A method and system for constructing a sports information intelligent service platform

Abstract

This invention relates to the field of information platform construction technology, and more particularly to a method and system for constructing a sports information intelligent service platform. The method includes the following steps: collecting user motion image data and user basic health data; extracting key body node position information during the user's movement; generating a user movement trajectory curve based on the key node position information, and calculating the deviation angle and distance from a standard trajectory template to obtain movement correctness data; constructing a user fitness risk assessment model based on the movement correctness data and user basic health data, and analyzing the difference from a preset safety threshold to generate tiered health guidance data; decomposing the tiered health guidance data into quantifiable progress units to generate progress tracking data. This invention, by combining user motion image data and basic health data, constructs an assessment model capable of identifying individualized risk factors, achieving more accurate safety guidance.

Description

Technical Field

[0001] This invention relates to the field of information platform construction technology, and in particular to a method and system for constructing a sports information intelligent service platform. Background Technology

[0002] The sports information intelligent service platform is a comprehensive system based on big data, artificial intelligence, the Internet of Things (IoT), and cloud computing technologies. It aims to integrate and analyze sports-related data from multiple sources (such as athlete physiological data, event information, venue management data, and user behavior data), and provide personalized services, real-time analysis, and predictive support through intelligent algorithms and models. The platform collects athletes' physical data and training performance through IoT devices, combines computer vision technology to perform tactical and movement analysis on match videos, and utilizes cloud computing technology for the storage and computation of massive amounts of data to achieve functions such as real-time event broadcasting, tactical optimization, spectator services, and venue management.

[0003] However, traditional sports information intelligent service platforms often have the following problems: Traditional systems mainly rely on the coach's subjective judgment or simple sensor data for movement assessment, lacking the ability to accurately capture and analyze subtle deviations in the user's movements. Traditional fitness guidance systems often set general long-term goals for progress, lacking quantifiable and trackable short-term progress indicators, making it difficult for users to perceive gradual improvement. Summary of the Invention

[0004] Therefore, the present invention needs to provide a method and system for constructing a sports information intelligent service platform to solve at least one of the above-mentioned technical problems.

[0005] To achieve the above objectives, a method for constructing a sports information intelligent service platform includes the following steps:

[0006] Step S1: Collect user motion image data and user basic health data; extract key body node position information during the user's movement; generate user movement trajectory curve based on key node position information, and calculate the deviation angle and distance from the standard trajectory template to obtain movement correctness data;

[0007] Step S2: Construct a user fitness risk assessment model based on movement correctness data and user health baseline data, analyze the difference from the preset safety threshold, and generate tiered health guidance data;

[0008] Step S3: Decompose the tiered health guidance data into quantifiable progress units to generate progress tracking data; perform anonymous user matching based on similar fitness characteristics and progress patterns on the progress tracking data to form interactive incentive data; establish an initial user exercise pattern map based on the movement correctness data, tiered health guidance data, and progress tracking data.

[0009] Step S4: Calculate the stability index of the data collection environment and the consistency index of user feedback for each data point in the initial user motion pattern map, and select high-confidence data points to construct a user motion evaluation model; based on the user motion evaluation model and interactive incentive data, analyze the operation fluency and usage frequency of each functional module, and implement a low-threshold progressive function unlocking mechanism in combination with interactive incentive data.

[0010] The present invention also includes a system for constructing a sports information intelligent service platform, used to execute the above-described method for constructing a sports information intelligent service platform, wherein the system for constructing the sports information intelligent service platform includes:

[0011] The motion trajectory analysis module is used to collect user motion image data and user basic health data; extract key body node position information during the user's movement; generate user motion trajectory curve based on key node position information; and calculate the deviation angle and distance from the standard trajectory template to obtain action correctness data.

[0012] The fitness risk assessment module is used to build a user fitness risk assessment model based on movement correctness data and user health baseline data, analyze the difference from the preset safety threshold, and generate tiered health guidance data.

[0013] The progress tracking and interaction matching module is used to decompose the hierarchical health guidance data into quantifiable progress units to generate progress tracking data; to perform anonymous user matching based on similar fitness characteristics and progress patterns on the progress tracking data to form interaction incentive data; and to build an initial user exercise pattern map based on movement correctness data, hierarchical health guidance data, and progress tracking data.

[0014] The user motion assessment module is used to calculate the stability index of the data collection environment and the consistency index of user feedback for each data point in the initial user motion pattern map, and to screen high-confidence data points to build a user motion assessment model. Based on the user motion assessment model and interactive incentive data, the module analyzes the smoothness of operation and frequency of use of each functional module, and implements a low-threshold progressive function unlocking mechanism in combination with interactive incentive data.

[0015] This invention, centered on the core application needs of a sports information intelligent service platform, constructs a closed-loop mechanism encompassing multi-source data collection, user behavior analysis, ability assessment, and functional guidance. This mechanism achieves a refined understanding and dynamic adaptation of users' movement status, behavioral habits, and learning abilities, significantly enhancing the platform's comprehensive capabilities in personalized services, data reliability assurance, and user experience optimization. In the data processing stage, by integrating movement correctness data, hierarchical health guidance data, and progress tracking data, a unified, time-aligned multi-source data integration table is constructed. Furthermore, environmental parameters such as light intensity, background complexity, and equipment stability are used to hierarchically manage the collection quality, ensuring the stability and reliability of the data foundation for subsequent model training and analysis tasks. Simultaneously, the user technical characteristic profiles and learning characteristic data established by the platform not only reveal typical deviations in user movement execution and the speed of new knowledge acquisition but also reflect potential dependency patterns between technical behaviors through feature association networks. This supports the establishment of a sports ability development model and generates short- to long-term ability growth prediction results, providing a scientific basis for the generation of personalized training paths. In terms of evaluation mechanisms, the platform assesses the authenticity and representativeness of data from two dimensions: environmental interference factors and user feedback performance. It generates a data collection environment stability index and a user feedback consistency index, forming a comprehensive data confidence evaluation model through a weighted mechanism. This allows the platform to automatically identify high-confidence data points during data screening and modeling, constructing a more accurate and robust motion assessment model. After model training, the platform further mines the behavioral trajectory information formed during user interaction with the system. By recording parameters such as operation duration, access frequency, and page path, it identifies peak usage periods and functional preference trends. Combined with motion ability profiles, it dynamically constructs a mapping relationship between user functional needs, providing behavioral-level basis for function recommendations. Furthermore, by analyzing user operation paths, decision node preferences, and interactive response characteristics, the platform can accurately identify their exploration behavior patterns and learning paths, thereby proactively adjusting and personalized push notifications for function adaptation and content delivery, effectively enhancing user engagement and platform stickiness. The platform also measures the user's proficiency and mastery of each functional module by calculating the operation completion time, retry frequency, and help information call status, and then dynamically adjusts the user's functional access permissions. In terms of feature guidance, a low-threshold, progressive feature unlocking strategy is introduced. This allows users to gradually master basic functions and receive automatic unlocking prompts for new features based on their behavior. This well-paced advancement mechanism lowers the learning threshold for using features and prevents usage obstacles caused by information overload or operational complexity. Simultaneously, the system dynamically adjusts the scoring weights and time limits in the unlocking mechanism based on user incentive response data, such as score achievement rate and continuous task completion cycle. This makes the unlocking process more flexible and adaptable, maintaining a sense of challenge for users while enhancing their motivation for continuous learning.Through the systematic implementation of the above methods, the platform can not only achieve a multi-dimensional understanding of individual users' sports behavior, learning pace, and operational preferences, but also accurately construct training suggestions, function push paths, and evaluation systems based on user characteristics, thereby improving the intelligence level and usability of the platform services. This entire approach, based on data and guided by user behavior, achieves an organic integration between technical models and human-computer interaction, promoting the platform's transformation from a "passive recorder" to an "active guide," demonstrating extremely high practicality and promotional potential in sports intelligent service scenarios. Attached Figure Description

[0016] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0017] Figure 1 This is a flowchart illustrating the steps of constructing a sports information intelligent service platform according to the present invention.

