Method, device and electronic equipment for detecting human fatigue state information

Abstract

This application discloses a method, apparatus, and electronic device for detecting human fatigue state information, belonging to the field of Internet of Things (IoT) technology. The method includes: acquiring a gait dataset and multiple preset peak thresholds; determining multiple anomaly ratios in the gait dataset based on the multiple preset peak thresholds, wherein each anomaly ratio corresponds one-to-one with a preset peak threshold; determining a target anomaly ratio from the multiple anomaly ratios, and setting the preset peak threshold corresponding to the target anomaly ratio as the target peak threshold, wherein the target anomaly ratio is the smallest among the multiple anomaly ratios; performing target processing on the gait dataset based on the target peak threshold to obtain a target dataset, wherein the target processing is used to correct the data in the gait dataset; and determining human fatigue state information based on the target dataset.

Description

Technical Field

[0001] This application belongs to the field of Internet of Things (IoT) technology, specifically relating to a method, device, and electronic device for detecting human fatigue state information. Background Technology

[0002] The human body needs to use lower limb strength in most sports, and the condition of the lower limbs has a great impact on the performance of the sport. With the accumulation of fatigue, excessive exercise and a large amount of fatigue can have adverse effects on safety and health. If the fatigue status of the human body can be monitored in real time, it can prevent the occurrence of dangers.

[0003] Existing fatigue condition detection technologies mainly involve optical capture equipment and force-measuring plates, but the accuracy of these methods is relatively poor. Summary of the Invention

[0004] The purpose of this application is to provide a method, apparatus, and electronic device for detecting human fatigue state information, which can solve the problem of poor accuracy of existing human fatigue state detection results.

[0005] In a first aspect, embodiments of this application provide a method for detecting human fatigue state information, the method comprising:

[0006] Acquire gait datasets and multiple preset peak thresholds;

[0007] Based on the multiple preset peak thresholds, multiple heterogeneity ratios are determined in the gait dataset, and the heterogeneity ratios correspond one-to-one with the preset peak thresholds.

[0008] A target heterogeneous ratio is determined from the plurality of heterogeneous ratios, and a preset peak threshold corresponding to the target heterogeneous ratio is determined as the target peak threshold, wherein the target heterogeneous ratio is the smallest of the plurality of heterogeneous ratios;

[0009] Based on the target peak threshold, the gait dataset is subjected to target processing to obtain a target dataset. The target processing is used to correct the data in the gait dataset.

[0010] Based on the target dataset, the fatigue state information of the human body is determined.

[0011] Secondly, embodiments of this application provide a device for detecting human fatigue state information, the device comprising:

[0012] The acquisition module is used to acquire gait datasets and multiple preset peak thresholds;

[0013] The first determining module is used to determine multiple heterogeneity ratios in the gait dataset based on the multiple preset peak thresholds, wherein the heterogeneity ratios correspond one-to-one with the preset peak thresholds.

[0014] The second determining module is used to determine a target outlier ratio from the plurality of outlier ratios, and to determine a preset peak threshold corresponding to the target outlier ratio as the target peak threshold, wherein the target outlier ratio is the smallest of the plurality of outlier ratios;

[0015] The processing module is used to perform target processing on the gait dataset based on the target peak threshold to obtain a target dataset, wherein the target processing is used to correct the data in the gait dataset;

[0016] The third determining module is used to determine the fatigue state information of the human body based on the target dataset.

[0017] Thirdly, embodiments of this application provide an electronic device, which includes a processor and a memory, wherein the memory stores a program or instructions that can run on the processor, and the program or instructions, when executed by the processor, implement the steps of the method described in the first aspect.

[0018] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0019] In this embodiment, each preset peak threshold can be assigned a corresponding heterogeneity ratio in the gait dataset. The preset peak threshold corresponding to the minimum target heterogeneity ratio is determined as the target peak threshold. Target processing is then performed on the gait dataset based on the target peak threshold to correct the data, thereby obtaining the target dataset. Finally, the target dataset is analyzed to obtain the fatigue state information of the human body. Because the target heterogeneity ratio is the smallest, the target dataset obtained by processing the gait dataset based on the target peak threshold has the highest degree of fit with the actual gait of the human body. Therefore, the fatigue state information of the human body determined based on the target dataset best matches the actual fatigue state of the human body, improving the accuracy of the detection results of the human fatigue state. Attached Figure Description

[0020] Figure 1 A flowchart illustrating the method for detecting human fatigue state information provided in this application embodiment;

[0021] Figure 2 A schematic diagram of the structure of the human fatigue state information detection device provided in the embodiments of this application;

[0022] Figure 3This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0023] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0024] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0025] The methods provided in this application will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0026] like Figure 1 As shown in the embodiments of this application, the method for detecting human fatigue state information includes the following steps:

[0027] Step S1: Obtain the gait dataset and multiple preset peak thresholds.

