating a sound signal time domain diagram in the sleep state;

step 200, preprocessing a sound signal, and cutting out a first error region signal and a last error region signal in a sound signal time domain diagram;

step 300, separating a respiratory signal and an interference signal from a sound signal time domain diagram according to the frequency and amplitude characteristics of the respiratory signal;

step 400, calculating a spectrogram of a respiratory signal according to a time domain graph of the respiratory signal, separately calculating the amplitude of the interference signal, analyzing the emotion of the interference signal by machine language learning, and refining the interference signal into an environmental noise interference signal and a sleep dream interference signal;

step 500, counting the waveform change of the respiration signal, and summarizing the sleep quality.

2. The method for detecting sleep quality through sound collection according to claim 1, wherein in step 200, the sound signal is preprocessed, and the specific steps of removing the collected error signal are as follows:

step 201, independently collecting a sound signal time domain diagram per hour to form a plurality of independent units related to sound signals;

step 202, determining a frequency range and an amplitude range which accord with the human respiratory signal according to the time domain and frequency domain characteristics of the human normal respiratory signal;

step 203, respectively acquiring respiratory signals in the independent unit of the front sleep stage and the independent unit of the rear sleep stage according to a time sequence;

and 204, marking the time point of the first appearance or disappearance of the breathing signal, cutting off the time period before the first appearance of the breathing signal and after the first disappearance of the signal, and taking the residual time of the independent unit of the sound signal as a time domain graph of the sound signal in the sleep state.

3. The method for detecting sleep quality through sound collection according to claim 2, wherein: in step 204, the time period between the first occurrences of the clipped respiratory signal is specifically the time to sleep, which is one of the criteria for measuring the quality of sleep.

4. The method for detecting sleep quality through sound collection according to claim 1, wherein in step 300, the breathing signal and the interference signal are separated from each sound signal independent unit according to the frequency and amplitude characteristics of the breathing signal, and the method comprises the following specific steps:

step 301, determining a respiratory waveform meeting the standards of the frequency range and the amplitude range of the respiratory signal from each sound signal independent unit corresponding to the frequency range and the amplitude range of the human respiratory signal;

step 302, periodically extending the respiratory waveforms to correspondingly form the respiratory waveforms of the independent units of each sound signal;

step 303, filtering the sound signal independent units according to the respiratory waveform adaptability, and separating a respiratory waveform from each sound signal independent unit;

and step 304, taking the sound signal independent unit with the separated respiratory waveform as an interference signal of mixed noise.

5. The method for detecting sleep quality through sound collection according to claim 1, wherein in step 400, the steps of processing the respiration waveform of the sound signal independent unit and the noise-mixed interference signal are as follows:

step 401, determining the breathing frequency of each sound signal independent unit according to the breathing waveform which extends periodically, and deducing and determining time periods corresponding to different sleep stages according to the difference of the breathing frequency and the noise in the corresponding breathing frequency time;

step 402, recognizing interference signals corresponding to different sleep stage time periods by voice, and refining the interference signals into environment noise interference signals and sleep dream interference signals;

step 403, analyzing the frequency and amplitude of the ambient noise interference signal, and determining the relationship between the ambient noise interference signal and the sleep stage;

step 404, analyzing the emotion of the sleep dream disturbance signal by using machine language learning, and determining the relationship between the sleep dream disturbance signal and the sleep stage.

6. The method for detecting sleep quality through sound collection according to claim 5, wherein: the breathing state is calculated in a sound collection and detection mode, and the sleep stage is divided into a sleep stage, a light sleep stage and a deep sleep stage according to the breathing frequency.

7. The method of claim 6, wherein the method comprises the following steps: the sleep quality is determined by the proportion of the deep sleep stage in the total sleep time.

8. The method for detecting sleep quality through sound collection according to claim 1, wherein in step 500, the waveform variation of the respiration signal is counted, and the specific steps for summarizing sleep quality are as follows:

step 501, counting respiratory waveforms of each sound signal independent unit, and generating a spectrogram of a respiratory signal corresponding to each respiratory waveform;

step 502, distinguishing the time length occupied by the respiratory signals with different frequencies, and dividing the respiratory signals into a sleep-in stage, a light sleep stage and a deep sleep stage according to different frequencies;

step 503, counting the proportion of the deep sleep stage in the total sleep time, corresponding to the sleep quality of different standards.

Technical Field

The embodiment of the invention relates to the technical field of sleep quality, in particular to a method for detecting sleep quality through sound collection.

Background

Modern society's life rhythm is fast, and is big, and the crowd of each age bracket all receives the puzzlement of sleep problem to a different extent, and sleep occupies one thirty percent of the life time, and various sleep problems such as sleep disorder and sleep disease can influence people's work and life to a different extent, but also can induce other many diseases. People are difficult to find abnormal problems in the sleeping process, and the sleeping problems appear in variable time, so the problems are found by continuous monitoring, and the earlier the problems are found, the related diseases can be better prevented.

