Non-contact life signal processing method in strong clutter environment

Abstract

The invention provides a non-contact life signal processing method in a strong clutter environment, and belongs to the field of data processing.The method comprises the steps that extreme points in a signal data sequence of a user and first and last signal data are recorded as to-be-corrected data; obtaining an amplitude adjustment parameter and a slope adjustment parameter of each piece of to-be-corrected data according to a numerical value difference and an acquisition time difference between each piece of to-be-corrected data and the to-be-corrected data on the left side and the right side of the to-be-corrected data; and combining the numerical value of the to-be-corrected data to obtain a corrected value of the to-be-corrected data, thereby obtaining upper and lower envelope lines. The invention aims to solve the problem that the signal data of a user is processed by an EMD (empirical mode decomposition) algorithm due to the fact that an upper envelope line and a lower envelope line which are obtained through extreme points in a signal data sequence are greatly influenced by noise because the noise in an environment where the user is located has a great influence on the signal data of the user. And the accuracy of the obtained heart rate signal and respiration signal is low.

Description

Technical Field

[0001] This invention belongs to the field of data processing, specifically relating to a non-contact life signal processing method under strong clutter environment. Background Technology

[0002] This invention relates to the interdisciplinary field of radar signal processing and biomedical engineering, and particularly to a non-contact vital sign monitoring method based on frequency-modulated continuous wave (FMCW) radar or ultra-wideband (UWB) radar. This method is suitable for high-precision extraction of weak respiratory and heartbeat signals in environments with strong clutter and low signal-to-noise ratio. Non-contact vital sign monitoring technology, especially the extraction of respiratory and heartbeat frequencies by sensing chest cavity micro-movements through radar, has broad application prospects in medical monitoring, disaster search and rescue, and sleep monitoring. However, in practical applications, vital sign signals such as respiratory and heartbeat are often interfered with by strong environmental noise, especially in environments with strong clutter and low signal-to-noise ratio, where these signals are easily submerged by noise. Traditional processing methods, such as Fourier transform, separate regular mixed periodic signals. However, vital signs affected by strong clutter are often not so regularly periodic, so Fourier transform cannot effectively address these problems, resulting in incomplete signal decomposition and difficulty in clearly distinguishing vital signs. This causes traditional methods, such as simple bandpass filtering and Fourier transform, to experience a sharp decline in performance under strong clutter environments. This makes it difficult for current technologies to accurately separate vital signs from signals collected by mattress sensors or other sensors such as PPG photoplethysmography (PPG). Therefore, there is an urgent need for a processing method that can effectively suppress strong clutter and accurately demodulate weak vital signs.

[0003] Since the Empirical Mode Decomposition (EMD) algorithm is suitable for decomposing complex biological signals into different intrinsic mode functions, this invention uses the EMD algorithm to extract the required weak vital signals from the acquired raw signals. However, in noisy environments, the EMD algorithm is easily affected by abnormal extreme points when constructing the upper and lower envelopes of the signal, leading to overshoot or undershoot in the envelope, which in turn affects the accuracy and precision of the signal decomposition, causing incorrect combination of signal components and thus affecting the final detection effect of vital signals. In other words, when the current EMD algorithm decomposes the acquired raw signals, the resulting respiratory and heart rate signals differ significantly from the actual signals. Summary of the Invention

[0004] To address the problem that when extracting the required signal from the raw user signal obtained in a non-contact manner, the original signal may contain noise, causing the upper and lower envelopes obtained from the extreme points of the original signal to be greatly affected by noise. This results in a significant difference between the extracted signal and the actual required signal when using the EMD algorithm to extract the required signal from the original signal. Therefore, this invention proposes a non-contact life signal processing method in a strong clutter environment.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] Acquire the user's signal data sequence;

[0007] The extreme points in the user's signal data sequence, as well as the first and last signal data, are all recorded as data to be corrected. The amplitude difference between each data to be corrected and the data to be corrected on its left and right sides is obtained based on the numerical difference between each data to be corrected and the surrounding data to be corrected. The amplitude adjustment parameter of each data to be corrected is obtained based on the amplitude difference between each data to be corrected and the surrounding data to be corrected, and the average of the amplitude differences between all data to be corrected and the surrounding data to be corrected.

[0008] Based on the numerical differences between each data point to be corrected and the data points to be corrected on its left and right sides, as well as the differences in collection time, the slope of each data point to be corrected is obtained; based on the difference between the slope of each data point to be corrected and the mean of the slopes of all data points to be corrected, the slope adjustment parameter of each data point to be corrected is obtained.

