Method and system for detecting wearing environment of wearable device

Abstract

The invention discloses a wearable device wearing environment detection method and system, and belongs to the technical field of wearable devices, and the method comprises the steps: S1, periodically collecting PPG signals through a built-in sensor of a wearable device, and carrying out the morphological analysis of the PPG signals to extract morphological features; s2, analyzing information theory indexes according to the collected PPG signals; s3, substituting the morphological characteristics and the information theory indexes into the constructed human body wearing discrimination model; and S4, comparing the output value of the human body discrimination model with a set discrimination threshold value, and determining the wearing environment of the wearable device. According to the method provided by the invention, the condition of distinguishing the living body from the non-living body can be effectively and remarkably improved under the condition that the complexity of peripheral hardware is not increased, and a better precondition is provided for accurately measuring the heart rate and the blood oxygen; the calculation method provided by the invention has universality and does not depend on hardware difference.

Description

Technical Field

[0001] The present invention belongs to the technical field of wearable devices, and in particular relates to a wearable device wearing environment detection method and system combining PPG morphology and information theory. Background Art

[0002] In smart watches and other wearable devices, physiological measurement technology based on photoplethysmography (PPG) signals is widely used to monitor users' heart rate, blood oxygen level and other health data. However, when the smart device is not worn correctly on the human body but placed on the surface of other objects, the collected PPG signals will have significant differences, resulting in inaccurate calculation of physiological parameters. Therefore, it is a crucial issue to distinguish whether the device is worn correctly on the human body and ensure the accuracy of the data.

[0003] Traditional smart watches or wearable devices usually determine whether the device is worn or removed by checking whether the PS value is greater than a certain threshold. However, this method alone is not very accurate in practical applications, and it is easy to misidentify non-human surfaces as worn. In addition, some methods use proximity capacitive sensors for judgment, which requires laying copper foil and other hardware materials inside the device, thereby increasing the manufacturing cost and complexity of the device. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the wearable device wearing environment detection method and system provided by the present invention solve the problems of low detection accuracy and increased equipment manufacturing cost and complexity of the existing related methods.

[0005] In order to achieve the above-mentioned purpose of the invention, the technical solution adopted by the present invention is: a method for detecting the wearing environment of a wearable device, comprising the following steps: S1, periodically collect PPG signals through the built-in sensors of wearable devices, and perform morphological analysis on them to extract morphological features; S2, analyzing information theory indicators based on the collected PPG signals; S3, substituting the morphological features and information theory indicators into the constructed human body wearing discrimination model; S4. Compare the output value of the human body discrimination model with the set discrimination threshold to determine the wearing environment of the wearable device.

[0006] Furthermore, the morphological features extracted in step S1 include the rise time and fall time of the pulse wave and the waveform amplitude.

[0007] Furthermore, in step S2, the information theory indicators include entropy, mutual information and redundancy.

[0008] Furthermore, for information theory indicators: The entropy is used to characterize the randomness of the PPG signal, and the analysis method is: The collected PPG signal is divided into several intervals according to the maximum to minimum value of the signal in the segment, and the probability of the signal appearing in each interval is calculated to obtain the entropy, which is expressed as:

[0009] In the formula, represents entropy, Indicates PPG signal In the intensity range i The probability of the interval, n is the number of intervals divided by the PPG signal; The mutual information is used to characterize the correlation between the currently collected PPG signal and the previous PPG signal, and the analysis method is: The collected PPG signal is divided into the first half and the second half, and the PPG signal of the first half and the second half is divided into several intervals according to the maximum value to the minimum value of the signal in the segment, and the corresponding signal occurrence probability is calculated. and ; According to the probability of signal occurrence in the first half and the second half and , where all possible value spaces of the signal are formed into Table, statistics the joint probability distribution of the signal in each interval grid, and then calculate the mutual information, the expression is:

[0010] In the formula, represents mutual information, represents the joint probability distribution of each interval grid, X and Y are the signal sets of the first half and the second half respectively; The redundancy is used to characterize whether the PPG signal contains redundant information, which is expressed as:

[0011] In the formula, Indicates the number of collected PPG signals.

[0012] Furthermore, in step S3, the human body wearing identification model constructed is:

[0013] In the formula, and They represent the rise time and fall time of the pulse wave in the morphological characteristics, respectively. Represents the waveform amplitude in the morphological characteristics, expresst PPG signal at all times The entropy of express t PPG signal at all times and t- PPG signal at 1 moment The mutual information of express t PPG signal at all times The redundancy of , , , and They are , , , and The weight coefficient of .