[0018] Figure 2 for Figure 1 A detailed flowchart of step S1;

[0019] Figure 3 for Figure 1 A detailed flowchart of step S2. Detailed Implementation

[0020] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0021] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0022] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0023] To achieve the above objectives, please refer to Figures 1 to 3 This invention provides a method for constructing a sports information intelligent service platform, the method comprising the following steps:

[0024] Step S1: Collect user motion image data and user basic health data; extract key body node position information during the user's movement; generate user movement trajectory curve based on key node position information, and calculate the deviation angle and distance from the standard trajectory template to obtain movement correctness data;

[0025] In this embodiment of the invention, a sports information intelligent service platform first collects motion image data of users performing preset fitness movements (such as squats, push-ups, lunges, etc.) using a multi-angle camera system deployed in gyms or home settings. Simultaneously, it acquires the user's basic health data, including weight, height, BMI, resting heart rate, and past sports injury records. This health data can be obtained through smart bracelets, smart body fat scales, and completed health questionnaires. After collecting the motion image data, the platform extracts the three-dimensional coordinate information of key body nodes of the user using posture recognition algorithms (such as OpenPose or MediaPipe). These key nodes include the head, shoulders, elbows, wrists, hip joints, knees, and ankles. Next, a complete motion trajectory curve of the user is generated by a time series tracking algorithm, and then compared frame by frame with a standard motion trajectory template. The angular deviation (e.g., the difference between the user's knee bending angle and the standard angle) and spatial displacement (i.e., the Euclidean distance between the user node and the standard trajectory node) between each key node and the standard node are calculated to obtain motion correctness data. This data is recorded with a sampling time of 0.1 seconds per frame. The generated data fields include "node number, time point, angle deviation (degrees), displacement distance (cm)".

[0026] Step S2: Construct a user fitness risk assessment model based on movement correctness data and user health baseline data, analyze the difference from the preset safety threshold, and generate tiered health guidance data;

[0027] After extracting user movement accuracy data, this embodiment of the invention constructs a personalized fitness risk assessment model based on the user's basic health data (such as joint flexibility indicators, muscle strength levels, etc.). This model uses a gradient-enhanced tree (GBDT)-based regression analysis method to map variables such as angle deviation and displacement distance to the risk level that may lead to sports injury or movement error, outputting a risk score for each type of movement. For example, if a user's knee abduction angle deviation is consistently higher than 15 degrees and the corresponding BMI is greater than 28, the system assesses that the knee pressure is excessive when performing squats, and the risk of injury is higher than the safety threshold. The platform sets this score as "high risk level". Based on the risk assessment results and the platform's preset risk stratification standards (such as setting the safety threshold to level 3 out of 5), the platform generates stratified health guidance data. This data includes corrective suggestions for the user's movements (such as "knee abduction control" and "core strength training"), suggestions for assistive devices (such as using elastic bands to control the trajectory), and training intensity limits, ensuring that the training suggestions generated by the platform are personalized, scientific, and executable.

[0028] Step S3: Decompose the tiered health guidance data into quantifiable progress units to generate progress tracking data; perform anonymous user matching based on similar fitness characteristics and progress patterns on the progress tracking data to form interactive incentive data; establish an initial user exercise pattern map based on the movement correctness data, tiered health guidance data, and progress tracking data.

[0029] In this embodiment of the invention, after obtaining stratified health guidance data, the platform refines the guidance suggestions into quantifiable progress units, that is, quantifies fitness goals according to the key points of movement improvement and their achievability. For example, if a user has a severe problem with knee abduction control, the problem is broken down into progress units such as "5 minutes of resistance band half-squat training daily" and "target abduction angle controlled within 10 degrees," each unit having time requirements, movement standards, and target indicators. The platform generates progress tracking data by periodically evaluating the user's performance, which consists of fields such as "progress unit number, completion status, trajectory improvement degree, and training frequency." Based on this, the platform uses clustering algorithms (such as K-Means) to analyze the fitness characteristics (such as physical condition, movement defect type, improvement path) and progress patterns (such as improvement speed, training frequency, etc.) of different users, achieving anonymous user matching and forming interactive incentive data, such as pushing "a challenge list of users with similar progress speeds" or forming virtual training groups for users with similar problems. Meanwhile, the platform performs semantic annotation and graph structure mapping on users' action correctness data, health guidance data, and progress tracking data to construct an initial user movement pattern graph. Graph nodes represent the execution status of various actions, and edges represent the transformation paths between actions or user state transition logic, supporting subsequent personalized training route recommendations.

[0030] Step S4: Calculate the stability index of the data collection environment and the consistency index of user feedback for each data point in the initial user motion pattern map, and select high-confidence data points to construct a user motion evaluation model; based on the user motion evaluation model and interactive incentive data, analyze the operation fluency and usage frequency of each functional module, and implement a low-threshold progressive function unlocking mechanism in combination with interactive incentive data.

[0031] In this embodiment of the invention, after constructing the initial user motion pattern map, the platform filters each data node in the map using the acquisition environment stability index and the user feedback consistency index. The acquisition environment stability index evaluates whether the data node was acquired under the same ambient lighting, shooting angle, and background interference conditions. It is calculated based on the standard deviation of brightness and background disturbance indicators in each video frame. If the standard deviation exceeds a platform-set threshold (e.g., brightness standard deviation greater than 15 pixel units), the reliability of the node's data is reduced. The user feedback consistency index is calculated by the correlation between the training satisfaction score submitted by the user after each training session and the actual improvement in movement. If the consistency is high (e.g., Pearson correlation coefficient greater than 0.7), the node's feedback is considered reliable. The platform retains only data nodes with high scores in both indices to construct a high-confidence user motion evaluation model. This model is used to accurately evaluate the user's current motion state and predict future movement execution risks or ability change trends. Based on this evaluation model and the aforementioned interactive incentive data, the platform further analyzes the operational smoothness (such as action recognition response time and feedback generation time) and user usage frequency (such as the number of weeks a certain training suggestion is called) of each functional module. If a functional module has low incentive participation, low user usage frequency, and the evaluation model predicts that the module has little impact on user progress, the advanced version will not be released for the time being. Conversely, for modules with good incentive effects and strong user stickiness, a low-threshold progressive function unlocking mechanism is implemented. For example, advanced challenge action training videos are automatically unlocked after a user completes three consecutive improvements in action accuracy to improve the platform's user activity and positive training feedback.

[0032] Preferably, step S1 includes the following steps:

[0033] Step S11: Collect multiple frames of continuous images of the user performing fitness movements through the camera of the smart terminal device to generate user motion image data;

[0034] In this embodiment of the invention, within a sports information intelligent service platform, users utilize smart terminal devices equipped with high-resolution cameras (such as smart fitness mirrors, AI-enabled camera TVs, or mobile phones) for fitness training. After the user selects a specific movement (such as bent-over rows, squats, or lunges), the platform activates the camera module, continuously capturing the user's complete training process at a frame rate of 15 frames per second. The acquisition cycle is the duration required to complete the movement (for example, completing a squat takes approximately 5 seconds, generating 75 consecutive frames of video data). This continuous video data constitutes the original user motion image data. The platform numbers each frame and records a timestamp to ensure accurate positioning of movement details under time-varying conditions in subsequent analysis. To enhance image data quality, the system automatically detects lighting sufficiency and the integrity of the subject (e.g., whether there is limb occlusion or overexposure in the image), ensuring that only high-quality motion image sequences meeting processing standards are acquired as the input basis for subsequent posture recognition and trajectory analysis.

[0035] Step S12: Collect the user's height, weight, age, and exercise habits to form the user's basic health data;

[0036] Before capturing motion images, the platform prompts users to fill in basic health information via a smart terminal interface and synchronizes data from wearable devices. Users need to manually or through authorization from their health records to fill in their height (in centimeters), weight (in kilograms), age (in years), and exercise habits (including weekly exercise frequency, duration of each exercise session, and main training programs). Exercise habit information is collected through a structured questionnaire and converted into quantifiable training level indicators. For example, "training 3 times a week for 30 minutes each time, preferring bodyweight training" will be categorized as "moderate frequency, low intensity trainer." Furthermore, the platform supports automatic acquisition of user physiological parameters, such as body fat percentage and resting heart rate, via Bluetooth connection to devices like smart body fat scales and heart rate bracelets, enriching the dimensions of health data. The system encapsulates this information into a unified "User Basic Health Data" object, with fields including basic physical characteristics, exercise frequency level, and training preference tags, serving as the basis for subsequent personalized movement risk analysis and recommended training templates.