[0009] Based on whether the amplitude difference between each data point to be corrected and its surrounding data points is greater than the average amplitude difference between all data points to be corrected and their surrounding data points, it is determined whether the value of each data point to be corrected should be corrected. If no correction is made, the value of the data point to be corrected is its corrected value. If correction is made, the corrected value of the data point to be corrected is obtained based on the average amplitude difference between all data points to be corrected and their surrounding data points, the value of the data point to be corrected, the amplitude adjustment parameter, and the slope adjustment parameter. Then, the upper and lower envelopes and the corrected signal data sequence are obtained. Through the EMD algorithm, the user's respiratory signal and heart rate signal are obtained.

[0010] Furthermore, the specific calculation formula for obtaining the amplitude difference between each data point to be corrected and its surrounding data points to be corrected is as follows:

[0011]

[0012] In the formula, Indicates the first The magnitude difference between the data to be corrected and the surrounding data to be corrected Indicates the first The value of the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The value of the data to be corrected. This represents the absolute value function.

[0013] Furthermore, the specific calculation formula for obtaining the amplitude adjustment parameter for each piece of data to be corrected is as follows:

[0014]

[0015] In the formula, Indicates the first The magnitude adjustment parameter for the data to be corrected. Indicates the first The magnitude difference between the data to be corrected and the surrounding data to be corrected This represents the mean of the magnitude differences between all data to be corrected and their surrounding data. This represents the maximum value among all the magnitude differences between the data to be corrected and its surrounding data. The brackets represent Iverson.

[0016] Furthermore, the specific formula for calculating the slope of each piece of data to be corrected is as follows:

[0017]

[0018] In the formula, Indicates the first The slope of the data to be corrected Indicates the first The value of the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The collection time of the data to be corrected. Indicates the first The collection time of the data to be corrected. Indicates the first The collection time of the data to be corrected. This represents the absolute value function.

[0019] Furthermore, the specific calculation formula for obtaining the slope adjustment parameter for each piece of data to be corrected is as follows:

[0020]

[0021] In the formula, Indicates the first The slope adjustment parameter for the data to be corrected. Indicates the first The slope of the data to be corrected This represents the mean of the slopes of all the data to be corrected. It represents 180 degrees. This represents the arctangent function.

[0022] Furthermore, if no correction is made, the specific calculation formula for the value of the data to be corrected, which is its correction value, is as follows:

[0023]

[0024] In the formula, Indicates the first Correction values ​​for the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The magnitude difference between the data to be corrected and the surrounding data to be corrected This represents the mean of the magnitude differences between all data to be corrected and the surrounding data to be corrected.

[0025] Furthermore, the specific calculation formula for obtaining the correction value of the data to be corrected is as follows:

[0026] when hour,

[0027]

[0028] In the formula, Indicates the first Correction values ​​for the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The slope adjustment parameter for the data to be corrected. Indicates the first The magnitude difference between the data to be corrected and the surrounding data to be corrected This represents the mean of the magnitude differences between all data to be corrected and their surrounding data. Indicates the first The magnitude adjustment parameter for the data to be corrected.

[0029] Furthermore, the specific steps for obtaining the upper and lower envelopes are as follows:

[0030] If the first If the value of the data to be corrected is a maximum point before correction, then when obtaining the upper and lower envelopes, the value of the first data point is... Each data point to be corrected will also be considered a maximum point after correction;

[0031] If the If the value of the data to be corrected is a local minimum point before correction, then when obtaining the upper and lower envelopes, the value of the first data point is... Each data point to be corrected is also considered a local minimum point after correction.

[0032] Before correcting all the data to be corrected, if the second to last data to be corrected is a maximum point, then when obtaining the upper and lower envelopes, the last data to be corrected will be a minimum point after correction. If the second to last data to be corrected is a minimum point, then when obtaining the upper and lower envelopes, the last data to be corrected will be a maximum point after correction.

[0033] Before correcting all the data to be corrected, if the second data to be corrected is a maximum point, then when obtaining the upper and lower envelopes, the first data to be corrected will be a minimum point after correction; before correcting all the data to be corrected, if the second data to be corrected is a minimum point, then when obtaining the upper and lower envelopes, the first data to be corrected will be a maximum point after correction.

[0034] In the user's signal data sequence, the value of each data to be corrected is modified to its corrected value; based on the data to be corrected as the maximum point when the upper and lower envelopes are obtained, the upper envelope in the EMD algorithm is obtained;

[0035] The lower envelope in the EMD algorithm is obtained by using the data to be corrected as the minimum point when obtaining the upper and lower envelopes.