[0014] Furthermore, in step S3, the weight coefficient Adjust according to the hardware settings of different wearable devices and use the noise intensity measured by the wearable device After that, As The weight adjustment factor.

[0015] Furthermore, in step S4, when the output value of the human body discrimination model is greater than a set discrimination threshold, the wearable device is worn on the human body, otherwise it is placed on the surface of the object.

[0016] A wearable device wearing environment detection system, comprising: PPG signal acquisition module: used to periodically collect PPG signals; Morphological feature extraction module: used to perform morphological analysis on the collected PPG signals and extract morphological features; Information theory index analysis module: used to perform information theory analysis on the collected PPG signals and extract information theory indexes; Wearing identification module: used to substitute the extracted morphological features and information theory indicators into the constructed human wearing identification model, and determine the wearing environment of the wearable device based on its output values.

[0017] Furthermore, the morphological features include the rise time and fall time of the pulse wave and the waveform amplitude.

[0018] The information theory indicators include entropy, mutual information and redundancy.

[0019] Furthermore, in the wearing discrimination module, the output value of the human body discrimination model is compared with a set discrimination threshold. When it is greater than the discrimination threshold, the wearable device is worn on the human body, otherwise it is placed on the surface of the object; The human body wearing discrimination model is expressed as:

[0020] In the formula, and They represent the rise time and fall time of the pulse wave in the morphological characteristics, respectively. Represents the waveform amplitude in the morphological characteristics, express t PPG signal at all times The entropy of express t PPG signal at all times and t- PPG signal at 1 moment The mutual information of express t PPG signal at all times The redundancy of , , , and They are , , , and The weight coefficient of .

[0021] The beneficial effects of the present invention are: 1. High-precision wearing status judgment: The method of the present invention is based on morphological analysis and information theory measurement of photoplethysmography (PPG) signals, which can provide more accurate wearing status judgment. Compared with the traditional method that relies on skin resistance (PS), PPG signals can capture more physiological characteristics, such as heart pulse waveforms, thereby significantly improving the accuracy of judging whether the device is worn on the human body. This method can effectively avoid false alarms caused by PS alone, especially in cases where the user is sweating or the contact is poor.

[0022] 2. No need to increase hardware costs: Compared with some methods that require proximity capacitive sensors to determine whether the device is worn on the human body, the present invention does not require additional hardware, such as copper foil, contact electrodes, etc. Only by collecting data through existing PPG sensors and performing algorithm processing can a high accuracy be achieved. This greatly reduces manufacturing costs, avoids major changes to existing device designs, and has high market adaptability.

[0023] 3. Flexibility to adapt to different hardware devices: Through experiments, most of the weight parameters in the human wearing discrimination model in the present invention are stable, and only the α2 parameter related to amplitude fluctuation needs to be adjusted according to different hardware devices. A method is further given to equivalently control the parameter without being affected by the device. This flexible parameter adjustment design can adapt to the differences in light intensity and unit measurement of different devices, so that the method can be applied to smart watch devices of different brands and models; this flexibility of the system ensures cross-device portability and adapts to the needs of different hardware platforms.

[0024] 4. Stable threshold judgment: The discrimination threshold θ in the present invention has been determined through experiments, which can ensure consistent judgment standards in different devices and scenarios. Stable threshold settings reduce the complexity of the system in practical applications, avoid frequent adjustments, and lower the threshold for use.

[0025] 5. Reduce false positives and improve user experience: Combining the physiological characteristics of PPG signals and information theory analysis, the present invention can effectively reduce false alarms and ensure that the device can work reliably when worn. When the device is placed on the surface of an object, the discriminant model of the present invention can also identify it in time to avoid unnecessary data recording or erroneous health monitoring data caused by wearing errors. Good accuracy and reduced false alarm characteristics enhance the user's sense of trust and bring a better user experience.

[0026] 6. Applicable to various application scenarios: The present invention is not only applicable to smart watches, but also to other wearable health monitoring devices, such as smart bracelets, heart rate monitors, etc. This technology can provide high-precision wearing status judgment for a wide range of wearable devices, thereby ensuring the reliability of health monitoring, sports tracking and medical data collection. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a flow chart of the wearable device wearing environment detection method provided by the present invention. DETAILED DESCRIPTION

[0028] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0029] The embodiment of the present invention provides a method for detecting a wearing environment of a wearable device, such as Figure 1 As shown, the following steps are included: S1, periodically collect PPG signals through the built-in sensors of wearable devices, and perform morphological analysis on them to extract morphological features; S2, analyzing information theory indicators based on the collected PPG signals; S3, substituting the morphological features and information theory indicators into the constructed human body wearing discrimination model; S4. Compare the output value of the human body discrimination model with the set discrimination threshold to determine the wearing environment of the wearable device.

