Health behavior monitoring system and method based on somatosensory interaction

Abstract

The invention discloses a health behavior monitoring system and method based on somatosensory interaction, and relates to the related field of somatosensory interaction, and the system comprises a starting module which is used for carrying out identity authentication after a user carries out starting interaction, and executing equipment starting after the identity authentication is passed; the task generation module is used for calling user data and configuring a reach task; the somatosensory interaction module is used for activating a tracking unit in the somatosensory interaction module to execute user authentication positioning and then carrying out reach monitoring on the user; the action decomposition module is used for performing action division on the reach monitoring data set through the calibration template and establishing action labels; and the health sensing module is used for initially configuring a health sensing channel based on the user portrait, performing multi-dimensional behavior health sensing according to the reach monitoring data set and the action label, and establishing a behavior health sensing result. The technical problem that existing health behavior monitoring is insufficient in action monitoring accuracy and health perception comprehensiveness is solved, and the technical effect that the health behavior monitoring result is more accurate and comprehensive is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of somatosensory interaction, in particular to a health behavior monitoring system and method based on somatosensory interaction. BACKGROUND

[0002] Health behavior monitoring plays a crucial role in helping people maintain good health and promptly discover potential health problems. Currently, traditional exercise monitoring devices such as wristbands, sports shoes, etc. are mainly used to monitor health behavior. These devices generally collect user exercise data and other information through built-in sensors. However, traditional monitoring methods are not accurate in monitoring action details, making it difficult to accurately classify and analyze complex actions such as touching, etc. On the other hand, such devices can only collect single-dimensional data, making it difficult to achieve multi-dimensional health perception and unable to comprehensively assess potential health problems that may exist in user health behavior.

[0003] In the current related technology, health behavior monitoring has the technical problems of insufficient action monitoring accuracy and comprehensive health perception. SUMMARY

[0004] The present application provides a health behavior monitoring system and method based on somatosensory interaction. When the user initiates interaction, the system first performs identity authentication, starts the device and calls user data to configure a personalized touch height task after authentication, then activates the somatosensory tracking module to locate the user, monitors the touch height action in real time and generates a data set, divides the four action stages of preparation, guide response, high point reach and recovery and labels them through the calibration template, and finally initializes the health perception channel combined with the user portrait, conducts multi-dimensional behavior health analysis based on action data and labels, and outputs the behavior health perception result. The technical means solves the technical problems of insufficient action monitoring accuracy and comprehensive health perception in existing health behavior monitoring, and achieves the technical effect of making the health behavior monitoring result more accurate and comprehensive.

[0005] The application provides a health behavior monitoring system based on somatosensory interaction, comprising: a starting module, configured to perform identity authentication on a user after the user performs a starting interaction, and perform device starting when the identity authentication is passed; a task generation module, configured to perform user data calling according to an identity authentication result, and configure a touch task according to a user data calling result, wherein the user data calling result comprises a user portrait and user historical touch data; a somatosensory interaction module, configured to perform touch monitoring of the user after a tracking unit in the somatosensory interaction module is activated after user authentication positioning is performed, and establish a touch monitoring data set; an action decomposition module, configured to perform action division of the touch monitoring data set by using a calibration template, and establish an action label, wherein the action label comprises labels of a preparation action, a guide response action, a high point reaching action and a recovery action; and a health perception module, configured to perform multi-dimensional behavior health perception based on the user portrait, the touch monitoring data set and the action label after a health perception channel is initially configured based on the user portrait, and establish a behavior health perception result.

[0006] In a possible implementation, in the health perception module, the multi-dimensional behavior health perception is performed based on the touch monitoring data set and the action label, and a behavior health perception result is established, and the following processing is performed: data segment division of the touch monitoring data set is performed according to the action label, and a preparation action sub-data segment, a guide response sub-data segment, a high point reaching sub-data segment and a recovery sub-data segment are established; a parameter set matched with the user portrait in the health perception channel is called, and a multi-dimensional behavior health perception path mapped with the sub-data segment is constructed; behavior characteristic perception of the mapped sub-data segment is performed by using the multi-dimensional behavior health perception path, and the behavior health perception result is established.

[0007] In a possible implementation, in the health perception module, the behavior characteristic perception of the mapped sub-data segment is performed by using the multi-dimensional behavior health perception path, and the behavior health perception result is established, and the following processing is performed: preparation behavior characteristic perception of the preparation action sub-data segment is performed by using a preparation action behavior health perception path, wherein the preparation behavior characteristic perception comprises start reaction time delay perception and trunk swing amplitude perception; reaction ability and balance stability analysis are performed by using the preparation behavior characteristic perception, and a first behavior health perception result is established; guide behavior characteristic perception of the guide response sub-data segment is performed by using a guide response behavior health perception path, wherein the guide behavior characteristic perception comprises target deviation distance perception and follow-up time delay perception; neural muscular control ability and attention maintenance analysis are performed by using the guide behavior characteristic perception, and a second behavior health perception result is established; and the behavior health perception result is established based on the first behavior health perception result and the second behavior health perception result.

[0008] In a possible implementation, in the health perception module, based on the first behavior health perception result and the second behavior health perception result, the following processing is performed: reach behavior characteristic perception of a high point reach sub-data segment is performed through a high point reach behavior health perception path, the reach behavior characteristic perception includes maximum height characteristic perception, upward speed characteristic perception, arm span amplitude characteristic perception; explosive force and upper limb joint activity analysis is performed according to the reach behavior characteristic perception, and a third behavior health perception result is established; recovery behavior characteristic perception of a recovery sub-data segment is performed through a recovery behavior health perception path, the recovery behavior characteristic perception includes landing stability perception, recovery time perception, and gravity center regression path perception; action integrity and recovery control ability analysis is performed according to the recovery behavior characteristic perception, and a fourth behavior health perception result is established; and a behavior health perception result is established based on the first behavior health perception result, the second behavior health perception result, the third behavior health perception result, and the fourth behavior health perception result.

[0009] In a possible implementation, in the health perception module, multi-dimensional behavior health perception is performed according to the touch height monitoring data set and the action label, and a behavior health perception result is established, and the following processing is further performed: after the touch height monitoring data set is time-identified, touch height frequency is counted, and a first additional behavior health perception result is established; action deformation perception is performed on the time-identified touch height monitoring data set, and a second additional behavior health perception result is established; and the behavior health perception result is compensated by using the first additional behavior health perception result and the second additional behavior health perception result.

[0010] In a possible implementation, the system further includes: a matching module, configured to perform abnormal health warning level matching according to the health perception result, and generate a matching result; and a warning reporting module, configured to configure a warning signal based on the matching result, and perform abnormal health behavior warning reporting.

[0011] In a possible implementation, the warning reporting module includes: a communication module, configured to establish device communication with a user wearing device; an interaction module, configured to establish a time sequence vital sign data set after performing wearing device interaction according to the device communication; and a compensation module, configured to configure a warning signal after compensating the matching result according to the time sequence vital sign data set.

[0012] In a possible implementation, the starting module further includes: if it is determined that the user is a new user, performing static data acquisition of the user, and establishing a first learning data set; obtaining user input data of the user, and establishing a second learning data set; creating a user portrait according to the first learning data set and the second learning data set, and performing device starting.