[0028] The gait dataset can come from inertial sensors in portable mobile devices, such as smartphones or fitness trackers. When a user carries a portable mobile device, the inertial sensors continuously record the user's gait data as they walk. Balance and posture control are level mechanisms of movement; muscle fatigue affects the stability and variability of movement, and the accumulation of fatigue causes changes in human gait. Accordingly, gait data can reflect human gait, and based on the gait, the user's fatigue state can be determined.

[0029] Gait datasets include at least one of the following: acceleration, angular velocity, angle, etc.

[0030] Furthermore, the values ​​of the multiple preset peak thresholds are all different. Different preset peak threshold values ​​allow for the determination of different numbers and values ​​of peak values ​​within the dataset. The preset peak thresholds can be set by the user in real time or configured by the computer based on empirical values.

[0031] After obtaining the gait dataset, the dataset can be filtered. Data filtering mainly uses filtering techniques from digital signal processing and image processing. Data filtering can remove introduced error points, redundant points, and measurement noise points from the gait dataset, thereby improving the accuracy of subsequent data analysis of the gait dataset.

[0032] After filtering the gait dataset, normalization is performed on the gait dataset, which transforms the originally dimensional data in the gait dataset into dimensionless scalars.

[0033] Step S2: Based on the multiple preset peak thresholds, determine multiple anomaly ratios in the gait dataset, wherein each anomaly ratio corresponds one-to-one with a preset peak threshold.

[0034] For a gait dataset, a preset peak threshold will correspond to multiple peaks. Based on the multiple peaks, multiple cycle length data can be determined, thereby determining the heterogeneity ratio corresponding to the multiple cycle length data. The heterogeneity ratio is used to represent the dispersion of the data. The larger the heterogeneity ratio, the higher the dispersion, and the smaller the heterogeneity ratio, the lower the dispersion.

[0035] Step S3: Determine the target dissident ratio from the plurality of dissident ratios, and determine the preset peak threshold corresponding to the target dissident ratio as the target peak threshold, wherein the target dissident ratio is the smallest among the plurality of dissident ratios.

[0036] The target heterogeneity ratio is the smallest among the multiple heterogeneity ratios. Therefore, the dispersion of the period length data corresponding to the target heterogeneity ratio is the smallest. Human gait exhibits periodicity during walking. The smallest dispersion in the period length data indicates the smallest difference between the various period lengths, meaning the period lengths are closest. The closer the period lengths are to each other, the higher the fit between the data processed based on the preset peak threshold and the human gait. A higher fit leads to more accurate gait analysis results.

[0037] Step S4: Based on the target peak threshold, perform target processing on the gait dataset to obtain a target dataset. The target processing is used to correct the data in the gait dataset.

[0038] Determining the target peak threshold by minimizing the target heterogeneity ratio, and then processing the gait dataset based on this threshold, can improve the fit between the processed data and real human gait. Specifically, the gait dataset is corrected to obtain the target dataset. Data correction can involve removing some gait data from the dataset based on the target peak threshold, or adding some gait data. This eliminates interference from abnormal conditions on the gait data, thereby improving the accuracy of subsequent analyses of human gait and fatigue states.

[0039] Step S5: Based on the target dataset, determine the fatigue state information of the human body.

[0040] Based on the target dataset after data correction of the gait dataset, the average value of the target dataset can be calculated. The average value is then compared with multiple preset value ranges, each of which corresponds to a human fatigue state. When the average value of the target dataset falls within the target value range, the human fatigue state corresponding to that target dataset is the human fatigue state corresponding to the target value range, thereby determining the human fatigue state information. Based on the fatigue state information, a prompt is issued to the user to prevent the occurrence of danger.