[0030] In step S1 of the embodiment of the present invention, the wearable device periodically collects PPG signals through the built-in photoelectric sensor. When the device is worn on the human body, the PPG signal will show an obvious heartbeat pulse waveform, and when the device is placed on the surface of an object, the PPG signal will show lower volatility or random noise.

[0031] In an optional embodiment of the present invention, the wearable device may be a smart watch, a smart bracelet, a heart rate monitor or other wearable health monitoring device.

[0032] In step S1 of the embodiment of the present invention, the extracted morphological features include the rise time and fall time of the pulse wave and the waveform amplitude.

[0033] In a specific embodiment of the present invention, when the device is worn on a human body, the heart pulse wave will form a specific rise time and fall time .

[0034] The waveform amplitude of the PPG signal when worn on the human body It will fluctuate periodically with the pulse, and when it is on the surface of an object, the fluctuation of the amplitude is significantly reduced.

[0035] By analyzing the periodicity of the signal, its autocorrelation coefficient in a given time window is calculated. If a periodic heartbeat signal is detected, it means that the device is worn on the human body.

[0036] In step S2 of the embodiment of the present invention, the information theory indicators include entropy, mutual information and redundancy; Specifically, entropy is used to characterize the randomness of PPG signals. The higher the entropy, the greater the randomness of the surface signal. The lower the entropy, the stronger the structure and regularity of the surface signal. The entropy analysis method in this embodiment is: The collected PPG signal is divided into several intervals according to the maximum to minimum value of the signal in the segment, and the probability of the signal appearing in each interval is calculated to obtain the entropy, which is expressed as:

[0037] In the formula, represents entropy, Indicates PPG signal In the intensity range i The probability of the interval, n is the number of intervals divided by the PPG signal.

[0038] Mutual information is used to characterize the correlation between the current collected PPG signal and the previous PPG signal. The larger the mutual information, the more relevant information there is between the current signal and the past signal. The mutual information analysis method in this embodiment is: The collected PPG signal is divided into the first half and the second half, and the PPG signal of the first half and the second half is divided into several intervals according to the maximum value to the minimum value of the signal in the segment, and the corresponding signal occurrence probability is calculated. and ; According to the probability of signal occurrence in the first half and the second half and , where all possible value spaces of the signal are formed into Table, statistics the joint probability distribution of the signal in each interval grid, and then calculate the mutual information, the expression is:

[0039] In the formula, represents mutual information, represents the joint probability distribution of each interval grid, X and Y They are the signal sets of the first half and the second half respectively.

[0040] Redundancy is used to characterize whether the PPG signal contains redundant information. It measures the signal redundancy in different time windows. The redundancy in this embodiment is expressed as:

[0041] In the formula, Indicates the number of collected PPG signals.

[0042] In an optional embodiment of the present invention, l The value is 250 points.

[0043] In step S3 of the embodiment of the present invention, a human body wearing discrimination model based on morphological and information theory metrics is constructed, and its output value is used to determine whether the device is worn on the human body. In this embodiment, the constructed human body wearing discrimination model is:

[0044] In the formula, and They represent the rise time and fall time of the pulse wave in the morphological characteristics, respectively. Represents the waveform amplitude in the morphological characteristics, express t PPG signal at all times The entropy of express t PPG signal at all times and t- PPG signal at 1 moment The mutual information of express t PPG signal at all times The redundancy of , , , and They are , , , and The weight coefficient of .

[0045] Furthermore, after experimental verification, since the parameter is the human body or the statistical measurement characteristics of signal information, it is different from the representation of signal strength without equipment and does not change with different equipment. Therefore, in the weight coefficient of the above model, , , and Relatively stable, only the weight coefficient Needs adjustment.

[0046] Specifically, in this embodiment, different devices have different light intensity measurement units, which leads to different ways of expressing the same signal amplitude. Therefore, in this embodiment, adjustments are made according to the hardware settings of different wearable devices, and the noise intensity measured by the wearable device is After that, As The weight adjustment coefficient is used to cancel the difference between devices.