[0013] In a possible implementation, the system comprises: a record module configured to perform a high-jump task completion evaluation according to the high-jump monitoring data set, and generate an evaluation record; and an incentive module configured to match reward points according to the evaluation record, and issue the reward points to a corresponding user account.

[0014] The application also provides a health behavior monitoring method based on somatosensory interaction, comprising: when a user performs a start interaction, performing identity authentication on the user, and performing device start when the identity authentication is passed; performing user data calling according to the identity authentication result, and configuring a high-jump task according to the user data calling result, wherein the user data calling result comprises a user portrait and user historical high-jump data; after the high-jump task configuration is completed, activating a tracking unit in a somatosensory interaction module to perform user authentication positioning, and then performing high-jump monitoring of the user to establish a high-jump monitoring data set; performing action division of the high-jump monitoring data set through a calibration template to establish action labels, wherein the action labels comprise labels of a preparation action, a guide response action, a high-point reaching action, and a recovery action; after initially configuring a health perception channel based on the user portrait, performing multi-dimensional behavior health perception according to the high-jump monitoring data set and the action labels to establish a behavior health perception result.

[0015] The application provides a health behavior monitoring system and method based on somatosensory interaction. When a user performs a start interaction, the system performs identity authentication on the user through a start module, performs device start when the identity authentication is passed, performs user data calling according to an identity authentication result through a task generation module, configures a high-jump task according to a user data calling result, wherein the user data calling result comprises a user portrait and user historical high-jump data, performs high-jump monitoring of the user after activating a tracking unit in a somatosensory interaction module to perform user authentication positioning, establishes a high-jump monitoring data set, an action division module performs action division of the high-jump monitoring data set through a calibration template to establish action labels, wherein the action labels comprise labels of a preparation action, a guide response action, a high-point reaching action, and a recovery action, initially configures a health perception channel based on the user portrait, and performs multi-dimensional behavior health perception according to the high-jump monitoring data set and the action labels through a health perception module to establish a behavior health perception result. The health behavior monitoring result is more accurate and comprehensive. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings of the embodiments of the present application will be briefly introduced below. In the present application, flowcharts are used to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the foregoing or the following operations are not necessarily performed in sequence. On the contrary, various steps can be processed in reverse order or simultaneously according to the needs. Meanwhile, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0017] Figure 1 A structural schematic diagram of a health behavior monitoring system based on somatosensory interaction provided by the embodiments of the present application.

[0018] Figure 2 A flowchart of a health behavior monitoring method based on somatosensory interaction provided by the embodiments of the present application.

[0019] Legend: start module 10, task generation module 20, somatosensory interaction module 30, action decomposition module 40, health perception module 50. DETAILED DESCRIPTION

[0020] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the embodiments of the present application can be implemented according to the content of the description, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described.

[0021] In order to make the purposes, technical solutions and advantages of the present application more clear, the following will further describe the present application in combination with the drawings, and the described embodiments should not be regarded as limiting the present application. All other embodiments obtained by those skilled in the art without making creative efforts belong to the scope of protection of the present application.

[0022] In the following description, "some embodiments" are described, which describe a subset of all possible embodiments, but it can be understood that "some embodiments" can be the same subset or different subset of all possible embodiments, and can be combined with each other without conflict. The term "first\second" is only to distinguish similar objects, and does not represent a specific order of the objects. The terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or server including a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or modules not clearly listed or inherent to these processes, methods, products or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as understood by those skilled in the art in the technical field of the present application. The terms used herein are only for the purpose of describing the embodiments of the present application.

[0023] Embodiments of the present application provide a health behavior monitoring system based on somatosensory interaction, as shown in the accompanying drawings Figure 1 The system comprises: A starting module 10 is configured to perform identity authentication on a user after the user performs a starting interaction, and perform device starting when the identity authentication is passed.

[0024] Specifically, identity authentication refers to verifying the legitimacy of a user's identity through specific technical means to ensure that only authorized users can use the device or system. Identity authentication can use biometric technology such as fingerprint recognition or facial recognition. The device is equipped with a fingerprint sensor or a camera, and when the user performs a starting interaction (such as pressing a fingerprint recognition area or facing the camera), the system completes identity authentication by comparing with the user's biometric information previously recorded. For example, a capacitive fingerprint sensor is used, which works on the principle of using the capacitance of the human body and the difference in the electric field formed by the fingerprint lines to identify the fingerprint pattern. For facial recognition, an infrared camera can be used to capture facial feature points to construct a 3D facial model for comparison. When the identity authentication is passed, the starting circuit is triggered to switch the device from standby state to working state. At the same time, the starting program starts to run, initializing the hardware and software systems of the device. For example, the starting program loads the operating system kernel of the device, initializes various sensors and interfaces, and prepares for the work of subsequent modules.

[0025] In one possible implementation, the starting module 10 further comprises: if it is determined that the user is a new user, performing static data collection of the user to establish a first learning data set; obtaining user input data of the user to establish a second learning data set; creating a user portrait according to the first learning data set and the second learning data set, and performing device starting.

[0026] Specifically, the user is distinguished as a new or old user, for example, when the user performs a start interaction, the device reads the current device fingerprint and compares it with the previously stored device fingerprint. If there is no matching device fingerprint, the user is determined to be a new user. When the user is determined to be a new user, the start module 10 starts a series of static data collection processes. For example, through the camera, body sensor and other hardware on the device, basic physical data such as height, weight, body proportion of the user are collected. At the same time, the touch screen or voice input module on the device will guide the user to input some basic information such as gender, age, basic health status (whether there is a chronic disease, history of allergy, etc.). The collected static data is sorted and classified to form a first learning data set. These data are used to provide basic physical information and background data for the user's subsequent health monitoring and personalized task configuration. The device will guide the user to input his own exercise habits, dietary preferences, health goals (such as wanting to improve exercise ability, enhance physical fitness, lose weight, etc.) through the touch screen or voice input, etc. For example, the user can choose his favorite sports type (such as basketball, football, swimming, etc.), the length of time available for exercise per week, whether he likes sweets, etc. The user input data is summarized and sorted to form a second learning data set, which reflects the user's subjective will and preference.

[0027] Based on the first learning data set and the second learning data set, the data is analyzed and processed to create a user portrait. The user portrait contains multi-dimensional information such as the user's physical characteristics, health status, exercise habits, dietary preferences, and health goals. Then, the device performs the start according to the user portrait, and prepares for subsequent health behavior monitoring and personalized task configuration.

[0028] This implementation can create a unique user portrait for each user by collecting static data of new users and obtaining user input data, so that the device can provide personalized health monitoring and task configuration for the user according to his personal characteristics, health status, exercise habits and dietary preferences, etc.

[0029] The task generation module 20 is configured to call user data according to the identity authentication result, and configure a high jump task according to a user data calling result, the user data calling result including a user portrait and user historical high jump data.

[0030] Specifically, the user data is stored in the cloud database, including user profile (such as basic information such as age, gender, height, weight, and characteristic information such as sports preference, health status) and user history data (such as the number of times of climbing, the highest height, the average height, and the like). Through network communication technology (such as Wi-Fi or Bluetooth), the device is connected to the cloud server to retrieve the corresponding user data. A rule-based expert system or a machine learning algorithm is used. Among them, the rule engine generates a climbing task according to the user profile and the user history data of climbing, according to the set rules (such as arranging a simple climbing task for a user with a younger age and poor sports foundation, and gradually increasing the difficulty as the age and sports ability improve); the machine learning algorithm predicts and configures a climbing task suitable for the user by learning a large amount of user data.