[0036] The non-contact vital signal processing method provided by this invention in a strong clutter environment has the following beneficial effects: When processing user signal data using the EMD algorithm to obtain user respiratory and heart rate signal data, the influence of respiration and heart rate on the numerical values ​​of the signal data is relatively small, while noise has a significant impact. Specifically, some extreme points in the user signal data are greatly affected by noise, resulting in significant noise influence on the acquired upper and lower envelopes. This leads to signal off-grid errors when processing user signal data in a strong noise environment using the EMD algorithm, causing inaccurate signal decomposition and potentially significant differences between the acquired respiratory and heart rate signals and the actual required signals. When correcting extreme points in the user signal data, this invention first considers the irregular occurrence of noise and its significant impact on the numerical values ​​of the user signal data, resulting in significant differences between noise-affected extreme points and their surrounding extreme points. Based on the difference between each extreme point and its surrounding extreme points, an amplitude adjustment parameter is obtained for each extreme point. Then, the amplitude adjustment parameter is adjusted according to the difference between each extreme point and its surrounding extreme points. The parameters are adjusted for each extreme point. This involves adaptively adjusting the amplitude and slope of the signal's extreme points, effectively eliminating noise-induced extreme points, reducing the impact of noise on their values, and mitigating its influence on the envelope fitting process. This allows for more accurate extraction of vital signs from the original signal. Since the above steps did not consider the difference in acquisition time between each extreme point and its surrounding extreme points when calculating the amplitude adjustment parameter, the amplitude adjustment parameter for a normal extreme point might be large if the acquisition interval between it and its surrounding normal extreme points is long. Therefore, based on the numerical differences between each extreme point and its surrounding extreme points, as well as the differences in acquisition time, a slope adjustment parameter is obtained for each extreme point. Based on the slope and amplitude adjustment parameters, the value of each extreme point is corrected, resulting in an upper and lower envelope less affected by noise. This reduces the impact of noise on the processing process and results when processing user signals using the EMD algorithm, thus effectively extracting weak vital signs, especially making the separation of respiration and heartbeat clearer, even in environments with strong clutter and low signal-to-noise ratio. Furthermore, compared with existing technologies, this method can extract target signals more accurately in noisy environments, significantly improving the accuracy and reliability of vital sign monitoring.

[0037] Furthermore, traditional methods often ignore endpoints or handle them inappropriately when processing signal data. This method, however, treats endpoints as extreme points and adjusts the processing based on the polarity of adjacent extreme points, thus ensuring the continuity and integrity of signal processing. This innovative approach ensures that even special fluctuations at signal endpoints are handled appropriately, avoiding the impact of traditional methods ignoring endpoints. Attached Figure Description

[0038] To more clearly illustrate the embodiments and design schemes of the present invention, the accompanying drawings required for this embodiment will be briefly described below. The drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is a flowchart of a non-contact life signal processing method under strong clutter environment according to an embodiment of the present invention. Detailed Implementation

[0040] To enable those skilled in the art to better understand and implement the technical solutions of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be construed as limiting the scope of protection of the present invention.

[0041] Example 1:

[0042] This invention provides a non-contact life signal processing method in a strong clutter environment, specifically as follows: Figure 1 As shown, it includes:

[0043] Step S001: Obtain the user's signal data sequence.

[0044] Specifically, when a user uses the smart mattress, sensors within the mattress, equipped with heart rate and respiration monitoring functions, collect the user's signal data in a non-contact manner, obtaining a sequence of the user's signal data. In this embodiment, when collecting the user's heart rate and respiration data, a sensor first obtains a sequence of the user's signal data. This sequence is then processed to separate the respiration and heart rate signals. That is, a single signal from the user may be influenced by both respiration and heartbeat. Furthermore, in this embodiment, there is only one user. Since this embodiment aims to acquire signal data sequences generated by heartbeat vibrations and respiratory vibrations, the sensor, when detecting a vibration, records the vibration's acquisition time and collects the vibration value, thus obtaining a single signal data point. In other words, each signal data point corresponds to a numerical value and a specific acquisition time.

[0045] At this point, the user's signal data sequence is obtained.

[0046] Step S002: Record the extreme points in the user's signal data sequence, as well as the first and last signal data, as data to be corrected; based on the numerical difference between each data to be corrected and the data to be corrected on its left and right sides, obtain the amplitude difference between each data to be corrected and the surrounding data to be corrected; based on the amplitude difference between each data to be corrected and the surrounding data to be corrected, and the average of the amplitude differences between all data to be corrected and the surrounding data to be corrected, obtain the amplitude adjustment parameter for each data to be corrected.