[0037] Step S13: Perform image preprocessing on the user motion image data and identify skeletal points to locate key body nodes of the user's shoulder, elbow, wrist, hip, knee, and ankle, and generate key body node position information;

[0038] After acquiring user motion image data, this embodiment of the invention first performs image preprocessing on each frame, mainly including image denoising, background matting, and human body region enhancement. Image denoising uses a Gaussian filter to smooth image edges and eliminate image particle interference that may occur during the shooting process; background matting uses a deep learning segmentation network (such as U^2-Net or BodyPix) to separate the user from the background, allowing the recognition algorithm to focus on the human body structure; human body region enhancement enhances the clarity of skeletal lines through brightness normalization and contrast improvement. Subsequently, the platform performs skeletal point recognition on the preprocessed image based on a lightweight pose estimation algorithm (such as BlazePose or HRNet), extracting and tracking the positions of key body nodes of the user. Key nodes include the shoulder (connecting the upper arm and torso), elbow (key point for forearm movement), wrist (hand control point), hip (core point affecting overall stability), knee (main force support point), and ankle (lower limb dynamic stability control point), etc. The position of each key node is represented by three-dimensional spatial coordinates, including lateral displacement, longitudinal displacement, and estimated depth value (in centimeters). The platform integrates the skeletal node position information at multiple moments into an array of "key body node position information" to construct dynamic motion trajectories.

[0039] Step S14: Generate the user's motion trajectory curve based on the key node location information, and calculate the deviation angle and distance from the standard trajectory template to obtain motion correctness data.

[0040] Based on the key node position information generated in step S13, this embodiment of the invention uses a motion trajectory reconstruction algorithm to model the three-dimensional motion path of each key node over time during the user's entire training process, forming a user motion trajectory curve. The specific processing flow is as follows: the system interpolates and smooths the spatial position of each key node in consecutive frames according to time sequence, connecting them to generate a continuous curved trajectory. Simultaneously, it comprehensively models the relative movement patterns between multiple nodes to ensure the accuracy of the overall motion coordination analysis. Next, the platform calls the corresponding motion template from the standard trajectory template library. The standard trajectory data in the template is collected and modeled by professional athletes under ideal conditions, possessing high reference value. The platform compares the user's trajectory with the standard trajectory point by point through frame synchronization and key point matching mechanisms, calculating the deviation angle (e.g., the difference between the knee joint bending angle and the standard angle) and offset distance (e.g., the perpendicular distance between the shoulder joint displacement direction and the standard curve) of key nodes in each frame. The units of deviation are angle (degrees) and distance (centimeters), respectively. For example, if the user's hip joint deviation angle is greater than 15 degrees and the ankle offset distance exceeds 10 centimeters when performing a lunge, the system will determine it as an abnormal posture. Ultimately, the platform aggregates the aforementioned angle and distance data into "action correctness data" and assigns an "action matching score" to each frame, serving as the core basis for subsequent action risk analysis and guidance suggestions.

[0041] Preferably, step S14 includes the following steps:

[0042] Step S141: Perform time-series analysis on the position information of key body nodes, record the position changes of each key node in continuous time frames, and generate raw data of node motion trajectory;

[0043] In this embodiment of the invention, after obtaining the array of key body node position information of the user, the platform activates the time-series analysis module to dynamically track the three-dimensional coordinate changes of each key node (such as shoulder, elbow, hip, knee, etc.) in continuous time frames (e.g., 15 frames per second, for a total of 75 frames sampled). The system constructs a time-series trajectory sequence for each node, with the frame number as the horizontal axis and the spatial coordinate as the vertical axis, to record its movement trajectory over time. For example, for the left knee node, the platform records its three-dimensional position changes from frame 1 to frame 75 as a three-dimensional vector sequence, forming the original node motion trajectory data. This process uses timestamp calibration to ensure that all node data strictly correspond to the same time step, while the system automatically fills in the missing position information caused by short-term occlusion to improve the continuity and completeness of the trajectory. Finally, the platform combines the time-series position data of all nodes to generate "original node motion trajectory data," which serves as the basis for subsequent curve construction and smoothing processing.

[0044] Step S142: Smooth the original data of the node motion trajectory to obtain the user motion trajectory curve data;

[0045] This invention applies a motion trajectory smoothing algorithm to the raw data of node motion trajectories to eliminate trajectory noise caused by user jitter, environmental interference, or recognition errors during movement. Specifically, it employs a moving average smoothing algorithm based on a sliding window. The window size is set to 5 frames, and the X, Y, and Z three-dimensional coordinate values ​​of each node are weighted and averaged. This ensures that the coordinate values ​​of each frame are the average of the adjacent frames, making the trajectory curve visually more coherent and mathematically more stable. For joint transitions (such as changes in hip and knee angles during a squat), the system introduces an edge-preserving filtering strategy (such as Savitzky-Golay filtering) to prevent the smoothing from weakening the characteristics of these transition points. The processed user motion trajectory curve data presents a continuous and smooth set of three-dimensional curves, more realistically reflecting the user's movement trends and suitable for subsequent dynamic analysis and trajectory comparison.

[0046] Step S143: Calculate the motion angle, velocity, and acceleration of the human motion chain segment formed between adjacent key nodes based on the user motion trajectory curve data, and construct a dynamic parameter set for the user motion chain segment;

[0047] This invention calculates the dynamic parameters of human motion chain segments formed between adjacent key nodes based on smoothed user motion trajectory curve data. A "motion chain segment" refers to a body structure segment formed by two adjacent key points (such as shoulder-elbow, hip-knee, knee-ankle), which rotates and moves over time during movement. The platform first constructs vectors between each pair of nodes, recording their directional changes in consecutive frames. The "motion angle" is obtained by calculating the degree of change in the angle between vectors in adjacent frames. Simultaneously, the system calculates the change in coordinate displacement of each motion chain segment vector between time frames, divided by the time interval, to obtain the "linear velocity." Further, the velocity change is time-differentiated to obtain "acceleration" data. All these parameters are combined into a "user motion chain segment dynamic parameter set," which accurately describes the motion characteristics of each segment of the user's body structure during training. For example, in a bent-over rowing motion, whether the motion angle of the elbow-shoulder chain segment reaches more than 45 degrees during the movement, and whether it is accompanied by excessively rapid swinging (excessive acceleration), can be reflected in this parameter set for subsequent motion recognition and accurate feedback.

[0048] Step S144: Match the user's dynamic parameter set of motion chain segments with the standard parameter set in the action library to determine the type of fitness action currently being performed by the user and generate action type recognition results;

[0049] This invention compares and matches the constructed user motion chain dynamic parameter set with the standard parameter set in the platform's motion standard library to identify the type of fitness movement the user is performing. The platform pre-establishes standard parameter models for common fitness movements (such as squats, deadlifts, and bent-over rows). Each model includes angle ranges, speed rhythms, and acceleration variation characteristics of multiple typical motion chain segments. For example, the "standard squat" model defines that during the squatting phase, the hip-knee chain must form an angle closure greater than 90 degrees within 1 second, and the ankle speed must not exceed 0.5 m / s. The platform uses feature vector similarity algorithms (such as cosine similarity or dynamic time warping) to match the user's dynamic parameter set with each standard model one by one, identifying the closest movement type. The system outputs "Movement Type Recognition Result," which includes the identified movement name, matching confidence score (between 0 and 1), and reference movement number, providing a basis for subsequent standard trajectory extraction and alignment operations.

[0050] Step S145: Extract the corresponding standard trajectory template data from the action standard library based on the action type recognition result, and perform spatiotemporal alignment between the user motion trajectory curve data and the standard trajectory template data to obtain normalized trajectory comparison data;

[0051] In this embodiment of the invention, based on the identified action type, the platform extracts standard trajectory template data for the corresponding action from the action standard library. This data consists of three-dimensional trajectory curves generated by professional athletes performing the action in a standard manner. The platform then performs a "spatiotemporal alignment" operation on the user's trajectory curve data and the standard trajectory to resolve differences between the two trajectories in terms of start time, action rhythm, and scale units. First, the Dynamic Time Warping (DTW) algorithm automatically aligns the user and the standard trajectory on the time axis, ensuring that the key nodes in the process from start to finish of the two actions match as closely as possible. Second, the platform normalizes the spatial scale by converting the two trajectories into a standard unit coordinate system based on height percentage, eliminating the influence of absolute coordinates on the trajectory caused by differences in body shape, and obtaining "normalized trajectory comparison data" for standard deviation calculation and error analysis.