[0031] The somatosensory interaction module 30 is used to activate the tracking unit in the somatosensory interaction module to perform user authentication positioning after the climbing task configuration is completed, and to perform climbing monitoring of the user to establish a climbing monitoring data set.

[0032] Specifically, the user is positioned using an infrared sensor, and a plurality of infrared transmitting and receiving devices are arranged around the device to determine the position of the user by triangulation. For example, infrared emitters are arranged at the four corners of the device, and the user wears an infrared reflective marker. After the receiver receives the reflected signal, the time difference of the signal arrival is calculated to determine the coordinate position of the user in space. A depth camera, such as a structured light depth camera or a ToF (Time of Flight) depth camera, is used to capture the user's actions. The depth camera can construct a 3D model of the user's body, monitor the user's body movements in real time, capture the details of the user's climbing action, and establish a climbing monitoring data set. The climbing monitoring data set is a collection of various data collected during the user's climbing action, including time series data, spatial coordinate data, and posture data, which are used for action analysis and health perception. For example, the structured light depth camera projects a specific light pattern, analyzes the deformation of the pattern to calculate the depth information, and obtains the depth coordinates of each part of the user's body, and then constructs the action data.

[0033] The action decomposition module 40 is used to divide the action of the climbing monitoring data set by calibrating the template, and to establish an action label, which includes the labels of the preparatory action, the guide response action, the high point reaching action, and the recovery action.

[0034] Specifically, the calibration template is a pre-defined standard high-reach action model, including a preparation action (such as a standing posture), a guide response action (such as a guide action of raising the arms upward), a high-point reaching action (such as an action of reaching the highest point with the fingers), and a recovery action (such as an action of recovering from the highest point to the initial posture). The dynamic time warping (DTW) algorithm or the hidden Markov model (HMM) algorithm is used to divide the action of the high-reach monitoring data set. For example, the DTW algorithm compares the continuous action data with the calibration template by calculating the optimal matching path between the time series data, determines the start and end points of the action, and establishes an action label. The action label is a classification and annotation of the action, which divides the continuous complex action into several sub-actions with clear characteristics and meanings, and is used for more detailed analysis and processing of the action.

[0035] The health perception module 50 is configured to perform multi-dimensional behavior health perception based on the high-reach monitoring data set and the action label after the health perception channel is initially configured based on the user portrait, and to establish a behavior health perception result.

[0036] Specifically, the health perception channel refers to a special data collection and analysis path set for different health dimensions (such as sports health, nutrition health, and skeletal health), which is used to comprehensively and deeply perceive the health status of the user. Different health perception channels are configured according to the user portrait. For example, for a child user, a sports health perception channel (such as monitoring the exercise intensity, exercise duration, and exercise frequency) and a nutrition health perception channel (such as analyzing the matching degree of food nutrients and exercise consumption) are mainly configured. For a teenager user, a skeletal development health perception channel (such as analyzing the stress condition of the skeleton through action monitoring) is added on this basis.

[0037] The deep learning algorithm (such as the convolutional neural network CNN or the long short-term memory network LSTM) is used to input the high-reach monitoring data set and the action label, and to combine other health-related data (such as heart rate and sleep data) to perform multi-dimensional behavior health perception. The CNN can extract the spatial features of the action data, and the LSTM can capture the time sequence features of the action, and finally establish a behavior health perception result. For example, by analyzing the speed, acceleration, and amplitude of the high-reach action, and combining the heart rate variation, the sports health status of the user is evaluated.

[0038] In a possible implementation, in the health perception module 50, multi-dimensional behavior health perception is performed according to the touch height monitoring data set and the action label, and a behavior health perception result is established, including: performing data segment division on the touch height monitoring data set according to the action label, establishing a preliminary action sub-data segment, a guide response sub-data segment, a high point reaching sub-data segment, and a recovery sub-data segment; calling a parameter set matched with a user portrait in a health perception channel, and constructing a multi-dimensional behavior health perception path mapped with the sub-data segment; and performing behavior characteristic perception on the mapped sub-data segment by using the multi-dimensional behavior health perception path, and establishing a behavior health perception result.

[0039] Specifically, according to the timestamp or data sequence position information labeled by the action label, the preliminary action sub-data segment, the guide response sub-data segment, the high point reaching sub-data segment, and the recovery sub-data segment are extracted from the complete touch height monitoring data set. For example, if the action label indicates that the preliminary action starts at the first second and ends at the third second of the data set, all related data (such as sensor data of displacement, speed, acceleration, etc.) in the first second to the third second of the touch height monitoring data set are extracted as the preliminary action sub-data segment, and other sub-data segments are extracted in the same way.

[0040] A plurality of parameter sets matched with different user portraits are pre-stored in the health perception channel. When it is necessary to call, the corresponding parameter set is searched and called according to the characteristics (such as age, gender, sports ability level, etc.) of the user portrait. For example, for a user portrait of a young sports enthusiast, the parameter set can include evaluation indexes of young sports intensity (such as maximum reaching height per unit time, action frequency, etc.), and judgment standards of action specification (such as correct angle range of each part of the body during the preliminary action, muscle exertion mode during the high point reaching, etc.).

[0041] For each sub-data segment, a series of analysis dimensions and corresponding algorithms or models are determined to form a perception path. Taking the preliminary action sub-data segment as an example, the perception path can include a “posture specification evaluation” dimension, and the corresponding algorithm is a human body posture estimation model (such as a human key point detection model in deep learning) based on analysis of body posture data in the sub-data segment to judge whether the preliminary posture is correct; a “muscle warm-up condition analysis” dimension, and the corresponding algorithm is to analyze the change of acceleration data and muscle electrical signals (acquired by muscle electrical sensors) to evaluate whether the muscle activity degree reaches the warm-up state. Each sub-data segment has a plurality of perception paths of different dimensions corresponding thereto.

[0042] The constructed multi-dimensional behavioral health perception path is used to map the behavioral characteristics of sub-data segments. Specifically, data in each sub-data segment is calculated and analyzed according to the algorithm or model for each dimension. For example, in the posture standardization assessment dimension, human posture data from the preparatory movement sub-data segment is input into a human posture estimation model. The model outputs angle information for various body parts, which is then compared with the preset standard angle range in the parameter set to determine whether the posture is standard. In the muscle warm-up analysis dimension, features are extracted and analyzed from acceleration data and electromyographic signals. Based on preset muscle warm-up assessment standards, a score for the degree of muscle warm-up is obtained. By synthesizing the analysis results from each dimension, a behavioral health perception result is finally established, forming a detailed report containing the user's health behavior performance at different stages of movement.

[0043] This approach, through detailed data segmentation and targeted perception path construction, enables refined analysis of the user's movement details and health status at each stage of the vertical jump process. This allows for the timely detection of potential health problems and movement defects, providing users with sufficient basis for improvement and technical support. It helps users exercise more scientifically, improve exercise results, prevent sports injuries, and achieve the goal of health management.