[0047] It should be noted that in this embodiment, the sensors in the smart mattress are used to acquire the user's heartbeat and breathing signals. However, the acquired signals may not be pure vital signals; they are often mixed signals containing various other signals, primarily formed by the superposition of vibrations from the heartbeat and breathing. Therefore, the acquired signals are processed to obtain the user's heartbeat and breathing signals.

[0048] It's important to further clarify that the vital signs signals extracted from the user are non-stationary and non-linear biological signals, which traditional Fourier transforms cannot process. The core idea of ​​the EMD algorithm is to adaptively decompose a complex signal into a series of intrinsic mode functions (EMFs), and these EMFs are likely to correspond to different physical processes in the original signal. Therefore, the EMD algorithm can be used to process the acquired signal data sequence to obtain the user's heartbeat and respiratory signals.

[0049] It's important to further clarify that, since we need to extract respiratory and heartbeat signals, the acquired raw signals are not two clear and separate perfect waveforms, but rather a mixed and complex signal affected by noise. Analysis of the characteristics of respiration and heartbeat clearly reveals that a complete breathing process involves two actions: inhalation and exhalation, with each breath taking longer than each heartbeat. That is, the signal generated by respiration should be a slow, low-frequency signal. The heartbeat, on the other hand, is a relatively short and rapid beating process, resulting in a high-frequency, fluctuating signal. When the two signals are superimposed, the overall signal should be a slowly undulating wave with rapid and dense sawtooth-like ripples superimposed on top. This is the respiratory-heartbeat coupled signal without noise interference.

[0050] It should be further noted that when collecting user signals, the acquired signal data is affected by several types of noise signals, including noise caused by the body and environmental interference. Furthermore, compared to the external environment in which the human body exists, the signals generated by fluctuations in breathing and heartbeat are quite weak, resulting in very small amplitude signals that need to be extracted, typically at the millivolt level or the raw ADC count. This makes the acquired signals highly susceptible to ambient noise interference. Even slight body movements can cause significant noise interference to the coupled signal; for example, a slight hand movement can produce large, irregular spikes with amplitudes far greater than heartbeat and breathing signals, sometimes completely drowning out the useful signal. In addition, changes in ambient light can also introduce noise, further interfering with the original coupled signal. Therefore, directly processing the acquired signals using the EMD algorithm to obtain the user's breathing and heart rate signals may yield results that differ from the actual values. Thus, certain improvements to the EMD algorithm are necessary.

[0051] It's important to further clarify that when using the EMD algorithm to extract the desired respiratory and heartbeat signals from complex signals, the first step is to generate upper and lower envelopes using the extreme points in the acquired signal, the first signal data, and the last signal data in the signal sequence. Then, a series of operations based on existing techniques are performed using these envelopes to obtain the respiratory and heartbeat signals. However, the presence of noise introduces intermittent components, impulses, or abrupt changes into the signal, causing sudden shifts in the distribution of extreme points. This makes the envelope generation process highly sensitive to these abrupt changes, leading to inaccurate envelope fitting. Consequently, during signal separation, signals of different frequencies are easily mixed together, resulting in the "off-network problem." In other words, the respiratory and heartbeat signals extracted from the acquired signal using the existing EMD algorithm may differ from the actual signals. Therefore, improvements are needed to the upper and lower envelope generation process to reduce the impact of the off-network problem. This involves correcting the values ​​of the extreme points, the first signal data, and the last signal data in the acquired signal data sequence to reduce the influence of noise.

[0052] It should be further explained that, because human respiratory and heartbeat signals are regular and their fluctuations are relatively weak, while noise is generated in a more disordered manner, the data to be corrected affected by noise will differ significantly in value from its surrounding data that is not affected by noise. Therefore, based on the difference between each data point to be corrected and its surrounding data in the collected signals, the numerical difference between each data point to be corrected and its surrounding data is obtained, that is, the amplitude difference between each data point to be corrected and its surrounding data. The value of each data point to be corrected is then adjusted accordingly.

[0053] It's important to further explain that because the signal values ​​affected by noise are much larger than the required respiratory and heart rate signals, the average amplitude difference between all data to be corrected and their surrounding data is amplified by the noise. Therefore, if the amplitude difference between a particular data point and its surrounding data is greater than the amplitude difference between all data points and their surrounding data, that data point is highly likely to be affected by noise, and thus its value needs adjustment. Furthermore, the greater the amplitude difference between the data point and its surrounding data, the greater the correction magnitude. Therefore, based on the amplitude difference between each data point and its surrounding data, an amplitude adjustment parameter is obtained for each data point to be corrected.