However, currently, monitoring sleep requires huge economic cost and professional technical personnel support, and the monitoring method is complex, has high professional requirements on the monitoring personnel, and is high in monitoring cost. In the prior art, algorithms for evaluating sleep quality by utilizing body motion signals mainly comprise two categories of linearity and nonlinearity, the linear algorithms are simple to realize, and a large number of predecessor research bases exist; the nonlinear algorithm is very complex to implement, requires a device with strong computing power, and currently, no explicit nonlinear model exists, and most importantly, the accuracy of sleep quality assessment based on single body data is low.

Disclosure of Invention

Therefore, the embodiment of the invention provides a method for detecting sleep quality through sound collection, so as to solve the problem that in the prior art, accurate sleep quality evaluation is difficult for single body data.

In order to achieve the above object, an embodiment of the present invention provides the following: a method for detecting sleep quality through sound collection is characterized by comprising the following steps:

step 100, collecting sound in a sleep state, and generating a sound signal time domain diagram in the sleep state;

step 200, preprocessing a sound signal, and cutting out a first error region signal and a last error region signal in a sound signal time domain diagram;

step 300, separating a respiratory signal and an interference signal from a sound signal time domain diagram according to the frequency and amplitude characteristics of the respiratory signal;

step 400, calculating a spectrogram of a respiratory signal according to a time domain graph of the respiratory signal, separately calculating the amplitude of the interference signal, analyzing the emotion of the interference signal by machine language learning, and refining the interference signal into an environmental noise interference signal and a sleep dream interference signal;

step 500, counting the waveform change of the respiration signal, and summarizing the sleep quality.

As a preferred embodiment of the present invention, in step 200, the sound signal is preprocessed to remove the collected error signal, and the specific steps are as follows:

step 201, independently collecting a sound signal time domain diagram per hour to form a plurality of independent units related to sound signals;

step 202, determining a frequency range and an amplitude range which accord with the human respiratory signal according to the time domain and frequency domain characteristics of the human normal respiratory signal;

step 203, respectively acquiring respiratory signals in the independent unit of the front sleep stage and the independent unit of the rear sleep stage according to a time sequence;

and 204, marking the time point of the first appearance or disappearance of the breathing signal, cutting off the time period before the first appearance of the breathing signal and after the first disappearance of the signal, and taking the residual time of the independent unit of the sound signal as a time domain graph of the sound signal in the sleep state.

As a preferred aspect of the present invention, in step 204, the time period between the first occurrences of the clipped respiratory signal is specifically the time to sleep, which is one of the criteria for measuring the quality of sleep.

As a preferred aspect of the present invention, in step 300, according to the frequency and amplitude characteristics of the respiratory signal, the respiratory signal and the interference signal are separated from each sound signal independent unit, and the specific steps are as follows:

step 301, determining a respiratory waveform meeting the standards of the frequency range and the amplitude range of the respiratory signal from each sound signal independent unit corresponding to the frequency range and the amplitude range of the human respiratory signal;

step 302, periodically extending the respiratory waveforms to correspondingly form the respiratory waveforms of the independent units of each sound signal;

step 303, filtering the sound signal independent units according to the respiratory waveform adaptability, and separating a respiratory waveform from each sound signal independent unit;

and step 304, taking the sound signal independent unit with the separated respiratory waveform as an interference signal of mixed noise.

As a preferred aspect of the present invention, in step 400, the specific steps of processing the respiration waveform of the sound signal independent unit and the noise-mixed interference signal are as follows:

step 401, determining the breathing frequency of each sound signal independent unit according to the breathing waveform which extends periodically, and deducing and determining time periods corresponding to different sleep stages according to the difference of the breathing frequency and the noise in the corresponding breathing frequency time;

step 402, recognizing interference signals corresponding to different sleep stage time periods by voice, and refining the interference signals into environment noise interference signals and sleep dream interference signals;

step 403, analyzing the frequency and amplitude of the ambient noise interference signal, and determining the relationship between the ambient noise interference signal and the sleep stage;

step 404, analyzing the emotion of the sleep dream disturbance signal by using machine language learning, and determining the relationship between the sleep dream disturbance signal and the sleep stage.

As a preferable scheme of the invention, the breathing state is calculated by a sound collection detection mode, and the sleep stage is divided into a sleep stage, a light sleep stage and a deep sleep stage according to the breathing frequency.

As a preferred aspect of the present invention, the sleep quality is determined by the ratio of the deep sleep phase to the total sleep time period.

As a preferred embodiment of the present invention, in step 500, the waveform variation of the respiration signal is counted, and the specific steps of summarizing the sleep quality are as follows:

step 501, counting respiratory waveforms of each sound signal independent unit, and generating a spectrogram of a respiratory signal corresponding to each respiratory waveform;

step 502, distinguishing the time length occupied by the respiratory signals with different frequencies, and dividing the respiratory signals into a sleep-in stage, a light sleep stage and a deep sleep stage according to different frequencies;

step 503, counting the proportion of the deep sleep stage in the total sleep time, corresponding to the sleep quality of different standards.