[0044] In one possible implementation, the health perception module 50 utilizes the multi-dimensional behavioral health perception path to perceive the behavioral characteristics of the mapped sub-data segments and establish behavioral health perception results. This includes: perceiving the preparatory behavioral characteristics of the preparatory action sub-data segments through the preparatory action behavioral health perception path, wherein the preparatory behavioral characteristics perception includes initial reaction delay perception and trunk sway amplitude perception; using the preparatory behavioral characteristics perception to perform reaction ability and balance stability analysis to establish a first behavioral health perception result; perceiving the guiding behavioral characteristics of the guiding response sub-data segments through the guided response behavioral health perception path, wherein the guiding behavioral characteristics perception includes target deviation distance perception and following delay perception; using the guiding behavioral characteristics perception to perform neuromuscular control ability and attention maintenance analysis to establish a second behavioral health perception result; and establishing a behavioral health perception result based on the first behavioral health perception result and the second behavioral health perception result.

[0045] Specifically, from the preparatory action sub-data segment, the time point (denoted as T0) when the action instruction is issued and the time point (denoted as T1) when the user's body starts to have a significant action response are determined, and the initial reaction time delay is T1-T0. For example, the action instruction is issued through device screen display or voice prompt, and the sensor (such as an acceleration sensor or a depth camera) of the device detects the time point when the user's body (such as the leg or arm) starts to move. Assuming that T0 is the time when the action instruction is issued, and the time when the acceleration sensor of the device detects that the acceleration of a part of the user's body first exceeds a set static threshold is T1, the time difference between the two can be calculated to obtain the initial reaction time delay. The calculated initial reaction time delay is compared with the average initial reaction time delay standard value of the same age group and the same type of movement. If the user's initial reaction time delay is significantly longer than the standard value, it indicates that the user's reaction ability may have a problem; if it is close to or shorter than the standard value, it indicates that the user's reaction ability is good.

[0046] The motion trajectory of the user's torso is captured in real time within the preparatory action sub-data segment using the depth camera or the inertial measurement unit (IMU) sensor of the device. By analyzing the displacement changes of the torso in the front-back, left-right, and vertical directions, the torso swing amplitude is calculated. For example, the depth camera can construct a 3D model of the user's torso, and by calculating the position coordinate differences of the model at different time points, the swing amplitude of the torso in each direction is obtained. Or the IMU sensor can measure the angular velocity and acceleration of the torso, and by performing mathematical processing such as integration on these data, the swing angle and displacement amplitude of the torso are obtained. According to the user's age, physical condition, and movement type, a reasonable torso swing amplitude range is set. If the user's torso swing amplitude exceeds this range, it indicates that the user's body balance stability is poor in the preparatory action stage; if it is within the range, it indicates that the user's balance stability is good.

[0047] In the guided response sub-data segment, the device pre-sets the target position of the guided action (such as a virtual target point displayed on the screen or a direction specified through voice prompt). The actual action trajectory of the user is tracked in real time through the depth camera or position sensor, and the deviation distance between the actual action trajectory of the user and the target position is calculated. For example, in a task that requires the user's arm to guide to touch a specific target point on the screen, the depth camera can accurately capture the position coordinates of the user's finger, and compare them with the coordinates of the target point to obtain the target deviation distance of each guided action. A range of allowed target deviation distance is set, and the average deviation distance and the standard deviation of the deviation are calculated based on the target deviation distance data of the user's multiple guided actions. If the average deviation distance is large and the standard deviation is also large, it indicates that the user's action accuracy is poor and the neuromuscular control ability needs to be improved; on the contrary, if the average deviation distance is small and the standard deviation is small, it indicates that the user's neuromuscular control ability is good.

[0048] For the case that the user needs to follow a specific rhythm or guide track in the guide response action, record the time point when the guide stimulus is emitted (e.g. the time point when the guide light flashes or the guide sound is played, denoted as T2) and the time point when the user starts to follow the action (denoted as T3). The follow-up delay is T3-T2. For example, in a guide action task of following the rhythm of jumping, the device plays a guide sound as the rhythm signal, and by detecting the starting time point of the user's body action (e.g. leg jumping or arm swinging), the follow-up delay is calculated. Similar to the starting reaction delay, the follow-up delay is compared with the standard follow-up delay of the same type of task. The standard follow-up delay is pre-set according to factors such as the user's exercise experience and skill level. If the user's follow-up delay is too long, it indicates that the user's attention allocation or neural system response coordination has a problem; if it is within a reasonable range, it indicates that the user's attention maintenance and neuromuscular coordination are good.

[0049] After the perception of the behavior characteristics of the preparatory action and the guide response action, the first behavior health perception result (reaction ability, balance stability analysis result) and the second behavior health perception result (neuromuscular control ability, attention maintenance analysis result) are integrated. A weighted average method can be used, and according to the importance of each analysis result to the overall health behavior, a corresponding weight is given. For example, the reaction ability weight is 0.3, the balance stability weight is 0.3, the neuromuscular control ability weight is 0.3, and the attention maintenance weight is 0.1 (the weight distribution can be adjusted according to actual application requirements), multiply the scores of each analysis index by the weight and add them up to get a comprehensive behavior health perception score. The comprehensive behavior health perception score and the analysis results of each sub-item are presented to the user or health management professional in the form of charts or written reports. For example, a behavior health radar chart is generated, each axis represents reaction ability, balance stability, neuromuscular control ability and attention maintenance, and the user can intuitively see the gap between his performance in each aspect and the health standard, as well as the overall behavior health status.

[0050] In a possible implementation, in the health perception module 50, the behavior health perception result is established based on the first behavior health perception result and the second behavior health perception result, including: performing touch behavior characteristic perception of a high point touch sub-data segment through a high point touch behavior health perception path, the touch behavior characteristic perception including maximum height characteristic perception, upward speed characteristic perception, arm span amplitude characteristic perception; performing explosive force and upper limb joint activity analysis according to the touch behavior characteristic perception to establish a third behavior health perception result; performing recovery behavior characteristic perception of a recovery sub-data segment through a recovery behavior health perception path, the recovery behavior characteristic perception including landing stability perception, recovery time perception, and gravity center regression path perception; performing action integrity and recovery control ability analysis according to the recovery behavior characteristic perception to establish a fourth behavior health perception result; and establishing the behavior health perception result based on the first behavior health perception result, the second behavior health perception result, the third behavior health perception result, and the fourth behavior health perception result.

[0051] Specifically, the maximum height reached by the user in the high touch sub-data segment is measured by using the depth camera or infrared ranging sensor of the device. For example, the depth camera can track the highest position coordinates of the user's fingers or touch objects (such as target points on the touch screen) in real time by constructing a 3D model of the user's body. Assuming that the starting position of the user is the ground, the highest point coordinates of the user's fingers in the air are captured by the depth camera, and the actual maximum height value is obtained through coordinate conversion (according to the focal length, image sensor size, and other parameters of the camera). The measured maximum height value is compared with the standard high touch height data under the conditions of the same age, the same gender, and the same height. The standard high touch height data can be obtained through large-scale sports tests. If the actual maximum height of the user is lower than a certain percentage of the standard value (such as lower than 80%), it indicates that the user has problems in explosive force or body flexibility; if it is close to or higher than the standard value, it indicates that the user has good explosive force and body flexibility.