[0054] Specifically, extreme points are obtained from the user's signal data sequence. Obtaining extreme points from a data sequence is a well-known technique and will not be elaborated upon in this embodiment. Furthermore, the extreme points are the maximum and minimum points.

[0055] Furthermore, the extreme points, the first signal data, and the last signal data in the user's signal data sequence are all recorded as data to be corrected. Among them, the first data to be corrected refers to the data with the shortest acquisition time in the user's signal data sequence; the second data to be corrected refers to the data with the second shortest acquisition time in the user's signal data sequence, and so on, to obtain the order value of each data to be corrected.

[0056] Furthermore, obtain the first The specific formula for calculating the amplitude difference between the data to be corrected and the surrounding data to be corrected is as follows:

[0057]

[0058] In the formula, Indicates the first The magnitude difference between the data to be corrected and the surrounding data to be corrected Indicates the first The value of the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The value of the data to be corrected. This represents the absolute value function. Where, when When the value is the first data to be corrected, The value of the second data point to be corrected in the user's signal data sequence; when When it is the value of the last piece of data to be corrected, This is the value of the second-to-last data to be corrected.

[0059] It should be noted that, The larger the value, the more significant the [value]. The greater the difference between the value of the first data point to be corrected and the value of the previous data point to be corrected, the more it indicates that the first... The greater the likelihood that a piece of data to be corrected is affected by noise; The larger the value, the more significant the [value]. The significant difference between the value of the first data point to be corrected and the value of the next data point to be corrected further illustrates that the first... The greater the likelihood that a piece of data to be corrected is affected by noise.

[0060] Furthermore, obtain the first The specific calculation formula for the magnitude adjustment parameter of each data point to be corrected is as follows:

[0061]

[0062] In the formula, Indicates the first The magnitude adjustment parameter for the data to be corrected. Indicates the first The magnitude difference between the data to be corrected and the surrounding data to be corrected This represents the mean of the magnitude differences between all data to be corrected and their surrounding data. This represents the maximum value among all the magnitude differences between the data to be corrected and its surrounding data. The parentheses represent Iverson's condition. Output 1 if the condition in the parentheses is true, and output 0 if the condition in the parentheses is false.

[0063] It should be noted that when When, explain the first The magnitude difference between the first data point to be corrected and its surrounding data points is greater than the mean of the magnitude differences between all data points to be corrected and their surrounding data points, further illustrating that the first... The greater the likelihood that the data to be corrected is affected by noise, the more likely it is to be affected by noise. ;when When, explain the first The magnitude difference between the first data point to be corrected and its surrounding data points is less than or equal to the mean of the magnitude differences between all data points to be corrected and their surrounding data points, indicating that the first data point... The likelihood of the data to be corrected being affected by noise is extremely small, therefore, it does not affect the first... The values ​​of the data to be corrected are adjusted at this time. ; The larger the value, the more significant the [value]. The greater the likelihood that the data to be corrected is affected by noise, the more necessary it is to correct the data of the first data point. A large adjustment is made to the value of the data to be corrected, therefore, according to Get the first The magnitude adjustment parameter for the data to be corrected.

[0064] At this point, the amplitude adjustment parameters for each data point to be corrected are obtained.

[0065] Step S003: Based on the numerical differences between each data point to be corrected and the data points to be corrected on its left and right sides, as well as the differences in acquisition time, obtain the slope of each data point to be corrected; based on the difference between the slope of each data point to be corrected and the average slope of all data points to be corrected, obtain the slope adjustment parameter of each data point to be corrected.

[0066] It should be noted that the numerical difference between two data points only reflects the change in value between the two data points, and does not reflect the relationship between the change in value and the difference in the acquisition time of the two data points. This means that a large difference between each data point and its surrounding data points may be due to normal variations over a longer period. Therefore, the slope of each data point is obtained based on the difference in acquisition time between each data point and its surrounding data points, as well as the numerical difference between each data point and its surrounding data points.

[0067] It should be further noted that, due to the rapid appearance and disappearance of noise during signal acquisition, and the significant impact of noise on signal values, the slope of the data to be corrected affected by noise is relatively large. Therefore, based on the slope of each data point to be corrected and the slopes of all data points to be corrected, a slope adjustment parameter is obtained for each data point to be corrected.