[0047] In an optional embodiment of the present invention, the values ​​of the above parameters are: [0.28~0.31], 0.5, is 0.2, [0.19~0.22], is [0.85~0.1], is 0.1, and the discrimination threshold is 14.5. It should be noted that in all the above parameters, if a range is given, the default median is preferred. It can be adjusted according to the actual effect, but it is not recommended to exceed the interval range.

[0048] In step S4 of the embodiment of the present invention, based on the above model, when the output value of the human body discrimination model is greater than the set discrimination threshold, the wearable device is worn on the human body, otherwise it is placed on the surface of the object.

[0049] The embodiment of the present invention further provides a wearable device wearing environment detection system based on the wearable device wearing environment detection method, comprising: PPG signal acquisition module: used to periodically collect PPG signals; Morphological feature extraction module: used to perform morphological analysis on the collected PPG signals and extract morphological features; Information theory index analysis module: used to perform information theory analysis on the collected PPG signals and extract information theory indexes; Wearing identification module: used to substitute the extracted morphological features and information theory indicators into the constructed human wearing identification model, and determine the wearing environment of the wearable device based on its output values.

[0050] In the embodiment of the present invention, the morphological features include the rise time and fall time of the pulse wave and the waveform amplitude; specifically, when the device is worn on the human body, the heart pulse wave will form a specific rise time and fall time. and fall time ; Under human wear status, the waveform amplitude of the PPG signal It will fluctuate periodically with the pulse, and when it is on the surface of an object, the fluctuation of the amplitude is significantly reduced; by analyzing the periodicity of the signal, its autocorrelation coefficient in a given time window is calculated. If a periodic heartbeat signal is detected, it means that the device is worn on the human body.

[0051] In this embodiment, information theory indicators include entropy, mutual information and redundancy; specifically, entropy is used to characterize the randomness of the PPG signal. The higher the entropy, the greater the randomness of the surface signal, and the lower the entropy, the stronger the structure and regularity of the surface signal; mutual information is used to characterize the correlation between the current collected PPG signal and the previous PPG signal. The larger the mutual information, the more relevant information the current signal has with the past signal; redundancy is used to characterize whether the PPG signal contains redundant information, which measures the signal redundancy in different time windows.

[0052] In the embodiment of the present invention, in the wearing discrimination module, the output value of the human body discrimination model is compared with the set discrimination threshold. When it is greater than the discrimination threshold, the wearable device is worn on the human body, otherwise it is placed on the surface of the object; The human body wearing discrimination model is expressed as:

[0053] In the formula, and They represent the rise time and fall time of the pulse wave in the morphological characteristics, respectively. Represents the waveform amplitude in the morphological characteristics, express t PPG signal at all times The entropy of express t PPG signal at all times and t- PPG signal at 1 moment The mutual information of express t PPG signal at all times The redundancy of , , , and They are , , , and The weight coefficient of .

[0054] Furthermore, after experimental verification, since the parameter is the human body or the statistical measurement characteristics of signal information, it is different from the representation of signal strength without equipment and does not change with different equipment. Therefore, in the weight coefficient of the above model, , , and Relatively stable, only the weight coefficient Needs adjustment.

[0055] Specifically, in this embodiment, different devices have different light intensity measurement units, which leads to different ways of expressing the same signal amplitude. Therefore, in this embodiment, adjustments are made according to the hardware settings of different wearable devices, and the noise intensity measured by the wearable device is After that, As The weight adjustment coefficient is used to cancel the difference between devices.

[0056] In the embodiment of the present invention, a specific application example of the above method is provided: In this embodiment, the above method is applied to a smart watch product. For the PPG signal, the acquisition time is set to 10 seconds, the sampling rate is 25 Hz, and the signal is subjected to a 0.4 Hz to 4 Hz bandpass filter. According to the above method, the discrimination results obtained are shown in Table 1: Table 1: Wearable device wearing environment detection results in different scenarios

[0057] Among the above 34 scenes, the recognition probability of 22 scenes was significantly improved from being completely unrecognizable by using the method of the present invention, and the living objects were effectively distinguished from the non-living objects.

[0058] The above method provided by the present invention can effectively improve the ability to distinguish between living and non-living objects without increasing the complexity of peripheral hardware, providing a better prerequisite for accurately measuring heart rate and blood oxygen; the calculation method provided by the present invention is universal and does not rely on hardware differences.