[0052] Through the acceleration sensor of the device, the speed change of the user's body (such as the arm or leg) during the upward process in the high point reaching sub-data segment is recorded. The acceleration sensor can measure the acceleration data of the user in real time, and the speed data can be obtained by integrating the acceleration data. For example, during the process of the user's arm touching the target point upward, the acceleration sensor records the acceleration value of the arm at certain time intervals (such as 0.01 seconds), and then calculates the speed value at each time point using numerical integration method. The maximum speed value and the average speed value during the upward process are calculated. According to the user's sports project and physical quality and other factors, a reasonable speed range is set. If the user's upward speed is lower than the range, it indicates that the user's muscle strength output efficiency is lower and the explosive power is insufficient; on the contrary, if the speed is within the reasonable range or higher, it indicates that the user's explosive power is better.

[0053] Using a depth camera or joint position sensor, the stretching angle and amplitude of the user's arm during the touching process in the high point reaching sub-data segment are captured. The depth camera can determine the positions of the shoulder joint, elbow joint and wrist joint by detecting the key points of the human body skeleton. For example, by calculating the distance from the shoulder joint to the wrist joint and the included angle between the arm and the body trunk, the arm stretching amplitude is obtained. Or the joint position sensor directly measures the flexion and extension angles of the elbow joint and shoulder joint to determine the stretching of the arm. According to the standards of ergonomics and sports physiology, a reasonable arm stretching amplitude range is determined. If the user's arm stretching amplitude is less than the range, it may affect the highest height of the touch, reflecting the limited range of motion of the upper limb joints; if it is within the reasonable range, it indicates that the range of motion of the user's upper limb joints is good, which is conducive to exerting better explosive power and touch effect.

[0054] Using the pressure sensor of the device (installed in the landing platform of the device or the smart insole worn by the user) and the inertial measurement unit (IMU) sensor, the body stability of the user when landing after completing the touch height action is detected in the recovery sub-data segment. The pressure sensor can measure the pressure distribution when the user's feet land, and the IMU sensor can measure the acceleration and angular velocity of the body. For example, the pressure sensor can detect whether the pressure is evenly distributed when the feet land, and the IMU sensor can detect the shaking degree of the body at the moment of landing (such as the acceleration change of the body center of gravity and the rotation angle change). The landing stability is judged by analyzing the pressure distribution data and the body shaking data. If the pressure is not evenly distributed when the feet land, and the body shaking amplitude is large (such as the body center of gravity acceleration exceeds a certain threshold, and the body rotation angle exceeds a certain range), it indicates that the landing stability is poor; on the contrary, if the pressure is evenly distributed and the body shaking amplitude is small, it indicates that the landing stability is good.

[0055] In the recovery sub-data segment, the time from the user completing the high point reaching action to the user's body returning to the initial posture of relative stability (such as standing with both feet on the ground, body upright) is recorded. Changes in the user's body posture can be detected by the depth camera or attitude sensor of the device. For example, the depth camera captures the user's body posture in real time, and when the speed and acceleration of each part of the user's body are detected to be lower than the set static threshold, it is considered that the user has returned to the initial posture, and the time difference from the completion of the high point reaching to this time is the recovery time. According to factors such as the user's exercise intensity and physical condition, a reasonable recovery time range is set. If the user's recovery time is too long, it indicates that the user's muscle fatigue is high, or the control ability in the action recovery phase is insufficient; if the recovery time is within the reasonable range, it indicates that the user's body control ability and recovery ability are good.

[0056] Using the depth camera and IMU sensor of the device, the regression path of the user's body center of gravity is tracked in real time in the recovery sub-data segment. The depth camera can construct a 3D model of the user's body, and calculate the position coordinate changes of the body center of gravity through the model. The IMU sensor can measure the acceleration and angular velocity of each part of the body, and combined with the human body kinematics model, the moving track of the body center of gravity is calculated. For example, in the process of the user falling from the high point and recovering to the standing posture, the depth camera and IMU sensor work together to accurately record the moving path of the body center of gravity in the front-back, left-right and vertical directions. According to the standard of sports biomechanics, a reasonable center of gravity regression path range is set. If the user's center of gravity regression path deviates from the ideal path too much (such as the body center of gravity appears obvious left-right swing or front-back collision during the regression process), it indicates that the user's action coordination and control ability in the recovery phase have problems; if the center of gravity regression path is relatively stable and close to the ideal path, it indicates that the user performs well in action integrity and recovery control ability.

[0057] The first behavioral health perception result (reaction ability, balance stability analysis result), the second behavioral health perception result (neuromuscular control ability, attention maintenance analysis result), the third behavioral health perception result (explosive force, upper limb joint activity analysis result), and the fourth behavioral health perception result (action integrity and recovery control ability analysis result) are comprehensively analyzed. A multi-dimensional data fusion method can be used, for example, a behavioral health perception evaluation model is constructed, each sub-result is taken as an input feature, and a comprehensive calculation is performed through weighted summation and other mathematical methods. For example, the reaction ability is assigned a weight of 0.2, the balance stability is assigned a weight of 0.2, the neuromuscular control ability is assigned a weight of 0.2, the attention maintenance is assigned a weight of 0.1, the explosive force is assigned a weight of 0.1, the upper limb joint activity is assigned a weight of 0.1, the action integrity is assigned a weight of 0.05, and the recovery control ability is assigned a weight of 0.05 (the weights can be adjusted according to actual application requirements and expert experience), the scores of each sub-result are multiplied by the weights and then added, to obtain a comprehensive behavioral health perception score. The behavioral health perception result is presented to the user in an intuitive graphical interface, for example, a comprehensive health dashboard is generated, which includes detailed scores and trend curves of each sub-item. The user can see his performance in different health behavior dimensions, such as the specific scores and change trends of reaction ability, balance stability, explosive force, etc., and the overall behavioral health status. At the same time, the system can provide text suggestions, according to the comprehensive analysis result, pointing out the user's advantages and aspects that need to be improved, and formulating a personalized health improvement plan for the user.

[0058] This implementation mode comprehensively evaluates the user's physical function performance in the exercise process through a multi-dimensional and multi-stage analysis method, and provides a comprehensive health behavior evaluation for the user.

[0059] In one possible implementation mode, in the health perception module 50, the multi-dimensional behavioral health perception is performed according to the touch height monitoring data set and the action label, and the behavioral health perception result is established, further comprising: after the touch height monitoring data set is time-identified, the touch height frequency is counted, and a first additional behavioral health perception result is established; the time-identified touch height monitoring data set is subjected to action deformation perception, and a second additional behavioral health perception result is established; and the first additional behavioral health perception result and the second additional behavioral health perception result are used to compensate the behavioral health perception result.

[0060] Specifically, after the body-sensing interaction module 30 completes the touch height monitoring, a touch height monitoring dataset with time stamps is obtained. For example, the dataset contains the start and end times of each touch height action of the user. By analyzing the time information in the dataset, the touch height actions are counted. Time series analysis algorithms can be used to identify the time interval of each complete touch height action, and when the time interval is greater than a set threshold (such as 2 seconds), it is considered a new touch height action. For example, in a 1-minute touch height monitoring dataset, 30 touch height actions are identified through time identification, i.e., the touch height frequency is 30 times / minute. The statistical touch height frequency is compared with the standard touch height frequency range of the same type of exercise. The standard range can be determined according to the user's age, physical fitness, and other factors. For example, for a young user, the standard touch height frequency range is 20-40 times / minute. If the user's actual touch height frequency is lower than the range, it indicates that the user's exercise ability and endurance need to be improved; if it is within the range, it indicates that the user's touch height exercise frequency is reasonable. This comparison result is used as the first additional behavior health perception result.