[0052] Step S146: Calculate the Euclidean distance deviation between the user trajectory and the standard trajectory at each key time point in the normalized trajectory comparison data, generate a point deviation sequence, and calculate the angle deviation between the user motion chain segment and the standard motion chain segment at each key time point, generate an angle deviation sequence.

[0053] This invention, based on normalized trajectory comparison data, sequentially compares the user's and standard action trajectories at several "key time points." Key time points refer to the moments in the action where posture changes are most significant and have the greatest impact on the action's effect (such as the lowest point of a squat, the turning point upon standing, etc.). The platform predefines the number and location of key time points for each action (e.g., a complete squat contains 5 key points). At each key point, the platform calculates the "Euclidean distance deviation" of the corresponding node, which is the spatial difference (in centimeters) between the user's node position and the standard node position in the three-dimensional coordinate system; simultaneously, it calculates the "angle deviation" of the same chain segment at that time point, which is the angle difference (in degrees) between the user's chain segment vector and the standard chain segment vector. The system generates two sequences: a "positional deviation sequence" reflecting positional errors and an "angle deviation sequence" reflecting structural posture errors. These two sequences are used to comprehensively evaluate the accuracy of the user's actions.

[0054] Step S147: Calculate the overall motion trajectory deviation score based on the position deviation sequence and angle deviation sequence to obtain motion correctness data.

[0055] In this embodiment of the invention, the positional deviation sequence and angle deviation sequence are input into the motion scoring module. A weighted scoring algorithm is used to score the overall motion deviation, generating "motion correctness data." The platform sets deviation sensitivity weights according to the motion type; for example, it focuses more on hip and knee angle deviations for squatting motions, and more on shoulder and elbow angles and upper body posture deviations for rowing motions. The scoring algorithm assigns weights to the positional and angle deviations at each key time point and performs normalization calculations. For example, a deviation less than 5 cm or 5 degrees is considered full marks, less than 10 is passing, and more than 10 is failing. The platform ultimately outputs a "motion correctness score" (0-100 points) and marks the concentrated deviation segments and corresponding joints, facilitating subsequent user guidance feedback and motion correction suggestions. This entire process lays the foundation for the platform to provide intelligent training effect evaluation and safety monitoring capabilities.

[0056] Preferably, step S2 includes the following steps:

[0057] Step S21: Extract motion deviation characteristic indicators from motion correctness data to form motion risk basic assessment data, where motion deviation characteristic indicators include the maximum deviation value, average deviation value, and deviation duration of key nodes;

[0058] After generating the action correctness data, the platform system in this embodiment of the invention will further analyze the data and extract action deviation feature indicators to form basic action risk assessment data. In this process, firstly, the point deviation sequence and angle deviation sequence are read in time series form, and the deviation trajectory of each key body node within the complete action cycle is statistically analyzed. Based on this, three core indicators are extracted: first, the maximum deviation value of the key node, used to measure the most significant deviation in the user's action; second, the average deviation value, reflecting the prevalence of deviation throughout the entire action; and third, the deviation duration, i.e., the time period during which the deviation value exceeds a threshold, used to reflect the length of time the user's non-standard actions last. For example, during squat training, if the angle between the user's knee and hip joints consistently deviates from the standard trajectory by more than 10 degrees for more than 1.5 seconds, the system will record this deviation as a high-risk feature. These features will be statistically analyzed according to action category and node type and stored in the basic action risk database, providing basic input for subsequent risk modeling.

[0059] Step S22: Extract user physiological characteristic parameters from user basic health data to generate user physiological risk factor data;

[0060] This invention acquires users' basic health data through user-end uploads or API calls. This data typically includes information such as age, gender, weight, height, history of sports injuries, skeletal health status (e.g., presence of articular cartilage degeneration), heart rate range, and flexibility score. The system extracts physiological characteristic parameters from this data and generates user physiological risk factor data. During processing, the platform uses a screening model based on medical knowledge graphs to extract features from the health data and automatically selects corresponding highly relevant factor combinations based on the type of action the user is currently performing. For example, for users performing jumping actions, the system prioritizes extracting knee joint stability parameters and ankle joint flexibility parameters as physiological risk factors. Furthermore, the platform introduces data standardization processing, normalizing the risk factors of different users to a range of 0 to 1, facilitating weighted modeling in subsequent risk calculations.

[0061] Step S23: Establish a correlation model between the basic assessment data of movement risk and the user's physiological risk factor data, calculate the degree of injury risk to the user for various movement deviations, and generate individualized risk assessment values;

[0062] In this embodiment of the invention, after acquiring motion deviation characteristic indicators and user physiological risk factor data, the platform constructs a correlation model between the two to calculate the risk level of sports injury caused by deviation during a user's performance of a certain type of movement. This model employs a risk inference model based on a combination of logistic regression and Bayesian networks. During the training phase, it uses a large amount of historical sample data to label the correspondence between motion deviation and injury consequences. During the prediction phase, it takes the current user's motion deviation indicators and their physiological risk factors as input and outputs an individualized injury risk value for that movement. For example, when the system detects that a user is performing a lunge with a knee valgus angle deviation greater than 15 degrees for more than 2 seconds, and the user has a history of knee ligament injury, the platform infers through the model that the individualized risk value corresponding to this movement is 0.82 (ranging from 0 to 1), indicating a high-risk level.

[0063] Step S24: Based on the individualized risk assessment value and the preset safety threshold, risk level stratification is performed to generate risk classification data;

[0064] This invention employs a risk level stratification process based on calculated individualized risk assessment values ​​and platform-set risk safety thresholds. Specifically, the system divides risk assessment values ​​between 0 and 1 into several risk level segments, such as 0.0 to 0.3 for low risk, 0.3 to 0.6 for medium risk, and 0.6 to 1.0 for high risk, assigning each assessment value to its corresponding level to form risk grading data. This data includes not only basic information such as action type, user ID, and action occurrence time, but also risk level labels, risk values, and key characteristic parameters affecting the risk assessment results (e.g., which deviation indicator contributes most to the risk). For example, if the platform records a user's "squat" action assessment value as 0.72, the system classifies it as "high risk" and indicates that the main source of this risk is "high average knee joint angle deviation value and long duration."

[0065] Step S25: Perform differentiated processing on action deviations of different risk levels based on risk classification data, and record action deviation patterns and guidance responses to obtain historical user feedback data;

[0066] In this embodiment of the invention, after acquiring risk classification data, the platform performs differentiated processing on movement deviations based on the risk level. For low-risk deviations, the system can automatically ignore them or only provide corrective suggestions. For medium-risk deviations, the system will alert the user and provide corrective guidance. If the deviation is high-risk, the system will immediately issue a mandatory pause suggestion and initiate a virtual coach-assisted teaching process. Simultaneously, it records the feedback processing strategy for each movement deviation and the user's response to that processing, forming historical user feedback data. For example, if a user is identified as high-risk by the system for performing a high-bend movement three times consecutively, the system not only provides a voice prompt to pause but also records whether the user made adjustments after the prompt, whether the movement deviation was alleviated after the adjustments, and stores the processing results in a structured manner.

[0067] Step S26: Visualize health guidance content based on risk grading data and user historical feedback data to generate tiered health guidance data.

[0068] This invention integrates risk grading data and user historical feedback data to generate tiered health guidance content and present it visually. The platform uses a graphics engine-based visualization module to overlay deviation points during each movement onto the user's skeletal model as a color heatmap. Simultaneously, a data panel displays the corresponding risk level and suggested actions. For example, if a user repeatedly experiences excessive hip joint movement during high knee exercises, the platform automatically generates a risk curve for that movement and overlays text suggestions, such as "It is recommended to reduce leg height and strengthen core muscle stability." Furthermore, the platform categorizes guidance strategies for different risk levels, providing daily training suggestions for low-risk users, pushing corrective video training content for medium-risk users, and recommending professional rehabilitation courses for high-risk users. This achieves personalized services for tiered health guidance data and supports subsequent use of the intelligent exercise recommendation module.

[0069] Preferably, step S25 includes the following steps:

[0070] Step S251: Extract the motion chain segment angle deviation value, speed control deviation value and motion stability index from the low-risk level motion deviation in the risk classification data, and quantitatively compare them with the preset standard motion model to identify the optimizable parameters and obtain the low-risk optimizable parameters.