[0041] In this embodiment, each preset peak threshold can be assigned a corresponding heterogeneity ratio in the gait dataset. The preset peak threshold corresponding to the minimum target heterogeneity ratio is determined as the target peak threshold. Target processing is then performed on the gait dataset based on the target peak threshold to correct the data, thereby obtaining the target dataset. Finally, the target dataset is analyzed to obtain the fatigue state information of the human body. Because the target heterogeneity ratio is the smallest, the target dataset obtained by processing the gait dataset based on the target peak threshold has the highest degree of fit with the actual gait of the human body. Therefore, the fatigue state information of the human body determined based on the target dataset best matches the actual fatigue state of the human body, improving the accuracy of the fatigue state detection results.

[0042] Optionally, determining multiple heterogeneity ratios in the gait dataset includes:

[0043] Obtain the first peak threshold among the plurality of preset peak thresholds;

[0044] Based on the first peak threshold, multiple first peaks are determined in the gait dataset, and all of the first peaks are greater than the first peak threshold.

[0045] Based on the plurality of first peak values, a plurality of first cycle lengths are determined, wherein the first cycle length is the distance between any two adjacent first peak values;

[0046] Based on the plurality of first period lengths, statistical target parameters are calculated, including the following parameters: the frequency of each first period length and the frequency corresponding to the mode of the plurality of first period lengths;

[0047] Based on the frequency of each first period length and the frequency of the mode of the plurality of first period lengths, the heterogeneity ratio of the period length corresponding to the first peak threshold is determined.

[0048] The first peak threshold is any one of the multiple preset peak thresholds. Based on the first peak threshold, a peak detection algorithm is applied to the gait dataset to obtain multiple first peaks, where each first peak is greater than the first peak threshold. The data between any two adjacent first peaks represents the gait data within one gait cycle. Each first peak is the starting point of the previous gait cycle and the ending point of the next gait cycle. The length between any two adjacent first peaks is the length of one first cycle.

[0049] Based on the first peak threshold T, the first peak result of the peak detection algorithm applied to the gait dataset is V(i), where i = 1, 2, 3, ..., n, and i represents the position of the first peak in the gait dataset. The first cycle length d(i) = V(i+1) - V(i), where i = 1, 2, 3, ..., n-1. The heterogeneity ratio of the first cycle length data is:

[0050] Vr=∑fi-fm / ∑fi=1-fm / ∑fi

[0051] Where fi is the frequency of each first period length, and fm is the frequency of the mode of the data for the first period length.

[0052] Furthermore, the preferred range for the preset peak threshold is (0.2, 0.8). When the preset peak threshold is too large, for example, greater than 0.8, only a few peaks can be detected, or even just one peak, resulting in a non-monopoly ratio of 0 for the period length. When the preset peak threshold is too small, for example, less than 0.2, all detected points are peaks, resulting in a non-monopoly ratio of 0 for the period length as well. Neither of these scenarios is realistic. Therefore, selecting the range (0.2, 0.8) can first eliminate interference from data that does not conform to reality, simplifying the process of selecting the target peak threshold, thereby improving the detection efficiency of the method in this application.

[0053] Multiple heterogeneity ratios are determined using the above method. In subsequent processes, the target peak threshold is determined by the minimum target heterogeneity ratio. The gait dataset is then processed based on the target peak threshold, ensuring a high degree of fit between the obtained target dataset and the real human gait.

[0054] Optionally, the step of performing target processing on the gait dataset based on the target peak threshold to obtain the target dataset includes:

[0055] Based on the target peak threshold, invalid peaks in the gait dataset are eliminated or hidden peaks are inserted into the gait dataset to obtain the target dataset.

[0056] Invalid peaks and hidden peaks are caused by abnormal conditions such as noise and malfunction of measuring equipment. By eliminating invalid peaks and inserting hidden peaks to correct the gait dataset, the interference of abnormal conditions on gait data can be eliminated, thereby improving the accuracy of subsequent analysis results on human gait and fatigue status.

[0057] Further, the step of eliminating invalid peaks in the gait dataset or inserting hidden peaks in the gait dataset based on the target peak threshold to obtain the target dataset includes:

[0058] Multiple second cycle lengths are determined based on the target peak threshold, where the second cycle length is the distance between two adjacent target peaks, and the target peak is the peak value in the gait dataset that is greater than the target peak threshold.

[0059] Calculate the average period length of the plurality of second period lengths;

[0060] In the gait dataset, a target data segment corresponding to the target cycle length is determined. The target cycle length is greater than a first value and less than a second value. The first value is the product of the difference between the preset value and the tolerance and the average cycle length. The second value is the product of the sum of the preset value and the tolerance and the average cycle length.