[0068] Specifically, to obtain the first The specific formula for calculating the slope of the data to be corrected is as follows:

[0069]

[0070] In the formula, Indicates the first The slope of the data to be corrected Indicates the first The value of the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The collection time of the data to be corrected. Indicates the first The collection time of the data to be corrected. Indicates the first The collection time of the data to be corrected. This represents the absolute value function. Where, when When the value is the first data to be corrected, This is the value of the second data point to be corrected. The time of collection for the second data to be corrected; when When the value is the last piece of data to be corrected, This is the value of the second-to-last data point to be corrected. This is the collection time for the second-to-last data to be corrected.

[0071] It should be noted that, A larger value indicates that the first... When the value of the first data to be corrected differs significantly from that of the previous data to be corrected, the first... The time of collection for the first data point to be corrected is also close to that of the previous data point to be corrected, that is, according to the first... The larger the slope value obtained by comparing the value of the data to be corrected with the value of the data to be corrected to its left and the collection time, the greater the slope value; A larger value indicates that the first... When the value of the first data point to be corrected differs significantly from that of the next data point to be corrected, the first... The collection time of the first data point to be corrected is close to that of the next data point to be corrected, that is, according to the first... The larger the slope value obtained by taking the value of the first data to be corrected, the value of the next data to be corrected, and the acquisition time, the greater the slope value.

[0072] Furthermore, obtain the first The specific calculation formula for the slope adjustment parameter of the data to be corrected is as follows:

[0073]

[0074] In the formula, Indicates the first The slope adjustment parameter for the data to be corrected. Indicates the first The slope of the data to be corrected This represents the mean of the slopes of all the data to be corrected. Represents the arctangent function. It represents 180 degrees.

[0075] It should be noted that when The larger the value, the more significant the slope. The greater the likelihood that the data to be corrected is affected by noise, the more likely it is to be affected by noise. The larger the value, the greater the output range of the arctangent function. arrive Between, therefore , making The output is between -1 and 1; then... This ensures that the slope adjustment parameter for all data to be corrected is between 0 and 1.

[0076] Thus, the slope adjustment parameters for each piece of data to be corrected are obtained.

[0077] Step S004: Determine whether to correct the value of each data point to be corrected based on whether the amplitude difference between each data point to be corrected and its surrounding data points is greater than the average amplitude difference between all data points to be corrected and their surrounding data points. If no correction is made, the value of the data point to be corrected is its corrected value. If correction is made, the corrected value of the data point to be corrected is obtained based on the average amplitude difference between all data points to be corrected and their surrounding data points, the value of the data point to be corrected, the amplitude adjustment parameter, and the slope adjustment parameter. Then, the upper and lower envelopes and the corrected signal data sequence are obtained. The user's respiratory signal and heart rate signal are obtained through the EMD algorithm.

[0078] It should be noted that, because noise has a significant impact on the numerical value of signal data, when When, explain the first The probability that the data to be corrected is affected by noise is extremely small, therefore the data to be corrected is not... The values ​​of the data to be corrected are then corrected. When, explain the first The numerical difference between the data to be corrected and its surrounding data to be corrected is much greater than the average, further illustrating that the first... The number of data points to be corrected is highly likely to be affected by noise, therefore the number of data points to be corrected is... The values ​​of the data to be corrected are corrected.

[0079] It should be further explained that since the polarities of the extreme points are arranged adjacently, the polarities of the first and last data points to be corrected can be obtained by determining whether the nearest extreme point around the first and last data points to be corrected is a maximum or a minimum.

[0080] It should be further noted that when correcting the values ​​of the data to be corrected, the values ​​of the data to be corrected, which are affected by noise, are much larger than the average difference in amplitude between all the data to be corrected and their surrounding data. This makes it possible that when That is, for the first When adjusting the value of a piece of data to be corrected, you can See as, That is, for the first When adjusting the values ​​of the data to be corrected, Based on, according to and right The value is corrected, and then the two are combined to obtain the first value. Correction values ​​for the data to be corrected.

[0081] Specifically, to obtain the first The specific formula for calculating the correction value of each piece of data to be corrected is as follows:

[0082]

[0083] In the formula, Indicates the first Correction values ​​for the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The slope adjustment parameter for the data to be corrected. Indicates the first The magnitude difference between the data to be corrected and the surrounding data to be corrected This represents the mean of the magnitude differences between all data to be corrected and their surrounding data. Indicates the first The magnitude adjustment parameter for the data to be corrected.