[0071] This invention further extracts optimizable motion parameters for motion deviations already categorized as low-risk. First, the system performs structural analysis on the raw motion data of this type of deviation, calculating the angular deviation values ​​between key joints segment by segment of the motion chain (e.g., shoulder-elbow-wrist chain, hip-knee-ankle chain). Combined with time series analysis, it derives speed control deviation values, i.e., the degree of difference between the user's speed fluctuation during a key motion segment and the speed curve of the standard model. Simultaneously, the platform introduces motion stability indicators for evaluation, including the smoothness of the motion trajectory (assessed through acceleration variance) and the range of center of gravity fluctuation (measured by the displacement amplitude of the user's center point). The system compares these three types of parameters one by one with the ideal parameters in the standard motion model, using threshold judgment and deviation ranking methods to identify motion chains and indicators that still have room for improvement within the low-risk range. For example, if a user's back contraction speed is relatively fast during rowing training, deviating from the standard model's average by 8%, but without causing significant instability, the system determines that this speed control indicator is a "low-risk optimizable parameter."

[0072] Step S252: Develop a progressive adjustment plan for low-risk optimizable parameters to generate general optimization recommendation data;

[0073] In this embodiment of the invention, after identifying low-risk, optimizable parameters, the platform automatically generates a progressive optimization scheme, forming general optimization suggestion data. This process is based on the principle of "small, incremental adjustments" in exercise learning theory, and plans the adjustment range in conjunction with the user's existing ability level. For example, for users with large deviations in speed control, the platform sets an optimization scheme with a daily target value improvement of no more than 5%, avoiding the user from developing new deviations due to excessive correction. At the same time, the system provides corresponding action demonstration videos and real-time voice prompts, such as prompting "Slow down the back stretching speed by 0.2 seconds in the next training set to maintain stable muscle tension." This optimization suggestion data is stored as a structured record, including the optimization target parameters, adjustment range, operation frequency, and a list of suitable training actions, and can be presented in a graphic and textual format on the user interface.

[0074] Step S253: Analyze the proximity of the motion deviation points to the human biomechanical safety limits for the medium-risk level motion deviations in the risk classification data, and calculate the potential stress concentration areas.

[0075] For moderate-risk movement deviations, this invention further analyzes the proximity of the deviation point to the human biomechanical safety limits. First, the system invokes a built-in human safety limit model. This model, combining the range of motion (ROM), tendon load limits, and maximum safe skeletal force values ​​defined in biomechanical research literature, maps the nodes of the user's deviation movement. Then, an inverse dynamics algorithm estimates the force path and load distribution of each key node under the deviation state, identifying potential "stress concentration areas." For example, in a bent-over press, if the shoulder flexion angle exceeds 90% of the physiological limit while the humeral internal rotation angle significantly deviates from the standard value, the system will determine that there is potential stress concentration in the acromion region and mark it as a moderate-risk warning point.

[0076] Step S254: Based on potential stress concentration areas and user physiological risk factor data, assess the risk of damage to the user's joints, muscles and ligaments caused by the deviation, and generate basic safety guidance data including descriptions of prohibited movements, definitions of safe movement ranges and recommendations for alternative movements;

[0077] In this embodiment of the invention, after identifying potential stress concentration areas, the platform further assesses the likelihood of damage to the user's joints, muscles, and ligaments caused by the movement deviation, based on the user's physiological risk factor data. The system uses mechanical simulation and individual parameter embedding models for cross-evaluation. If a user has a background of articular cartilage degeneration and the load on the shoulder rotator muscles is consistently high during a deviation movement, the system will label the deviation as a "high stress-physiological vulnerability" combined risk and automatically generate basic safety guidance data containing three parts: first, "description of prohibited movements," clearly indicating to the user that certain movements should be avoided; second, "definition of safe movement range," clearly showing the user's acceptable range of joint movement through illustration; and third, "recommendation of alternative movements," recommending movements with similar functional goals but lower biomechanical burden, such as replacing high-load presses with resistance band lateral raises. The above information is integrated to generate a personalized guidance document, which can be synchronized to the user's mobile terminal.

[0078] Step S255: Extract the degree and duration of the movement deviations exceeding the safety threshold from the high-risk level movement deviations in the risk classification data, and classify the errors based on the high-risk movement deviations to obtain error type data, which includes technical errors, physical limitation errors, or errors with potential health problems.

[0079] In this embodiment of the invention, for high-risk movement deviations, the platform needs to further extract two indicators: "the degree to which the deviation exceeds the safety threshold" and "the duration." The system first locates the abrupt change point of the deviation based on the time-series data of the user's entire movement process, and calculates the difference between the peak deviation value and the preset safety threshold. For example, if the hip joint adduction angle deviates from the standard posture by more than 25 degrees and is maintained for more than 3 seconds, exceeding the upper limit of the threshold by more than 50%, the system then calls a well-trained error recognition model (integrating expert rules and supervised learning methods) to classify and identify the high-risk movement deviation, determining that the error type belongs to one of the following three categories: First, "technical errors," such as errors in movement path control or muscle group coordination disorders; second, "physical limitation errors," such as insufficient flexibility or muscle strength leading to the inability to complete the movement; and third, "potential health problem errors," such as repeated deviations accompanied by abnormal physiological reactions (such as abnormal heart rate fluctuations), indicating that the user may have unidentified health problems. The system outputs error type data for subsequent personalized intervention.

[0080] Step S256: Generate professional consultation guidance data based on error type data;

[0081] This invention generates professional consultation guidance data based on the aforementioned error type data. For "technical errors," the system pushes a specialized corrective training plan designed by a sports coach and guides the user to book online video guidance courses. For "physical limitation errors," the platform pushes a low-intensity training plan based on physical fitness improvement and provides training progress management tools. For "potential health problem errors," the platform will forcibly trigger a health warning mechanism, advising the user to suspend related training, pushing consultation appointment portals for sports rehabilitation experts or sports medicine physicians, and guiding the user to a designated medical examination institution for specialized assessment if necessary. The aforementioned professional consultation guidance data is stored in the form of a structured data package, including error type markings, intervention suggestion types, recommended resource links, and tracking feedback interfaces, supporting cross-module interactive use.

[0082] Step S257: Record the user's behavioral response after receiving general optimization suggestion data, basic safety guidance data, or professional consultation guidance data, including the time delay in accepting guidance, the completion rate of implementing guidance, and the persistence in continuously applying guidance, and generate user historical feedback data.

[0083] This invention records and models the entire process of user behavior responses after receiving various types of guidance data (including general optimization suggestions, basic safety guidance, or professional consultation guidance). The platform obtains the following data through a user behavior monitoring interface: "time delay in receiving guidance" (i.e., the time taken for the user's first response after receiving a prompt), "completion rate of instruction execution" (analyzing whether adjustments were adequate by comparing subsequent action data with target guidance parameters), and "persistence in applying guidance" (statistically tracking whether the user consistently completes training according to the guidance within a certain period). For example, if a user receives shoulder joint safety guidance data and begins modifying their movements within 3 hours, resulting in a decrease in shoulder angle deviation of over 15%, and maintains the new posture for three consecutive days, the platform rates the user as "high response-high completion-high persistence," generates historical user feedback data, and uploads it to the user behavior modeling module for subsequent personalized content recommendations and risk control strategy adjustments.

[0084] Of particular importance, step S26 includes the following steps:

[0085] Analyze historical user feedback data to determine user response speed and completion rate for different presentation formats, and identify the most effective information delivery channels.

[0086] Based on historical user feedback data, assess users' understanding and preference for technical terms, and establish a user terminology acceptance model;

[0087] Based on the terminology acceptance model, adjust the frequency of use and explanation depth of professional terms in general optimization suggestion data, basic safety guidance data, or professional consultation guidance data to generate language complexity adaptation data.

[0088] Personalized guidance data is generated based on language complexity and information delivery channels.

[0089] A tiered health guidance data structure is constructed based on risk grading data and personalized guidance data.