[0061] If the second cycle length on both sides of the target data segment in the gait dataset is less than the first value, the target peak corresponding to the second cycle length is eliminated; or, if the second cycle length on both sides of the target data segment in the gait dataset is greater than the second value, a hidden peak is added to the gait dataset to obtain the target dataset.

[0062] Based on the target peak threshold, a peak detection algorithm is applied to the gait dataset to obtain multiple target peaks. The length between any two adjacent target peaks is defined as the second cycle length, and the average cycle length E of the multiple second cycle lengths is calculated. The first and second values ​​are determined based on the average cycle length E and the tolerance of the gait dataset. The tolerance of the gait dataset is:

[0063]

[0064] Where K represents the kurtosis of the gait dataset, and the value of K is [1, +∞). The kurtosis K is determined as follows:

[0065]

[0066] Where Xi is the gait data sequence, σ is the standard deviation of the gait data, μ is the mean of the gait data, and n is the number of gait data.

[0067] The first value is (1-δ)E. The second value is (1+δ)E. The default value is 1.

[0068] The target cycle length is determined from multiple second cycle lengths based on the first and second values. The target cycle corresponding to the target cycle length is found in the gait dataset, and the gait data corresponding to the target cycle is determined as the target data segment.

[0069] The target cycle length must be greater than the first value and less than the second value. It should be noted that when there are multiple target cycle lengths that meet the conditions among multiple second cycle lengths, any one of the corresponding gait data segments can be selected as the target data segment.

[0070] The gait dataset is analyzed based on the target data segment. Specifically, the lengths of the second cycles on both sides of the target data segment are analyzed sequentially. If the length of the second cycle on both sides of the target data segment is less than the first value, it indicates that the second cycle length is too small, and the gait cycle data is likely caused by noise. The target peak corresponding to this gait cycle data is an invalid peak, and the invalid peak is removed from the gait dataset. If the length of the second cycle on both sides of the target data segment is greater than the second value, it indicates that the second cycle length is too long, and the gait cycle data contains hidden cycles that were not counted due to equipment limitations. Hidden peaks are added to this gait cycle. If the length of the second cycle on both sides of the target data segment is greater than the first value but less than the second value, it indicates that the gait cycle corresponding to this second cycle length is a valid cycle.

[0071] Eliminating invalid peaks can remove the adverse effects of noise on gait datasets, while adding hidden peaks can eliminate the adverse effects of incomplete data collection due to equipment limitations, thereby improving the accuracy of data analysis.

[0072] Optionally, if the second cycle length on both sides of the target data segment in the gait dataset is greater than the second value, a hidden peak is added to the gait dataset to obtain the target dataset, including:

[0073] The target peak value corresponding to the second cycle length is determined as the third peak value;

[0074] A fourth peak is determined from the gait dataset. This fourth peak is located on the same side of the target data segment as the third peak, and is the farthest from the target data segment with a value greater than 0.

[0075] The period length between the third peak and the fourth peak is defined as the third period length;

[0076] Based on the third cycle length and the average cycle length, a third value is determined, which is the quotient of the third cycle length and the average cycle length.

[0077] If the length of the third period is greater than the product of the first value and the third value, and the length of the third period is less than the product of the second value and the third value, a hidden peak is inserted between the third peak and the fourth peak, wherein the number of the hidden peaks is the same as the third value.

[0078] The third cycle length P = Va - Vb, where Va represents the third peak value corresponding to the second cycle length which is greater than the second value, and Vb represents the non-zero peak value that is located on the same side of the target data segment as the third peak value and is the farthest from the target data segment.

[0079] The third value is: The ROUND function rounds a number to the nearest whole number.

[0080] The hidden peak between the third and fourth peaks is determined based on the first, second, and third values, as well as the length of the third period. Specifically:

[0081]

[0082]

[0083] else set S(i)=0

[0084] Where (1-δ)E represents the first value and (1+δ)E represents the second value. S(i) represents the state of the target peak corresponding to the second period length which is greater than the second value. When the target peak state is 0, it means that the target peak is an invalid peak and needs to be eliminated. When the target peak is a positive integer, it means that there are hidden periods and hidden peaks that need to be inserted in the gait period corresponding to the second period length which is greater than the second value.

[0085] If the length of the third cycle is greater than the product of the first and third values, and less than the product of the second and third values, a hidden peak is inserted between the third and fourth peak values. Otherwise, the target peak value is output as 0, thus eliminating the target peak value.