[0061] The touch height monitoring dataset with time stamps is analyzed using a deep learning algorithm (such as a convolutional neural network). Based on the 3D model data of the user's actions captured by the depth camera, the convolutional neural network is trained to recognize action deformation features. For example, during the training phase, a large number of 3D model data labeled with normal actions and deformed actions (such as abnormal arm bending angle, insufficient leg stretching, etc.) are used as training samples. In the actual perception process, the user's touch height monitoring data with time stamps is input into the trained model, and the model outputs the deformation degree and type of the user's action. According to the deformation degree and type output by the model, the quality of the user's action is evaluated. For example, if the model detects that the user's arm bending angle during the touch height process is less than 30 degrees smaller than the standard action, and similar deformation occurs multiple times, it is considered that the user has obvious action deformation problems, which may affect the exercise effect and increase the risk of injury. This evaluation result is used as the second additional behavior health perception result.

[0062] On the basis of the existing behavioral health perception results (for example, analysis results based on the preparation, guidance, touch, and recovery stages), the first and second additional behavioral health perception results are combined for compensation. For example, a weighted fusion method can be used, with the weight of the original behavioral health perception result set to 0.7, the weight of the first additional result set to 0.15, and the weight of the second additional result set to 0.15 (the weights can be adjusted according to actual applications and expert experience). The final behavioral health perception result after compensation is obtained through weighted calculation. Among them, the first additional behavioral health perception result supplements the evaluation of the user's movement ability from the perspective of movement frequency, and the second additional behavioral health perception result supplements the evaluation of the user's movement safety and effect from the perspective of action quality. The two additional results can make up for the shortcomings of the original analysis method in these aspects, making the final behavioral health perception result more comprehensive and accurate.

[0063] This implementation can effectively reduce the error of the behavioral health perception result by introducing two additional perception dimensions, namely high-frequency touch and action deformation perception. For example, analysis based only on the action stage may ignore the differences in movement frequency of the user and the influence of action deformation on the movement effect and safety. Through the compensation of the additional results, the user's healthy behavior can be more accurately evaluated, such as accurately identifying those who have acceptable single-action quality but low movement frequency or frequent action deformation, resulting in poor overall movement effect.

[0064] In one possible implementation, the system further includes a matching module for matching the abnormal health warning level according to the health perception result, generating a matching result; and a warning reporting module for configuring a warning signal based on the matching result and performing abnormal health behavior warning reporting.

[0065] Specifically, the matching module is built-in with a set of warning level rule library, which defines the abnormal health warning level corresponding to different health perception results. For example, the warning levels are divided into slight abnormality (first-level warning), moderate abnormality (second-level warning), and severe abnormality (third-level warning). After the health perception module outputs the behavioral health perception result, the matching module reads the result and compares it with the rules in the warning level rule library. For example, if the behavioral health perception result shows that the user's explosive power is 30% lower than the standard value and there are serious problems with action integrity, the matching module determines that it belongs to severe abnormality (third-level warning) according to the rules in the rule library, and records the matching result. The matching result includes the user's identification information, the specific description of the abnormal health behavior, and the corresponding warning level. For example, for the user with insufficient explosive power and poor action integrity, the matching result is recorded as "User ID: XXX, Abnormal Health Behavior: Insufficient Explosive Power and Poor Action Integrity, Warning Level: Third-Level Warning".

[0066] The pre-warning reporting module configures corresponding pre-warning signals according to the matching results generated by the matching module. The pre-warning signals can adopt various ways, such as visual signals (screen display warning information, indicator light flickering), auditory signals (alarm sound) or tactile signals (vibration feedback). For example, for a three-level pre-warning, a red flashing indicator light and a high-pitched alarm sound are configured, and detailed abnormal health behavior description and suggested measures are displayed on the device screen; for a two-level pre-warning, a yellow indicator light and a medium-pitched alarm sound are configured, and brief abnormal information is displayed on the screen; for a one-level pre-warning, a blue indicator light and a low-pitched alarm sound are configured, and only a prompt information is displayed on the screen. The pre-warning reporting module issues a pre-warning to the user according to the configured pre-warning signal mode. For example, after the user completes the high-reach action, the device immediately starts the corresponding pre-warning signal according to the matching result.

[0067] This implementation can timely pre-warn the abnormal health behavior of the user through the cooperative work of the matching module and the pre-warning reporting module, which helps the user to be reminded in the early stage of the problem and to take measures in time, such as adjusting the exercise intensity and correcting the action posture, thereby effectively preventing potential health risks and avoiding more serious health problems caused by neglecting abnormal health behavior.

[0068] In a possible implementation, the pre-warning reporting module comprises: a communication module for establishing device communication with the user's wearable device; an interaction module for establishing a time-series vital sign data set after interacting with the wearable device according to the device communication; a compensation module for configuring a pre-warning signal after compensating the matching result according to the time-series vital sign data set.

[0069] Specifically, the communication module in the pre-warning reporting module establishes a connection with the user's wearable device (such as a smart bracelet, a smart watch, a heart rate monitor, etc.) by using Bluetooth, Wi-Fi or other wireless communication protocols. For example, when the device starts, the communication module automatically scans the available wearable devices around, the user selects the wearable device to be connected on the device, and inputs the corresponding pairing information (such as device name, password, etc.), and after completing the pairing, a device communication link is established. The communication module is responsible for the sending and receiving of data, ensuring the stability and security of data transmission. For example, through the Bluetooth protocol, instructions and data are sent to the wearable device at a certain data transmission rate (such as 1 Mbps), while receiving the data feedback from the wearable device. After the communication module establishes communication with the wearable device, the basic information of the wearable device can be obtained, such as device type, battery capacity, signal strength, etc. At the same time, the subsequent data interaction is prepared to ensure that the pre-warning reporting module can timely obtain the vital sign data of the wearable device.

[0070] The interaction module interacts with the wearable device through the communication module. For example, it sends a data request instruction to the wearable device, asking it to send physiological data such as heart rate, blood pressure, and blood oxygen saturation at a set time interval (e.g., every second). After receiving the request, the wearable device sends the real-time monitoring data back to the early warning module through the communication module. The interaction module organizes and stores the received data, establishing a time-series physiological data set. For example, it arranges heart rate data in chronological order to form a heart rate time-series data set, with each data point containing a timestamp and corresponding heart rate value. The time-series physiological data set contains the user's physiological changes over a period of time. The interaction module can perform preliminary processing on the data set, such as data cleaning (removing outliers and noise) and data fusion (integrating different types of data). For example, by analyzing the time series of heart rate and blood pressure data, it can identify the user's physiological change trend during exercise.

[0071] The compensation module analyzes the time-series physiological data set in combination with the matching results generated by the matching module. For example, if the matching results indicate that the user has a moderate abnormal health behavior (level two warning), but the time-series physiological data set shows that the user's heart rate and blood pressure are within the normal range, and the blood oxygen saturation is high, indicating that the user's physiological state is relatively stable. The compensation module compensates for the matching results based on these physiological data, such as reducing the warning level from level two to level one. The compensation algorithm can use a rule-based method, such as defining a series of compensation rules: "When heart rate is within the normal range and blood pressure is normal, reduce the warning level by one." It can also use a machine learning method to learn the relationship between physiological data and health behavior through model training, and automatically make compensation decisions. After compensation, the compensation module configures the warning signal according to the new warning level. For example, for level one warning, configure a blue indicator light and a low-pitched alarm sound, and display a brief prompt on the device screen: "The user's health behavior has a slight abnormality, please adjust the exercise method."