[0090] This invention analyzes user response time and completion rate for different types of guidance content presentation formats (such as text descriptions, dynamic illustrations, voice broadcasts, video demonstrations, and combinations of text and images) based on user historical feedback data and behavioral tracking records. Response speed is defined as the time interval between a user's first attempt to perform a related action after receiving guidance content. The completion rate is calculated by comparing the suggested action parameters in the guidance content with the actual action data generated after the user's execution. For example, analyzing a user's training data over 30 consecutive days revealed that the user's average response time for voice broadcast guidance content was 12 minutes, with a completion rate of 82%; while the response time for text-based guidance content was as high as 50 minutes, with a completion rate of only 41%. The platform employs a multi-dimensional response scoring model, combining information format, response latency, and completion rate to construct a scoring matrix, ultimately outputting that the most suitable information delivery channel for the user is a combination of "voice broadcasts + dynamic illustrations." This result will serve as the basis for generating subsequent personalized guidance formats. Data tags related to technical terms are extracted from user feedback interaction logs and training behaviors. For example, user behavior is recorded after terms such as "scapular stability," "isometric contraction," or "hip flexor activation" appear, including whether they consult term explanations, whether they delay execution due to term confusion, and whether they actively request simplification. Simultaneously, the platform categorizes terms into three types: basic, intermediate, and advanced. Combining user response time, consultation frequency, and execution deviation rate, a term acceptance modeling method based on decision tree algorithms is used to assign acceptance weights to each term category. For example, if a user's average consultation time for the intermediate-advanced term "lower crossed syndrome" is 70 seconds, and the execution deviation of the guidance content is higher than 20%, the platform classifies this term as a "low acceptance term," with a corresponding weight of only 0.3 in the user's term acceptance model. The term acceptance model ultimately forms a personalized term acceptance threshold curve for each user, used to guide language complexity adaptation strategies. Based on the output of the term acceptance model, the frequency and level of terms used in the current guidance content are automatically retrieved and compared with the user's acceptance curve for filtering and replacement. Specifically, the platform uses a semantic substitution template library to replace low-acceptance terms with synonymous but more accessible expressions, such as replacing "winged scapula" with "scapular eversion." While retaining the terms, it adds annotations or pop-up explanation modules to clarify their meaning and training relevance. The platform uses a content complexity calculation model to evaluate the Flesch reading difficulty index of the text before and after adjustment, and saves both the adjusted and original versions as language complexity adaptation data, assigning them labels (such as "standard version," "simplified version," "illustrated enhanced version," etc.) for the platform to flexibly call upon according to user needs when outputting content. For example, for users with a term acceptance threshold of less than 0.5, a "simplified version + semantic annotation" content version is output. This data provides a language-level adaptation basis for subsequent personalized guidance content generation.Combining the user's optimal information delivery channel (such as voice broadcast + dynamic illustration) with the "simplified + semantically annotated" content version from the language complexity adaptation data, the system dynamically assembles the final output format using a content generation engine. The system generates multimodal content packages based on a combination of "carrier format + language style + interaction method," for example, outputting "short voice commands + interactive animations + simplified terminology explanations." The voice content is kept under 20 seconds, the animation annotates joint movement trajectories, and provides an interactive button for users to click and view terminology explanations. The platform assigns a unique ID to personalized guidance format data and binds it to the user ID, recording its display timing, usage frequency, and execution effect to continuously optimize the output strategy. For example, the guidance output for a sports rehabilitation user during post-operative training is: "Please slowly raise your arm outward to shoulder level while keeping your shoulder pressed against the wall (this is a simplified term for preventing scapular eversion). You can watch the animation demonstration on the right and listen to the voice prompts to complete the movement," achieving highly adaptive personalized communication. By integrating risk levels of movement deviations (low, medium, and high risk), personalized guidance formats, and language adaptation strategies, a guidance content template system is constructed to meet users' tiered health intervention needs. For low-risk deviations, the system matches general optimization suggestions, primarily using simplified terminology and illustrations. For medium-risk deviations, basic safety guidance data is matched, employing a medium-complexity format of illustrations, audio, and terminology annotations. For high-risk deviations, professional consultation guidance data is invoked, including original medical terminology, explanatory documents, and doctor appointment links. The platform dynamically adjusts the tiered strategy of the matched content based on users' historical execution records and terminology acceptance models. For example, if a user's acceptance of the term "rotator cuff tear risk" in the high-risk category is only 0.4, the platform will add an illustration and synchronized audio explanation of this term to the high-risk guidance data, and simultaneously send a doctor's guidance link via SMS, constructing a tiered health guidance data structure of "high risk - low terminology acceptance - multimodal re-guidance" for refined management. This tiered health guidance data serves as the core reference for the platform's intelligent intervention decisions and can be dynamically invoked in subsequent training, rehabilitation, or consultations.

[0091] Preferably, the step S3 of decomposing the stratified health guidance data into quantifiable progress units includes:

[0092] The hierarchical health guidance data is decomposed into a structured form to identify the core improvement goals in the health guidance content, and the improvement goal extraction data is obtained.

[0093] Extract motion angle control, muscle stability, movement rhythm, and joint range of motion from the data of the improvement target to form the structural data of the improvement target;

[0094] Based on the improved target structure data, a phased progress standard is designed, and the long-term improvement target is decomposed into three levels of progress targets: short-term, medium-term and long-term, to obtain multi-time domain progress target mapping data.

[0095] The multi-time-domain progress target mapping data is refined into tiny progress units that can be directly measured in daily training, including the reduction value of angle deviation, the percentage improvement of stability, and the index of improvement of movement smoothness, forming a micro-progress evaluation module;

[0096] The micro-progress assessment module is transformed into specific numerical indicators and visual representations, thereby generating quantifiable progress unit data.

[0097] The data collection frequency and feedback cycle of the progress unit are set, and the micro-progress achieved during the user's training process is recorded to obtain progress tracking data.

[0098] After constructing the hierarchical health guidance data, this embodiment of the invention employs a semantic understanding and contextual relationship analysis model (such as a nested intent recognition algorithm based on BERT) to structure the guidance content, transforming the training objectives contained in the text into explicit "improvement goals." For example, for the advanced guidance content of "maintaining scapular stability while performing bent-over rows," the platform automatically breaks it down into two core improvement goals: "enhancing muscle stability in the scapular region" and "controlling the stability of the upper limb rowing angle." This process first performs entity recognition and verb target mapping on the guidance statements, extracting action control words such as "maintain" and "perform," and then performs context matching with the training parts (such as "scapula" and "upper limb") to generate a preliminary target list. Subsequently, the platform prioritizes and removes redundancy based on task word frequency, action result type, and associated risk level, ultimately forming structured "improvement goal extraction data." This data includes fields such as target name, associated actions, affected parts, and priority level, serving as the foundational data for subsequent action-level analysis. From the extracted goals, four core variables involving quantifiable dimensions are selected: motion angle control (such as squatting angle accuracy), muscle stability (such as the stability of electromyographic signals in the scapular region), movement rhythm (such as the consistency of movement execution time), and joint range of motion (such as hip abduction amplitude). The system uses a mapping model between goals and sensor data to annotate the goal fields with measurable parameters within the platform. For example, for the goal of "improving shoulder abduction control," the platform extracts "shoulder abduction angle error less than 5 degrees" as an angle control sub-goal from the angle control category, and "average electromyographic fluctuation rate of the target muscle group less than 15%" as a stability sub-goal from the muscle stability category. A unified structured data entry is constructed, recording the calculation method, required equipment (such as inertial sensors and electromyographic devices), and units for each sub-goal. Finally, the platform generates "improvement goal structure data," organized in JSON format. Each entry includes the goal dimension, reference standard, measurement method, and current user starting state, serving as input data for phased goal division and micro-progress monitoring. Based on the difference between the initial value and the expected target value in the improvement target structure data, combined with the user's training frequency and risk level, a progressive interval strategy is adopted to divide the target into three stages: short-term (e.g., within 7 days), medium-term (e.g., within 21 days), and long-term (e.g., within 60 days). The system divides the total improvement path into arithmetic or weighted increment target intervals and sets reasonable increments with reference to the average user progress in the platform's big data. For example, if a user's "shoulder abduction angle control error" starts at 12 degrees and the platform's target is to control it within 3 degrees, the system sets the short-term target to 9 degrees, the medium-term target to 6 degrees, and the long-term target to 3 degrees, decreasing by 3 degrees every 7 days. Similarly, the "muscle stability volatility" target gradually decreases from the current 25% to 10%, with the platform adjusting the reduction rate for each stage to 5% and 10% based on the training load.All phased goals are mapped to a unified time-axis progress model, generating "multi-time-domain progress goal mapping data," which is used for dynamic feedback and user goal awareness management. Each phased goal is further refined into measurable parameters for daily training units, such as a daily goal of reducing angle error by 0.5 degrees, decreasing electromyographic fluctuation rate by 1%, and reducing execution time fluctuation by 2%. The platform quantifies these changes by defining "micro-progression units" (i.e., small-scale improvement indicators that can be observed after a single training session or multiple training sessions) and establishing a micro-progression indicator library in the system. Taking angle control as an example, the "angle deviation reduction value" is defined as the difference between the average posture of a user's action and the target action, in degrees; muscle stability is expressed as the percentage decrease in the standard deviation of electromyographic signals; and the movement smoothness improvement index is calculated by combining the smoothness of the acceleration curve and the frequency of movement interruptions. All these quantitative indicators are encapsulated into a "micro-progression evaluation module," which is automatically called by the platform after each training session to output evaluation results. The visualization template engine transforms micro-progress indicators into graphical data displays, such as line charts showing the trend of angle error changes, radar charts showing the improvement of muscle stability in various directions, and bar charts comparing the uniformity of movement rhythm at different times. Each indicator is assigned a clear numerical unit and stage label, and items that have not met the target are color-coded (e.g., red warning or yellow alert). For example, if a user's shoulder abduction angle deviation decreases from 5.2 degrees to 4.8 degrees on a given day, the platform displays "Today's Progress: Angle Deviation Reduced by 0.4 Degrees (Achievement Rate 80%)", while the "Fluency Index" chart shows a 12% improvement compared to last week. All charts and values ​​are linked to "Progress Unit Data," which structurally records fields such as date, training program, indicator type, target value, actual value, and achievement rate, supporting subsequent automatic aggregation and trend analysis. The system sets data collection frequencies (e.g., recording data for each training session for high-frequency users, and summarizing data weekly for low-frequency users) and feedback cycles (e.g., daily summaries and weekly trend reports) for different user types. It employs a rolling window statistical model to continuously sample and compare "progress unit data" against the target, identifying trend changes in real time. For example, with a feedback cycle of every three training sessions, the platform automatically extracts angle control and fluency indicators from the latest three training data points to fit a trend line, determining whether the user is currently progressing, stagnating, or regressing. The platform provides the analysis results to the user in the form of a "micro-progress heatmap" or "weekly progress curve," while also storing them in "progress tracking data." This data supports subsequent adjustments to personalized training plans and further optimization of the platform's health guidance strategies. The system also automatically triggers a mechanism to adjust guidance content or terminology when a user experiences prolonged stagnation, forming a closed-loop optimization mechanism.