[0086] Furthermore, the number of hidden peaks is consistent with the third value. Based on the number of hidden peaks, hidden peaks are inserted proportionally between the third and fourth peaks. The value of the hidden peaks can be the average of the third and fourth peaks, or it can be a geometric or arithmetic sequence with the third peak as the starting point and the fourth peak as the ending point. By inserting hidden peaks in the gait cycle corresponding to the second cycle length, which is greater than the second value, the fitting degree between the target dataset and the real human gait is improved, thereby improving the accuracy of the detection results of the detection method in this embodiment.

[0087] Optionally, the number of gait datasets is multiple, and before performing target processing on the gait datasets based on the target peak threshold to obtain the target dataset, the method further includes:

[0088] Obtain the target peak threshold corresponding to each of the multiple gait datasets, and the variance of the cycle length corresponding to each gait dataset under the corresponding target peak threshold;

[0089] A reference gait dataset is determined from a plurality of said gait datasets, wherein the variance of the cycle length corresponding to the reference gait dataset is minimized under the corresponding target peak threshold;

[0090] The step of eliminating invalid peaks in the gait dataset or inserting hidden peaks in the gait dataset based on the target peak threshold to obtain the target dataset includes:

[0091] Based on the target peak threshold corresponding to the reference gait dataset, invalid peaks in the reference gait dataset are eliminated or hidden peaks are inserted into the reference gait dataset to obtain the target dataset.

[0092] When the gait dataset has multiple data types or a large number of datasets, a target peak threshold is obtained for each gait dataset. Based on the target peak threshold, the variance of the corresponding cycle length is determined. The smallest variance in cycle length indicates less fluctuation in the dataset, meaning the cycle length is closer to the actual human gait and has the highest degree of fit. Therefore, the gait dataset with the smallest variance in cycle length is selected as the reference gait dataset. Data processing and analysis are then performed on the reference gait dataset, further improving the accuracy of the detection results of the detection method in this embodiment.

[0093] Optionally, determining the fatigue state information of the human body based on the target dataset includes:

[0094] The target dataset is input into a pre-trained fatigue state recognition model to identify the fatigue state of the human body, so that the fatigue state recognition model outputs the fatigue state result of the human body.

[0095] The fatigue state recognition model is a neural network model used to identify the fatigue state of the human body based on a dataset.

[0096] By using a pre-trained fatigue state recognition model to analyze and identify the target dataset, the fatigue state of the human body can be obtained quickly, thus improving the detection efficiency of the method in this embodiment.

[0097] Optionally, the method further includes inputting the target dataset into a pre-trained fatigue state recognition model to identify the fatigue state of the human body, so that before the fatigue state recognition model outputs the fatigue state information of the human body:

[0098] Acquire information on human fatigue status and the corresponding dataset;

[0099] The dataset corresponding to the human fatigue state is used as the feature value, and the human fatigue state information is used as the label value to train the initial model.

[0100] If the initial model after training meets the preset conditions, the initial model after training will be determined as the fatigue state recognition model.

[0101] A large amount of human fatigue state information and corresponding datasets are input into an initial model for training. The initial model can be a support vector machine, random forest, gradient boosting tree, or similar model. Multiple trained initial models can be obtained based on various algorithms. From these models, a fatigue state recognition model is selected. The accuracy of this model is greater than or equal to a preset accuracy, and the mean absolute error (MAE) is less than a preset error. The accuracy rate indicates the model's accuracy in identifying fatigue states, while the MAE indicates the degree of deviation in the model's identification. Inputting the target dataset into the fatigue state recognition model improves the accuracy of the model's recognition results, thereby improving the accuracy of the detection results of the method in this embodiment.

[0102] The execution subject of the method for detecting human fatigue state information provided in this application embodiment can be a human fatigue state information detection device. Taking the execution of the human fatigue state information detection method by a human fatigue state information detection device as an example, in conjunction with the appendix... Figure 2 This application describes the human fatigue state information detection device 200 provided in the embodiments of the present application. The human fatigue state information detection device 200 includes:

[0103] The acquisition module 201 is used to acquire gait datasets and multiple preset peak thresholds;

[0104] The first determining module 202 is used to determine multiple heterogeneity ratios in the gait dataset based on the multiple preset peak thresholds, wherein the heterogeneity ratios correspond one-to-one with the preset peak thresholds.