[0072] After the compensation module completes the warning signal configuration, it sends the configuration information to other parts of the early warning module (such as visual, auditory, and tactile signal generators), ensuring that the warning signal can be accurately reported to the user in the specified manner.

[0073] This implementation method obtains the user's time-series physiological data set through communication and interaction with the wearable device, and can compensate the matching results by integrating physiological information. This makes the warning decision not only based on the behavior health perception results, but also considers the user's real-time physiological state, avoiding the possible warning errors caused by relying solely on behavior data. For example, the user may have a slight behavioral abnormality, but the physiological signs show that his physical condition is good. Through the compensation mechanism, unnecessary over-warning can be avoided, improving the accuracy of the warning decision.

[0074] In a possible implementation, the system comprises: a record module configured to perform a high-jump task completion evaluation according to the high-jump monitoring data set, and generate an evaluation record; and an incentive module configured to match reward points according to the evaluation record, and issue the reward points to a corresponding user account.

[0075] Specifically, the record module acquires the complete high-jump monitoring data set output by the somatosensory interaction module 30, including the number of high jumps of the user, the height of each high jump, the time of action completion, and the like. Meanwhile, in combination with the action label information, it is determined whether the user has completed the high jump task according to the standard action procedure. For example, by analyzing the action label, it is determined whether the user has completed the four stages of the preparatory action, the response action, the high point reaching action, and the recovery action. According to the set evaluation rules, if the number of completions reaches a preset target (for example, 10 times) and the height of each high jump is not lower than a minimum standard (for example, 200 cm), it is determined that the task completion quality is high. The record module arranges these evaluation results into the evaluation record, including the basic information of the user (such as the user ID), the specific indicators of task completion (such as the actual number of completions, the average height, and the like), and the comprehensive evaluation grade (such as excellent, good, qualified, and unqualified).

[0076] The incentive module has a reward point rule library built in, and matches the corresponding reward points according to the evaluation record. For example, for a user with an excellent comprehensive evaluation grade, 100 points are rewarded; for a good user, 70 points are rewarded; for a qualified user, 30 points are rewarded; and for an unqualified user, no points are rewarded. The point rule library can also include other factors, such as the number of consecutive task completions, to increase the flexibility of the incentive. The incentive module issues the matched reward points to the corresponding user account through an interface associated with the user account. For example, the point data is sent to the App account or the cloud account of the user through network communication, and the point balance of the account is updated at the same time. The user can view the details of his or her points in the account, such as the number of points obtained this time and the cumulative points.

[0077] This implementation can effectively encourage users to actively participate in the high jump task through the evaluation of the record module and the reward point mechanism of the incentive module. In order to obtain higher reward points, the user will have more motivation to complete the task and improve the task completion quality, thereby increasing the stickiness and use frequency of the user to the system.

[0078] The embodiment of the application adopts the technical means that when a user initiates interaction, the system first performs identity authentication, starts the device and calls user data to configure a personalized touch height task after the authentication passes, then activates a somatosensory tracking module to locate the user, monitors touch height actions in real time and generates a data set, divides four action stages of preparation, guide response, high point reach and recovery through a calibration template and labels them, and finally initializes a health perception channel in combination with a user portrait, performs multi-dimensional behavior health analysis based on action data and labels, and outputs behavior health perception results, to solve the technical problems of insufficient action monitoring accuracy and health perception comprehensiveness of existing health behavior monitoring, and achieve the technical effects of more accurate and comprehensive health behavior monitoring results.

[0079] In the foregoing, reference is made to Figure 1 The health behavior monitoring system based on somatosensory interaction according to the embodiment of the application is described in detail. Next, the health behavior monitoring method based on somatosensory interaction according to the embodiment of the application will be described with reference to Figure 2 The health behavior monitoring method based on somatosensory interaction according to the embodiment of the application is described in detail. Next, the health behavior monitoring method based on somatosensory interaction according to the embodiment of the application will be described with reference to

[0080] The health behavior monitoring method based on somatosensory interaction according to the embodiment of the application is used to solve the technical problems of insufficient action monitoring accuracy and health perception comprehensiveness of existing health behavior monitoring, and achieve the technical effects of more accurate and comprehensive health behavior monitoring results.

[0081] The health behavior monitoring method based on somatosensory interaction comprises: when a user initiates interaction, performing identity authentication on the user, and executing device start when the identity authentication passes; performing user data calling according to the identity authentication result, configuring a touch height task according to the user data calling result, wherein the user data calling result comprises a user portrait and user historical touch height data; after the touch height task configuration is completed, activating a tracking unit in a somatosensory interaction module to perform user authentication positioning, then performing touch height monitoring of the user, and establishing a touch height monitoring data set; performing action division of the touch height monitoring data set through a calibration template, establishing action labels, wherein the action labels comprise labels of preparation action, guide response action, high point reach action and recovery action; after a health perception channel is initially configured based on the user portrait, performing multi-dimensional behavior health perception according to the touch height monitoring data set and the action labels, and establishing behavior health perception results.

[0082] The multi-dimensional behavior health perception result can be further established based on the first behavior health perception result and the second behavior health perception result.

[0083] The multi-dimensional behavior health perception result can be further established based on the first behavior health perception result and the second behavior health perception result.

[0084] The multi-dimensional behavior health perception result can be further established based on the first behavior health perception result and the second behavior health perception result.

[0085] The multi-dimensional behavior health perception is performed according to the touch height monitoring data set and the action label, and a behavior health perception result is established, and the method can further include: after the touch height monitoring data set is time-identified, the touch height frequency is counted, a first additional behavior health perception result is established, the time-identified touch height monitoring data set is subjected to action deformation perception, a second additional behavior health perception result is established, and the first additional behavior health perception result and the second additional behavior health perception result are used for compensating the behavior health perception result.

[0086] The method can further include: according to the health perception result, an abnormal health warning level is matched to generate a matching result; and based on the matching result, a warning signal is configured, and an abnormal health behavior warning is reported.

[0087] The method can further include: based on the matching result, a warning signal is configured, which can further include: device communication with a user wearing device is established; after the wearing device is interacted according to the device communication, a time-series vital sign data set is established; and after the matching result is compensated according to the time-series vital sign data set, the warning signal is configured.

[0088] The method can further include: when the user performs a start interaction, the user is subjected to identity authentication, and when the identity authentication is passed, the device is started, and the method can further include: if it is determined that the user is a new user, the static data of the user is collected to establish a first learning data set; user input data of the user is acquired to establish a second learning data set; and a user portrait is created according to the first learning data set and the second learning data set, and the device is started.

[0089] The method can further include: according to the touch height monitoring data set, a touch height task completion evaluation is performed to generate an evaluation record; and according to the evaluation record, a reward point is matched, and the reward point is issued to a corresponding user account.

[0090] The health behavior monitoring system based on somatosensory interaction provided in the embodiments of the application can execute the health behavior monitoring method based on somatosensory interaction provided in any embodiment of the application, has the corresponding function modules and beneficial effects of the execution method.