[0099] Preferably, the anonymous user matching of progress tracking data based on similar fitness characteristics and progress patterns in step S3 includes:

[0100] Time series analysis of progress tracking data is performed to identify user progress rate, stagnation period and breakthrough point characteristics, and to generate progress pattern characteristic data.

[0101] Based on the progress pattern feature data, an anonymous user group with similar fitness goals, physical conditions and initial ability levels is extracted from the pre-set user database to form a candidate matching user pool.

[0102] The similarity between the progress pattern feature data and each user in the candidate matching user pool is calculated to identify anonymous users with high similarity to the progress trajectory, thereby generating user matching results;

[0103] Based on the user matching results, extract the successful breakthrough strategies and incentive factors of the matched users, and combine them with the current user's progress pattern feature data to generate personalized incentive strategies;

[0104] Transform personalized incentive strategies into interactive incentive data that includes success story sharing, breakthrough point prompts, and virtual competition mechanisms.

[0105] This invention receives progress unit data output from the user's micro-progress evaluation module and constructs time series data by combining it with timestamp information. A sliding window method (e.g., using a 7-day window, sliding for 1 day at a time) is employed to analyze the user's training data over multiple time periods. The system identifies the user's progress rate within a specific time period by calculating the rate of change of progress within each window. When the progress value fluctuates minimally or stagnates within multiple consecutive windows, the system determines it as a training stagnation period. When the system detects that a progress indicator (e.g., the improvement index of motion smoothness or the reduction value of angle deviation) significantly increases beyond a certain threshold in a short period (e.g., an increase of more than 20% compared to the average of the previous stage), the system marks it as a breakthrough point. The platform encodes features such as progress rate, stagnation period, and breakthrough point in a structured manner to form progress pattern feature data, which serves as the basis for subsequent user clustering and matching. Based on the progress pattern feature data, the system retrieves the accessed user feature database, which contains a large number of anonymous users' fitness goals (such as core strength enhancement, shoulder joint mobility improvement, etc.), basic physical conditions (such as height, weight, gender, body fat percentage), and initial ability levels (such as initial controllable movement angles, basic stability scores, etc.). The system first uses feature standardization methods to process various indicators of the current user and database users. In the standardized space, the K-nearest neighbor algorithm is used to initially screen out user groups that have similarity to the current user in terms of goals and ability levels (Euclidean distance or Manhattan distance is less than a certain threshold, such as 0.3), forming a candidate matching user pool, which serves as the input basis for the next step of precise matching. For each user in the aforementioned candidate matching user pool, their complete historical progress pattern feature data (including progress rate curves and breakthrough point distributions at each training stage) is extracted. Then, the progress pattern features of the current user are compared with those of each candidate user using Dynamic Time Warping (DTW). The DTW method can capture the similarity of two progress trajectories at different time scales. The platform sets a similarity threshold (e.g., similarity score greater than 0.85) to filter out users with similar progress trajectories. Finally, it outputs a number of anonymous users with the highest matching scores (e.g., Top 5) to form the user matching results, which are used for subsequent successful strategy mining. Based on the anonymous user group identified in the user matching results, we analyze their behavioral changes before and after the breakthrough point, extract their successful breakthrough strategies (such as adjusting training frequency, adding auxiliary training methods, and changing training time periods) and motivating factors (such as participating in virtual competitions, receiving positive feedback, and setting phased rewards). Through the strategy encoding module, we categorize these contents and label their associated progress pattern characteristics. For example, we establish a correspondence between "adding low-intensity stretching recovery training" and "slowing down the improvement of stability before the breakthrough point". Then, combined with the current user's progress pattern characteristic data, we match the most relevant set of breakthrough strategies to form personalized incentive strategies, and output them in a label structure (such as [phased stagnation + shoulder joint limitation] → [strategy: adjust training angle + light stretching assistance]) for the next step of conversion.Personalized incentive strategies are transformed into multimodal interactive incentive data through a content generation module, comprising three main modules: 1) A success story sharing module generates personalized text and image content by matching user corpus materials and progress trajectory graphs, such as showcasing the story of "User A, who also had limited shoulder joint, broke through the 20-degree angle limitation within two weeks through daily small-angle shoulder swing training"; 2) A breakthrough point prompt module pushes real-time prompts based on current user data, such as "A breakthrough in movement stability is expected after the next training session"; 3) A virtual competition mechanism module sets challenge goals based on the user's current stage and the average level of matched users (such as "Increase the stability score by 5% within 3 days"), and enhances training motivation through leaderboards, points rewards, etc., ultimately generating a visual incentive interface for the terminal.

[0106] Preferably, step S3, which involves establishing an initial user movement pattern map based on movement correctness data, tiered health guidance data, and progress tracking data, includes:

[0107] A user data fusion preprocessing mechanism is established based on action correctness data, tiered health guidance data, and progress tracking data to generate a multi-source data integration table;

[0108] Perform time synchronization processing on the multi-source data integration table, establish a unified time axis reference system, and form a time-series synchronized dataset;

[0109] Extract common user action deviation patterns and technical defect characteristics from action correctness data to form user technical characteristic profiles;

[0110] Based on the analysis of hierarchical health guidance data and progress tracking data, user acceptance and execution effectiveness of different types of guidance, as well as the characteristics of ability development pace, user learning characteristic data are constructed.

[0111] Based on user technical feature profiles and user learning characteristic data, identify mutual influence relationships and dependency patterns, and establish feature association networks;

[0112] Based on the feature association network, a predictive model for the development of user motor ability is constructed, and short-term, medium-term and long-term ability predictions are made to generate a user potential assessment report.

[0113] The user potential assessment report is used to generate a visualization scheme that includes node definition, connection rules and graph hierarchy. Visual optimization and interaction design are carried out based on highlighting important nodes, coloring critical paths and implementing interactive functions, thereby forming an initial user motion pattern graph.