[0059] The present invention uses specific embodiments to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

[0060] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A wearable device wearing environment detection method, characterized in that: The following steps are involved: S1, periodically collect PPG signals through the built-in sensors of wearable devices, and perform morphological analysis on them to extract morphological features; S2, analyzing information theory indicators based on the collected PPG signals; S3, substituting the morphological features and information theory indicators into the constructed human body wearing discrimination model; S4. Compare the output value of the human body discrimination model with the set discrimination threshold to determine the wearing environment of the wearable device.

2. The wearable device wearing environment detection method according to claim 1, characterized in that: The morphological features extracted in step S1 include the rise time and fall time of the pulse wave and the waveform amplitude.

3. The wearable device wearing environment detection method according to claim 1, characterized in that: In step S2, the information theory indicators include entropy, mutual information and redundancy.

4. The wearable device wearing environment detection method according to claim 2, characterized in that: For information theoretic metrics: The entropy is used to characterize the randomness of the PPG signal, and the analysis method is: The collected PPG signal is divided into several intervals according to the maximum to minimum value of the signal in the segment, and the probability of the signal appearing in each interval is calculated to obtain the entropy, which is expressed as: In the formula, represents entropy, Indicates PPG signal In the intensity range i The probability of the interval, n is the number of intervals divided by the PPG signal; The mutual information is used to characterize the correlation between the currently collected PPG signal and the previous PPG signal, and the analysis method is: The collected PPG signal is divided into the first half and the second half, and the PPG signal of the first half and the second half is divided into several intervals according to the maximum value to the minimum value of the signal in the segment, and the corresponding signal occurrence probability is calculated. and ; According to the probability of signal occurrence in the first half and the second half and , where all possible value spaces of the signal are formed into Table, statistics the joint probability distribution of the signal in each interval grid, and then calculate the mutual information, the expression is: In the formula, represents mutual information, represents the joint probability distribution of each interval grid, X and Y are the signal sets of the first half and the second half respectively; The redundancy is used to characterize whether the PPG signal contains redundant information, which is expressed as: In the formula, Indicates the number of collected PPG signals.

5. The wearable device wearing environment detection method according to claim 1, characterized in that: In step S3, the human body wearing identification model constructed is: In the formula, and They represent the rise time and fall time of the pulse wave in the morphological characteristics, respectively. Represents the waveform amplitude in the morphological characteristics, express t PPG signal at all times The entropy of express t PPG signal at all times and t- PPG signal at 1 moment The mutual information of express t PPG signal at all times The redundancy of , , , and They are , , , and The weight coefficient of .

6. The wearable device wearing environment detection method according to claim 5, characterized in that: In step S3, the weight coefficient Adjust according to the hardware settings of different wearable devices and use the noise intensity measured by the wearable device After that, As The weight adjustment factor.

7. The wearable device wearing environment detection method according to claim 1, characterized in that: In step S4, when the output value of the human body discrimination model is greater than a set discrimination threshold, the wearable device is worn on the human body, otherwise it is placed on the surface of the object.

8. A wearable device wearing environment detection system, implemented based on the wearable device wearing environment detection method according to any one of claims 1 to 7, characterized in that: include: PPG signal acquisition module: used to periodically collect PPG signals; Morphological feature extraction module: used to perform morphological analysis on the collected PPG signals and extract morphological features; Information theory index analysis module: used to perform information theory analysis on the collected PPG signals and extract information theory indexes; Wearing identification module: used to substitute the extracted morphological features and information theory indicators into the constructed human wearing identification model, and determine the wearing environment of the wearable device based on its output values.

9. The wearable device wearing environment detection system according to claim 8, characterized in that: The morphological characteristics include the rise time and fall time of the pulse wave and the waveform amplitude; The information theory indicators include entropy, mutual information and redundancy.

10. The wearable device wearing environment detection system according to claim 8, characterized in that: In the wearing identification module, the output value of the human body identification model is compared with the set identification threshold. When it is greater than the identification threshold, the wearable device is worn on the human body, otherwise it is placed on the surface of the object; The human body wearing discrimination model is expressed as: In the formula, and They represent the rise time and fall time of the pulse wave in the morphological characteristics, respectively. Represents the waveform amplitude in the morphological characteristics, express t PPG signal at all times The entropy of express t PPG signal at all times and t- PPG signal at 1 moment The mutual information of express t PPG signal at all times The redundancy of , , , and They are , , , and The weight coefficient of .

Images