The embodiment of the invention has the following advantages:

(1) the sleep quality is inferred according to the total sleep duration and the proportion of the deep sleep stage by collecting the sound during sleep, counting the respiratory frequency, identifying the sleep-in stage, the light sleep stage and the deep sleep stage in sleep, the realization mode is simple, the operation is convenient, the sleep quality can be used by being integrated in a mobile phone, and a user can detect the sleep condition of the user at any time;

(2) the method can analyze the relationship between the sleep quality and the external environment noise and the dream condition according to the noise corresponding to the sleep stage, conveniently adjust the factors influencing the sleep and improve the sleep quality.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It should be apparent that the drawings in the following description are merely exemplary, and that other embodiments can be derived from the drawings provided by those of ordinary skill in the art without inventive effort.

Fig. 1 is a schematic flow chart of a sleep quality detection method according to an embodiment of the present invention;

Detailed Description

The present invention is described in terms of particular embodiments, other advantages and features of the invention will become apparent to those skilled in the art from the following disclosure, and it is to be understood that the described embodiments are merely exemplary of the invention and that it is not intended to limit the invention to the particular embodiments disclosed. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

As shown in fig. 1, the present invention provides a method for detecting sleep quality through sound collection, comprising the steps of:

step 100, collecting the sound in the sleep state, and generating a sound signal time domain diagram in the sleep state.

Step 200, sound signal preprocessing, and cutting out the first and last error region signals in the sound signal time domain diagram.

In step 200, the specific steps of preprocessing the sound signal and removing the collected error signal are as follows:

step 201, independently collecting a sound signal time domain diagram per hour to form a plurality of independent units related to sound signals;

step 202, determining a frequency range and an amplitude range which accord with the human respiratory signal according to the time domain and frequency domain characteristics of the human normal respiratory signal;

step 203, respectively acquiring respiratory signals in the independent unit of the front sleep stage and the independent unit of the rear sleep stage according to a time sequence;

and 204, marking the time point of the first appearance or disappearance of the breathing signal, cutting off the time period before the first appearance of the breathing signal and after the first disappearance of the signal, and taking the residual time of the independent unit of the sound signal as a time domain graph of the sound signal in the sleep state.

The time period between the first occurrence of the clipped respiratory signal is in particular the time to sleep, which is one of the criteria for measuring the quality of sleep.

And step 300, separating the respiratory signal and the interference signal from the sound signal time domain diagram according to the frequency and amplitude characteristics of the respiratory signal.

In step 300, according to the frequency and amplitude characteristics of the respiratory signal, the respiratory signal and the interference signal are separated from each sound signal independent unit, and the specific steps are as follows:

step 301, determining a respiratory waveform meeting the standards of the frequency range and the amplitude range of the respiratory signal from each sound signal independent unit corresponding to the frequency range and the amplitude range of the human respiratory signal;

step 302, periodically extending the respiratory waveforms to correspondingly form the respiratory waveforms of the independent units of each sound signal;

step 303, filtering the sound signal independent units according to the respiratory waveform adaptability, and separating a respiratory waveform from each sound signal independent unit;

and step 304, taking the sound signal independent unit with the separated respiratory waveform as an interference signal of mixed noise.

Step 400, calculating a spectrogram of the respiratory signal according to the time domain diagram of the respiratory signal, separately calculating the amplitude of the interference signal, analyzing the emotion of the interference signal by machine language learning, and refining the interference signal into an environmental noise interference signal and a sleep dream interference signal.

The specific steps of processing the respiratory waveform of the sound signal independent unit and the interference signal of the mixed noise are as follows:

step 401, determining the breathing frequency of each sound signal independent unit according to the breathing waveform which extends periodically, and deducing and determining time periods corresponding to different sleep stages according to the difference of the breathing frequency and the noise in the corresponding breathing frequency time;

step 402, recognizing interference signals corresponding to different sleep stage time periods by voice, and refining the interference signals into environment noise interference signals and sleep dream interference signals;

step 403, analyzing the frequency and amplitude of the ambient noise interference signal, and determining the relationship between the ambient noise interference signal and the sleep stage;

step 404, analyzing the emotion of the sleep dream disturbance signal by using machine language learning, and determining the relationship between the sleep dream disturbance signal and the sleep stage.

The breathing state is calculated in a sound collection and detection mode, and the sleep stage is divided into a sleep stage, a light sleep stage and a deep sleep stage according to the breathing frequency.

The sleep quality is determined by the proportion of the deep sleep stage in the total sleep time.

Step 500, counting the waveform change of the respiration signal, and summarizing the sleep quality.

Step 501, counting respiratory waveforms of each sound signal independent unit, and generating a spectrogram of a respiratory signal corresponding to each respiratory waveform;

step 502, distinguishing the time length occupied by the respiratory signals with different frequencies, and dividing the respiratory signals into a sleep-in stage, a light sleep stage and a deep sleep stage according to different frequencies;

step 503, counting the proportion of the deep sleep stage in the total sleep time, corresponding to the sleep quality of different standards.

Although the invention has been described in detail above with reference to a general description and specific examples, it will be apparent to one skilled in the art that modifications or improvements may be made thereto based on the invention. Accordingly, such modifications and improvements are intended to be within the scope of the invention as claimed.