[0105] The second determining module 203 is used to determine a target outlier ratio from the plurality of outlier ratios, and to determine a preset peak threshold corresponding to the target outlier ratio as the target peak threshold, wherein the target outlier ratio is the smallest of the plurality of outlier ratios;

[0106] The processing module 204 is used to perform target processing on the gait dataset based on the target peak threshold to obtain a target dataset, wherein the target processing is used to correct the data in the gait dataset;

[0107] The third determining module 205 is used to determine the fatigue state information of the human body based on the target dataset.

[0108] Optionally, the first determining module 202 is further configured to:

[0109] Obtain the first peak threshold among the plurality of preset peak thresholds;

[0110] Based on the first peak threshold, multiple first peaks are determined in the gait dataset, and all of the first peaks are greater than the first peak threshold.

[0111] Based on the plurality of first peak values, a plurality of first cycle lengths are determined, wherein the first cycle length is the distance between any two adjacent first peak values;

[0112] Based on the plurality of first period lengths, statistical target parameters are calculated, including the following parameters: the frequency of each first period length and the frequency corresponding to the mode of the plurality of first period lengths;

[0113] Based on the frequency of each first period length and the frequency of the mode of the plurality of first period lengths, the heterogeneity ratio of the period length corresponding to the first peak threshold is determined.

[0114] Optionally, the processing module 204 is also used for:

[0115] Based on the target peak threshold, invalid peaks in the gait dataset are eliminated or hidden peaks are inserted into the gait dataset to obtain the target dataset.

[0116] Optionally, the processing module 204 is also used for:

[0117] Multiple second cycle lengths are determined based on the target peak threshold, where the second cycle length is the distance between two adjacent target peaks, and the target peak is the peak value in the gait dataset that is greater than the target peak threshold.

[0118] Calculate the average period length of the plurality of second period lengths;

[0119] In the gait dataset, a target data segment corresponding to the target cycle length is determined. The target cycle length is greater than a first value and less than a second value. The first value is the product of the difference between the preset value and the tolerance and the average cycle length. The second value is the product of the sum of the preset value and the tolerance and the average cycle length.

[0120] If the second cycle length on both sides of the target data segment in the gait dataset is less than the first value, the target peak corresponding to the second cycle length is eliminated; or, if the second cycle length on both sides of the target data segment in the gait dataset is greater than the second value, a hidden peak is added to the gait dataset to obtain the target dataset.

[0121] Optionally, the processing module 204 is also used for:

[0122] The target peak value corresponding to the second cycle length is determined as the third peak value;

[0123] A fourth peak is determined from the gait dataset. This fourth peak is located on the same side of the target data segment as the third peak, and is the farthest from the target data segment with a value greater than 0.

[0124] The period length between the third peak and the fourth peak is defined as the third period length;

[0125] Based on the third cycle length and the average cycle length, a third value is determined, which is the quotient of the third cycle length and the average cycle length.

[0126] If the length of the third period is greater than the product of the first value and the third value, and the length of the third period is less than the product of the second value and the third value, a hidden peak is inserted between the third peak and the fourth peak, wherein the number of the hidden peaks is the same as the third value.

[0127] Optionally, the device 200 further includes a selection module for:

[0128] Obtain the target peak threshold corresponding to each of the multiple gait datasets, and the variance of the cycle length corresponding to each gait dataset under the corresponding target peak threshold;

[0129] A reference gait dataset is determined from a plurality of said gait datasets, wherein the variance of the cycle length corresponding to the reference gait dataset is minimized under the corresponding target peak threshold;

[0130] Based on the target peak threshold corresponding to the reference gait dataset, invalid peaks in the reference gait dataset are eliminated or hidden peaks are inserted into the reference gait dataset to obtain the target dataset.

[0131] Optionally, the third determining module 205 is also used for:

[0132] The target dataset is input into a pre-trained fatigue state recognition model to identify the fatigue state of the human body, so that the fatigue state recognition model outputs the fatigue state result of the human body.

[0133] The fatigue state recognition model is a neural network model used to identify the fatigue state of the human body based on a dataset.

[0134] Optionally, the device 200 further includes a training module for:

[0135] Acquire information on human fatigue status and the corresponding dataset;

[0136] The dataset corresponding to the human fatigue state is used as the feature value, and the human fatigue state information is used as the label value to train the initial model.

[0137] If the initial model after training meets the preset conditions, the initial model after training will be determined as the fatigue state recognition model.