[0084] It should be noted that when When, explain the first The numerical difference between the first data point to be corrected and its surrounding data points is quite close to the average of the numerical differences between all data points to be corrected and their surrounding data points, further illustrating that the first... The likelihood of the first data point being affected by noise is almost zero, therefore the second data point is not corrected. The values ​​of the data to be corrected are corrected; when When, explain the first The numerical difference between the first data point to be corrected and its surrounding data points is much greater than the average, further illustrating that the first... The number of data points to be corrected is highly likely to be affected by noise, therefore the number of data points to be corrected is... The values ​​of the data to be corrected are corrected; The larger the value, the more likely it is that the th... The values ​​of the data to be corrected are significantly revised at this time. The smaller the value, the better. The value was adjusted downwards significantly.

[0085] At this point, the numerical value of each data point to be corrected is obtained.

[0086] Furthermore, correcting the value of each piece of data to be corrected does not affect the polarity of the data. That is, if the first... If the data point to be corrected was a local maximum before correction, then when obtaining the upper and lower envelopes, the first... The data point to be corrected will also be a local maximum after correction; if the first... If the data point to be corrected was a local minimum before correction, then when obtaining the upper and lower envelopes, the first... Each data point to be corrected is also considered a local minimum after correction.

[0087] Furthermore, before correcting all the data to be corrected, if the second-to-last data to be corrected is a maximum point, then when obtaining the upper and lower envelopes, the last data to be corrected will be a minimum point after correction; before correcting all the data to be corrected, if the second-to-last data to be corrected is a minimum point, then when obtaining the upper and lower envelopes, the last data to be corrected will be a maximum point after correction.

[0088] Furthermore, before correcting all the data to be corrected, if the second data to be corrected is a maximum point, then when obtaining the upper and lower envelopes, the first data to be corrected will be a minimum point after correction; before correcting all the data to be corrected, if the second data to be corrected is a minimum point, then when obtaining the upper and lower envelopes, the first data to be corrected will be a maximum point after correction.

[0089] Furthermore, in the user's signal data sequence, the value of each data point to be corrected is modified to its corrected value, resulting in a corrected signal data sequence. Then, based on the data points to be corrected that were considered maxima when the upper and lower envelopes were acquired, the upper envelope in the EMD algorithm is obtained; and based on the data points to be corrected that were considered minima when the upper and lower envelopes were acquired, the lower envelope in the EMD algorithm is obtained. The corrected signal data sequence contains both all non-corrected data and all data points to be corrected.

[0090] Furthermore, based on the obtained upper and lower envelopes, the EMD algorithm is used to separate multiple signal data sequences from the corrected signal data sequence. Then, using existing technology, based on the characteristic that signals caused by respiration are low-frequency, slow signals, while signals caused by heartbeats are high-frequency, fluctuating signals, the user's respiratory signal data sequence and heart rate signal data sequence are obtained from the separated multiple signal data sequences. That is, the user's respiratory signal and heart rate signal are obtained. The use of the EMD algorithm to divide a signal data sequence into multiple signal data sequences is a known existing technique and will not be elaborated upon in this embodiment. Similarly, obtaining the respiratory signal data sequence and heart rate signal data sequence from multiple signal data sequences based on characteristics is also a known existing technique and will not be elaborated upon in this embodiment.

[0091] This concludes the embodiment.

Claims

1. A non-contact life signal processing method in a strong clutter environment, characterized in that, include: Acquire the user's signal data sequence; The extreme points in the user's signal data sequence, as well as the first and last signal data, are all recorded as data to be corrected; Based on the numerical difference between each data point to be corrected and the data points to be corrected on its left and right sides, the amplitude difference between each data point to be corrected and the surrounding data points to be corrected is obtained. The amplitude adjustment parameter for each data point to be corrected is obtained based on the amplitude difference between each data point to be corrected and the surrounding data points to be corrected, and the average amplitude difference between all data points to be corrected and the surrounding data points to be corrected. Based on the numerical differences between each data point to be corrected and the data points to be corrected on its left and right sides, as well as the differences in collection time, the slope of each data point to be corrected is obtained; based on the difference between the slope of each data point to be corrected and the mean of the slopes of all data points to be corrected, the slope adjustment parameter of each data point to be corrected is obtained. Whether to correct the value of each data point to be corrected depends on whether the difference in amplitude between each data point to be corrected and its surrounding data points is greater than the average difference in amplitude between all data points to be corrected and their surrounding data points. If no correction is made, the value of the data to be corrected will be the corrected value. If correction is required, the correction value of the data to be corrected is obtained based on the mean of the amplitude differences between all the data to be corrected and the surrounding data to be corrected, the value of the data to be corrected, the amplitude adjustment parameter, and the slope adjustment parameter. Then, the upper and lower envelopes and the corrected signal data sequence are obtained. The user's respiratory signal and heart rate signal are obtained through the EMD algorithm.