[0091] Although various references are made in the present application to certain modules in the system according to the embodiments of the application, however, any number of different modules can be used and run on the user terminal and / or server, and each unit and module included is only divided according to the functional logic, but is not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of each functional unit are only for the convenience of mutual differentiation, and do not limit the protection scope of the application.

[0092] The foregoing DETAILED DESCRIPTION, including the above section titled "Detailed Description," is not to be taken as limiting the scope of the application. Various modifications, combinations, and equivalents can be apparent to those skilled in the art and can be made once the nature of the application is understood. Any modification, combination, or equivalent, which falls within the principles and the scope of the present application, is intended to be included in the present application. In some instances, the actions or steps can be performed in different order from those described herein, and still achieve desirable results. Additionally, the process depicted in the figures can not necessarily require the particular order shown or sequential order to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous.

Claims

1. A health behavior monitoring system based on motion-sensing interaction, characterized in that, The system includes: The startup module is used to authenticate the user after the user performs a startup interaction, and to start the device after the authentication is successful. The task generation module is used to retrieve user data based on the identity authentication result and configure the height-reaching task based on the user data retrieval result. The user data retrieval result includes user profile and user historical height-reaching data. The motion-sensing interaction module is used to activate the tracking unit in the motion-sensing interaction module after the height-reaching task is configured, perform user authentication and positioning, monitor the user's height-reaching activity, and establish a height-reaching monitoring dataset. The action decomposition module is used to divide the action into the height monitoring dataset using a calibration template and establish action labels. The action labels include labels for preparatory actions, guiding response actions, high point reaching actions, and recovery actions. The health perception module is used to initially configure the health perception channel based on the user profile, and then perform multi-dimensional behavioral health perception based on the height monitoring dataset and the action tags to establish behavioral health perception results.

2. The health behavior monitoring system based on somatosensory interaction as described in claim 1, characterized in that, In the health perception module, multidimensional behavioral health perception is performed based on the vertical reach monitoring dataset and the action tags to establish behavioral health perception results, including: Based on the action tags, the data segmentation of the height monitoring dataset is carried out to establish a preparatory action sub-data segment, a guiding response sub-data segment, a high point reach sub-data segment, and a recovery sub-data segment. Call the parameter set in the health perception channel that matches the user profile to construct a multi-dimensional behavior health perception path that maps to the sub-data segments; The behavioral characteristics of the mapped sub-data segments are perceived using the multi-dimensional behavioral health perception path, and behavioral health perception results are established.

3. The health behavior monitoring system based on somatosensory interaction as described in claim 2, characterized in that, In the health perception module, the behavioral feature perception of the mapped sub-data segments is performed using the multi-dimensional behavioral health perception path to establish behavioral health perception results, including: The preparatory behavior characteristics of the preparatory action sub-data segment are perceived through the preparatory action behavior health perception path. The preparatory behavior characteristics perception includes the perception of the initial reaction delay and the perception of the trunk swing amplitude. Using the aforementioned pre-behavioral characteristics, reaction ability and balance stability are analyzed to establish the first behavioral health perception results. The guidance behavior characteristics of the guidance response sub-data segment are perceived through the guidance response behavior health perception path. The guidance behavior characteristics perception includes target deviation distance perception and follow-up delay perception. The neuromuscular control ability and attention maintenance are analyzed using the aforementioned guided behavioral feature perception to establish the second behavioral health perception results. A behavioral health perception result is established based on the first behavioral health perception result and the second behavioral health perception result.

4. The health behavior monitoring system based on somatosensory interaction as described in claim 3, characterized in that, In the health perception module, a behavioral health perception result is established based on the first behavioral health perception result and the second behavioral health perception result, including: The high-point reach behavior health perception path is used to perceive the reach behavior characteristics of the high-point reach sub-data segment. The reach behavior characteristic perception includes maximum height characteristic perception, upward speed characteristic perception, and arm span characteristic perception. Based on the perceived characteristics of the reach behavior, explosive force and upper limb joint range of motion are analyzed to establish the third behavior health perception results; The recovery behavior characteristics of the recovery sub-data segment are perceived through the recovery behavior health perception path. The recovery behavior characteristics perception includes landing stability perception, recovery time perception, and center of gravity return path perception. Based on the perceived characteristics of the recovery behavior, the integrity of the movement and the ability to recover control are analyzed to establish the fourth behavior health perception result; A behavioral health perception result is established based on the first behavioral health perception result, the second behavioral health perception result, the third behavioral health perception result, and the fourth behavioral health perception result.

5. The health behavior monitoring system based on somatosensory interaction as described in claim 1, characterized in that, The health perception module, which performs multidimensional behavioral health perception based on the vertical reach monitoring dataset and the action tags, and establishes behavioral health perception results, also includes: After time-stamping the height monitoring dataset, the frequency of height measurements is counted, and the first additional behavioral health perception result is established. Perform motion deformation perception on the height monitoring dataset after time stamping, and establish a second additional behavioral health perception result; Compensation for behavioral health perception results is performed using the first additional behavioral health perception result and the second additional behavioral health perception result.

6. The health behavior monitoring system based on somatosensory interaction as described in claim 1, characterized in that, The system also includes: The matching module is used to match abnormal health warning levels based on the health perception results and generate matching results; The early warning module is used to configure early warning signals based on the matching results and to issue early warnings for abnormal health behaviors.

7. The health behavior monitoring system based on somatosensory interaction as described in claim 6, characterized in that, The early warning dispatch module includes: The communication module is used to establish device communication with the user's wearable device; The interaction module is used to establish a time-series vital sign dataset after interacting with the wearable device based on the device communication. The compensation module is used to compensate the matching results based on the time-series vital sign dataset and then configure an early warning signal.

8. The health behavior monitoring system based on somatosensory interaction as described in claim 1, characterized in that, The startup module also includes: If a user is determined to be a new user, static data collection of the user is performed to establish the first learning dataset; Obtain user input data and build a second learning dataset; User profiles are created based on the first learning dataset and the second learning dataset, and device startup is performed.

9. The health behavior monitoring system based on somatosensory interaction as described in claim 1, characterized in that, The system includes: The recording module is used to evaluate the completion of the height-reaching task based on the height-reaching monitoring dataset and generate an evaluation record. The incentive module is used to match reward points based on the evaluation records and distribute the reward points to the corresponding user accounts.

10. A health behavior monitoring method based on somatosensory interaction, characterized in that, The method is implemented by the health behavior monitoring system based on somatosensory interaction as described in any one of claims 1-9, and the method includes: After the user initiates the startup interaction, the user's identity is authenticated. Once the authentication is successful, the device is started. User data is retrieved based on the identity authentication result, and a height-reaching task is configured based on the user data retrieval result. The user data retrieval result includes user profile and user historical height-reaching data. After the height measurement task is configured, the tracking unit in the motion-sensing interaction module is activated to perform user authentication and positioning, and then the user's height measurement is monitored to establish a height measurement monitoring dataset. The action classification of the height monitoring dataset is performed using a calibration template, and action labels are established. The action labels include labels for preparatory actions, guided response actions, high point reaching actions, and recovery actions. After initially configuring the health perception channel based on the user profile, multi-dimensional behavioral health perception is performed based on the height monitoring dataset and the action tags to establish behavioral health perception results.

Images