[0138] The human fatigue state information detection device 200 provided in this application embodiment can acquire gait dataset and multiple preset peak thresholds, thereby determining multiple heterogeneity ratios, and determining a target heterogeneity ratio among the multiple heterogeneity ratios, thereby determining a target peak threshold, and then processing the gait dataset according to the target peak threshold to obtain a target dataset, and finally determining the human fatigue state information according to the target dataset.

[0139] It should be noted that the human fatigue state information detection device 200 provided in this application embodiment can realize all the technical processes of the above-mentioned human fatigue state information detection method and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0140] The device in this application embodiment can be an electronic device or a component within an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other devices besides a terminal. For example, the electronic device can be a mobile phone, tablet computer, laptop computer, PDA, in-vehicle electronic device, mobile internet device (MID), augmented reality (AR) / virtual reality (VR) device, robot, wearable device, ultra-mobile personal computer (UMPC), netbook, or personal digital assistant (PDA), etc. Non-mobile electronic devices can also be servers, network attached storage (NAS), personal computers (PCs), televisions (TVs), ATMs, or self-service machines, etc. This application embodiment does not specifically limit the scope of the device.

[0141] Optionally, such as Figure 3 As shown, this application embodiment also provides an electronic device 300, including a processor 301 and a memory 302. The memory 302 stores a program or instructions that can run on the processor 301. When the program or instructions are executed by the processor 301, they implement the various steps of the above method embodiments and can achieve the same technical effect. To avoid repetition, they will not be described again here.

[0142] It should be noted that the electronic devices in the embodiments of this application include the mobile electronic devices and non-mobile electronic devices described above.

[0143] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above method embodiments and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0144] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0145] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0146] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0147] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for detecting human fatigue state information, characterized in that, The method includes: Acquire gait datasets and multiple preset peak thresholds; Based on the multiple preset peak thresholds, multiple heterogeneity ratios are determined in the gait dataset, and the heterogeneity ratios correspond one-to-one with the preset peak thresholds. A target heterogeneous ratio is determined from the plurality of heterogeneous ratios, and a preset peak threshold corresponding to the target heterogeneous ratio is determined as the target peak threshold, wherein the target heterogeneous ratio is the smallest of the plurality of heterogeneous ratios; Based on the target peak threshold, the gait dataset is subjected to target processing to obtain a target dataset. The target processing is used to correct the data in the gait dataset. Based on the target dataset, the fatigue state information of the human body is determined; The determination of multiple heterogeneity ratios in the gait dataset includes: Obtain the first peak threshold among the plurality of preset peak thresholds; Based on the first peak threshold, multiple first peaks are determined in the gait dataset, and all of the first peaks are greater than the first peak threshold. Based on the plurality of first peak values, a plurality of first cycle lengths are determined, wherein the first cycle length is the distance between any two adjacent first peak values; Based on the plurality of first period lengths, statistical target parameters are calculated, including the following parameters: the frequency of each first period length and the frequency corresponding to the mode of the plurality of first period lengths; Based on the frequency of each first period length and the frequency of the mode of the plurality of first period lengths, the heterogeneity-modality ratio of the period length corresponding to the first peak threshold is determined. The step of performing target processing on the gait dataset based on the target peak threshold to obtain the target dataset includes: Based on the target peak threshold, invalid peaks in the gait dataset are eliminated or hidden peaks are inserted into the gait dataset to obtain the target dataset.

2. The method as described in claim 1, characterized in that, The step of eliminating invalid peaks in the gait dataset or inserting hidden peaks in the gait dataset based on the target peak threshold to obtain the target dataset includes: Multiple second cycle lengths are determined based on the target peak threshold, where the second cycle length is the distance between two adjacent target peaks, and the target peak is the peak value in the gait dataset that is greater than the target peak threshold. Calculate the average period length of the plurality of second period lengths; In the gait dataset, a target data segment corresponding to the target cycle length is determined. The target cycle length is greater than a first value and less than a second value. The first value is the product of the difference between a preset value and the tolerance and the average cycle length. The second value is the product of the sum of the preset value and the tolerance and the average cycle length. If the second cycle length on both sides of the target data segment in the gait dataset is less than the first value, the target peak corresponding to the second cycle length is eliminated; or, if the second cycle length on both sides of the target data segment in the gait dataset is greater than the second value, a hidden peak is added to the gait dataset to obtain the target dataset.