2. The non-contact life signal processing method under strong clutter environment according to claim 1, characterized in that, The specific calculation formula for obtaining the amplitude difference between each data point to be corrected and its surrounding data points to be corrected is as follows: In the formula, Indicates the first The magnitude difference between the data to be corrected and the surrounding data to be corrected Indicates the first The value of the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The value of the data to be corrected. This represents the absolute value function.

3. The non-contact life signal processing method under strong clutter environment according to claim 1, characterized in that, The specific calculation formula for obtaining the amplitude adjustment parameter of each data point to be corrected is as follows: In the formula, Indicates the first The magnitude adjustment parameter for the data to be corrected. Indicates the first The magnitude difference between the data to be corrected and the surrounding data to be corrected This represents the mean of the magnitude differences between all data to be corrected and their surrounding data. This represents the maximum value among all the magnitude differences between the data to be corrected and its surrounding data. The brackets represent Iverson.

4. The non-contact life signal processing method under strong clutter environment according to claim 1, characterized in that, The specific formula for calculating the slope of each data point to be corrected is as follows: In the formula, Indicates the first The slope of the data to be corrected Indicates the first The value of the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The collection time of the data to be corrected. Indicates the first The collection time of the data to be corrected. Indicates the first The collection time of the data to be corrected. This represents the absolute value function.

5. The non-contact life signal processing method under strong clutter environment according to claim 1, characterized in that, The specific calculation formula for the slope adjustment parameter of each data point to be corrected is as follows: In the formula, Indicates the first The slope adjustment parameter for the data to be corrected. Indicates the first The slope of the data to be corrected This represents the mean of the slopes of all the data to be corrected. It represents 180 degrees. This represents the arctangent function.

6. The non-contact life signal processing method under strong clutter environment according to claim 1, characterized in that, If no correction is made, the specific calculation formula for the value of the data to be corrected, which is its correction value, is as follows: In the formula, Indicates the first Correction values ​​for the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The magnitude difference between the data to be corrected and the surrounding data to be corrected This represents the mean of the magnitude differences between all data to be corrected and the surrounding data to be corrected.

7. The non-contact life signal processing method under strong clutter environment according to claim 1, characterized in that, The specific formula for calculating the correction value of the data to be corrected is as follows: when hour, In the formula, Indicates the first Correction values ​​for the data to be corrected. Indicates the first The value of the data to be corrected. Indicates the first The slope adjustment parameter for the data to be corrected. Indicates the first The magnitude difference between the data to be corrected and the surrounding data to be corrected This represents the mean of the magnitude differences between all data to be corrected and their surrounding data. Indicates the first The magnitude adjustment parameter for the data to be corrected.

8. The non-contact life signal processing method under strong clutter environment according to claim 1, characterized in that, The specific steps for obtaining the upper and lower envelope lines are as follows: If the first If the value of the data to be corrected is a maximum point before correction, then when obtaining the upper and lower envelopes, the value of the first data point is... Each data point to be corrected will also be considered a maximum point after correction; If the If the value of the data to be corrected is a local minimum point before correction, then when obtaining the upper and lower envelopes, the value of the first data point is... Each data point to be corrected is also considered a local minimum point after correction. Before correcting all the data to be corrected, if the second to last data to be corrected is a maximum point, then when obtaining the upper and lower envelopes, the last data to be corrected will be a minimum point after correction. If the second to last data to be corrected is a minimum point, then when obtaining the upper and lower envelopes, the last data to be corrected will be a maximum point after correction. Before correcting all the data to be corrected, if the second data to be corrected is a maximum point, then when obtaining the upper and lower envelopes, the first data to be corrected will be a minimum point after correction; before correcting all the data to be corrected, if the second data to be corrected is a minimum point, then when obtaining the upper and lower envelopes, the first data to be corrected will be a maximum point after correction. In the user's signal data sequence, the value of each data point to be corrected is modified to its corrected value; The upper envelope in the EMD algorithm is obtained by using the data to be corrected as the maximum point when obtaining the upper and lower envelopes. The lower envelope in the EMD algorithm is obtained by using the data to be corrected as the minimum point when obtaining the upper and lower envelopes.

Images