3. The method as described in claim 2, characterized in that, When the second cycle length on both sides of the target data segment in the gait dataset is greater than the second value, a hidden peak is added to the gait dataset to obtain the target dataset, including: The target peak value corresponding to the second cycle length is determined as the third peak value; A fourth peak is determined from the gait dataset. This fourth peak is located on the same side of the target data segment as the third peak, and is the farthest from the target data segment with a value greater than 0. The period length between the third peak and the fourth peak is defined as the third period length; Based on the third cycle length and the average cycle length, a third value is determined, which is the quotient of the third cycle length and the average cycle length. If the length of the third period is greater than the product of the first value and the third value, and the length of the third period is less than the product of the second value and the third value, a hidden peak is inserted between the third peak and the fourth peak, wherein the number of the hidden peaks is the same as the third value.

4. The method according to any one of claims 1 to 3, characterized in that, The number of gait datasets is multiple. Before performing target processing on the gait datasets based on the target peak threshold to obtain the target dataset, the method further includes: Obtain the target peak threshold corresponding to each of the multiple gait datasets, and the variance of the cycle length corresponding to each gait dataset under the corresponding target peak threshold; A reference gait dataset is determined from a plurality of said gait datasets, wherein the variance of the cycle length corresponding to the reference gait dataset is minimized under the corresponding target peak threshold; The step of eliminating invalid peaks in the gait dataset or inserting hidden peaks in the gait dataset based on the target peak threshold to obtain the target dataset includes: Based on the target peak threshold corresponding to the reference gait dataset, invalid peaks in the reference gait dataset are eliminated or hidden peaks are inserted into the reference gait dataset to obtain the target dataset.

5. The method according to any one of claims 1 to 3, characterized in that, The step of determining the fatigue state information of the human body based on the target dataset includes: The target dataset is input into a pre-trained fatigue state recognition model to identify the fatigue state of the human body, so that the fatigue state recognition model outputs the fatigue state result of the human body. The fatigue state recognition model is a neural network model used to identify the fatigue state of the human body based on a dataset.

6. The method as described in claim 5, characterized in that, The method further includes inputting the target dataset into a pre-trained fatigue state recognition model to identify the fatigue state of the human body, and before the fatigue state recognition model outputs fatigue state information of the human body, the method further includes: Acquire information on human fatigue status and the corresponding dataset; The dataset corresponding to the human fatigue state is used as the feature value, and the human fatigue state information is used as the label value to train the initial model. If the initial model after training meets the preset conditions, the initial model after training will be determined as the fatigue state recognition model.

7. A device for detecting human fatigue state information, characterized in that, The device includes: The acquisition module is used to acquire gait datasets and multiple preset peak thresholds; The first determining module is used to determine multiple heterogeneity ratios in the gait dataset based on the multiple preset peak thresholds, wherein the heterogeneity ratios correspond one-to-one with the preset peak thresholds. The second determining module is used to determine a target outlier ratio from the plurality of outlier ratios, and to determine a preset peak threshold corresponding to the target outlier ratio as the target peak threshold, wherein the target outlier ratio is the smallest of the plurality of outlier ratios; The processing module is used to perform target processing on the gait dataset based on the target peak threshold to obtain a target dataset, wherein the target processing is used to correct the data in the gait dataset; The third determining module is used to determine the fatigue state information of the human body based on the target dataset; The first determining module is also used for: Obtain the first peak threshold among the plurality of preset peak thresholds; Based on the first peak threshold, multiple first peaks are determined in the gait dataset, and all of the first peaks are greater than the first peak threshold. Based on the plurality of first peak values, a plurality of first cycle lengths are determined, wherein the first cycle length is the distance between any two adjacent first peak values; Based on the plurality of first period lengths, statistical target parameters are calculated, including the following parameters: the frequency of each first period length and the frequency corresponding to the mode of the plurality of first period lengths; Based on the frequency of each first period length and the frequency of the mode of the plurality of first period lengths, the heterogeneity-modality ratio of the period length corresponding to the first peak threshold is determined. The processing module is also used for: Based on the target peak threshold, invalid peaks in the gait dataset are eliminated or hidden peaks are inserted into the gait dataset to obtain the target dataset.

8. An electronic device, characterized in that, The electronic device includes a processor and a memory, the memory storing programs or instructions that can run on the processor, the programs or instructions being executed by the processor to implement the steps of the method for detecting human fatigue state information as described in any one of claims 1 to 6.

9. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions, which, when executed by a processor, implement the steps of the method for detecting human fatigue state information as described in any one of claims 1 to 6.

Images