[0114] This invention integrates various data generated by users during training. First, the system obtains real-time scores of user training movements from the movement correctness recognition module, including data such as angle deviation, posture error, and disordered movement execution sequence. Simultaneously, it extracts personalized training suggestions, movement decomposition information, and control priorities recommended to users from the hierarchical health guidance system, and synchronously reads micro-progression values ​​and feedback periodic records recorded by the progress tracking module. The platform establishes a unified data interface using user ID and timestamp as anchors through a data fusion preprocessing mechanism. It performs linear imputation or historical data backfilling for missing items using data cleaning functions. For health guidance data with significant structural differences, a structure mapping transformation strategy is adopted (e.g., mapping guidance suggestion text labels to structured control elements). Finally, the data is arranged and combined into data record units with a unified format according to time, constructing a multi-source data integration table for subsequent multi-dimensional modeling and analysis. After obtaining the multi-source data integration table, time synchronization processing is required for records from different data sources. Since motion capture data during user training is typically collected at the second level, while health guidance feedback and progress tracking data are mostly collected at the daily or weekly granularity, the platform employs a timeline alignment algorithm to map all data onto a unified training day-level timeline. Specifically, this involves setting uniform training time slices (e.g., each time slice is one day), downsampling high-frequency data (e.g., taking the maximum or average value within the day), and upsampling low-frequency data (e.g., using forward padding or moving averages). This ensures that each time slice contains complete information such as motion correctness, health guidance acceptance status, and progress feedback, forming a time-series synchronized dataset. This lays a temporal consistency foundation for subsequent user behavior feature modeling. Based on the time-series synchronized dataset, the platform uses statistical clustering and pattern recognition methods to analyze user motion correctness data, extracting common motion deviation patterns and technical defect characteristics. The platform first uses the Local Outlier Factor algorithm to identify recurring postural anomalies (such as knee valgus and shoulder compensation) in the training sequences. Then, it uses K-means clustering to classify these anomaly types. Simultaneously, it combines error values ​​with keyframe positions to extract key technical deficiencies (such as insufficient core strength leading to lower limb instability and delayed upper limb movements). All these features are written into the user's technical feature profile in a standardized manner. Each record includes the name of the deviated movement, its frequency, error magnitude, impact range, and speculated cause, becoming key input features for subsequent ability development prediction. Combining the user's technical feature profile, the platform further analyzes the user's acceptance of different types of guidance content and the training feedback effect from hierarchical health guidance data and progress tracking data.The platform uses a weighted scoring mechanism to analyze the magnitude of micro-progresses users make within a specific time window after receiving specific types of health guidance (such as rhythm guidance, movement relief, and local activation). It also assesses the user's progress cycle (i.e., the number of days required from receiving guidance to achieving the goal) in each training mode, extracting characteristics of ability development rhythm, such as "rapid response to muscle activation suggestions and delayed response to rhythm control suggestions." Finally, the platform uses a tag combination method to construct user learning characteristic data, including dimensions such as guidance preference categories, feedback sensitivity, and training adaptation rhythm, providing a basis for ability development prediction and personalized content delivery. Based on user technical feature profiles and user learning characteristic data, a feature association network is constructed using graph relationship modeling. Specifically, each type of technical feature and learning characteristic is used as a node, and edge weight connections are established between nodes through co-occurrence frequency, causal order, and correlation indices. For example, "movement delay" and "slow rhythm adaptation" form a strong connection edge due to high-frequency co-occurrence in the same training phase. The platform further uses PageRank or graph embedding algorithms to identify key nodes at the center of the network and their influence paths. This feature association network reflects the dynamic dependency between the root causes of user capability limitations and training adaptation methods, serving as a crucial structural foundation for predicting subsequent capability development trends. Based on this feature association network, the platform designs a user motor capability development prediction model. A Graph Convolutional Network (GCN) is used to learn the representation of user node features, and the user's historical progress time series is used as the time input channel. After fusing the graph structure and time series data, the model can output predicted values ​​for key capability indicators (such as stability score, movement accuracy, reaction speed, etc.) of the user in the short (7 days), medium (30 days), and long (90 days) periods. The platform uses historically matched user evolution trajectories to adjust the confidence level of the prediction results and outputs a user potential assessment report. The report indicates the user's current capability position, the predicted upper limit value, and suggested development paths and breakthrough point predictions. The core capability nodes and their evolution paths from the user potential assessment report are visualized into a graph, forming an initial user motor pattern graph. Specifically, it includes three aspects: 1) The node definition module transforms the user's ability components (such as motion control, rhythm control, and joint flexibility) into graph nodes; 2) The connection rules establish directed edges between nodes based on the feature association network and set edge weights according to the training influence path and time dependence; 3) The graph hierarchy is divided into a basic ability layer, a comprehensive ability layer, and a target ability layer.The platform further optimizes the front-end interface visually, including highlighting important nodes by size and color, using gradient lines to enhance user attention on critical paths, and allowing users to view training suggestions and predicted trends by clicking on nodes, thus enabling interactive functionality. Finally, it outputs a visual result in the form of a graph to guide users in long-term training path planning and dynamic target adjustment.

[0115] Preferably, step S4 includes the following steps:

[0116] Step S41: Extract environmental feature parameters from each data point in the initial user motion pattern map during data acquisition, and combine them with image clarity score, skeleton point recognition confidence and tracking stability score to form a data acquisition environment quality assessment table, where environmental feature parameters include light intensity, background complexity and device stability.

[0117] After constructing the initial user motion pattern map, this embodiment of the invention further extracts environmental feature parameters from the video segments or image frames corresponding to each data point in the map to assess the degree of external interference affecting data quality. Specifically, the platform uses built-in or external sensor modules to collect data on light intensity (in Lux), background complexity of the camera equipment (calculated by the number of edges in the image background area or the dispersion of the color histogram), and equipment stability (measured by IMU to measure camera shake frequency and amplitude). The platform then combines these with internal processing indicators extracted from the image processing module, such as image sharpness score (calculated by image gradient, texture richness, or Laplacian variance), skeletal point recognition confidence (derived from the confidence interval score of each skeletal point by the pose recognition model), and tracking stability score (measuring the degree of positional fluctuation of the same skeletal point in consecutive frames), to comprehensively analyze each data point and summarize it into a data acquisition environment quality assessment table. This table records the external environmental interference factors and internal image processing quality corresponding to each motion data point, providing a basis for subsequent judgment of data reliability.

[0118] Step S42: Based on the consistency performance of each data point in the data acquisition environment quality assessment table, assess the degree of environmental influence on each data point and generate the acquisition environment stability index.

[0119] After obtaining the data acquisition environment quality assessment table, this embodiment of the invention evaluates the degree of environmental influence on each data point based on the consistency between the environmental interference index and the image processing stability index. The platform first normalizes the fluctuation of each data point across three external parameters (illuminance, background complexity, and equipment stability) and sets weighting coefficients (e.g., background complexity weight 0.4, illumination intensity weight 0.3, and equipment stability weight 0.3), comparing the corresponding image quality degradation to construct an influence mapping relationship. Then, correlation analysis is used to determine the strength of the linear relationship between the image quality of the data point and environmental fluctuations, thereby judging its environmental sensitivity. For data points whose image quality is significantly affected by environmental changes, their acquisition environment stability index is low; conversely, it is high. The platform generates an acquisition environment stability index for each data point by comprehensively calculating the stability scores of all data points, used to mark whether the data was generated under ideal acquisition conditions.

[0120] Step S43: Record and analyze user feedback behavior on the health guidance data provided by the system, including acceptance, completion status and subjective evaluation, assess the time stability and content consistency of user feedback, and establish a user feedback consistency index;

[0121] This invention comprehensively records and analyzes user feedback behavior regarding the provided health guidance data during system use. This behavioral data includes three aspects: first, whether the user accepts the system's recommended health guidance tasks, i.e., "acceptance"; second, whether the user completes the tasks according to the guidance and uploads training records, as "execution completion status"; and third, whether the user provides subjective evaluations, such as marking effectiveness or providing written feedback, as "subjective evaluation." The platform records the timestamps and content of this feedback data and performs time-series analysis to extract the temporal stability (i.e., whether user feedback occurs regularly and is consistent with the training cycle) and content consistency (i.e., whether the same type of guidance receives consistent evaluations and whether opinions change during repeated executions). Through the fusion of these dimensions, the platform constructs a user feedback consistency index for each user. Users with high indices represent those whose feedback has stability, usability, and high credible reference value.

[0122] Step S44: The stability index of the collected environment and the consistency index of user feedback are weighted and fused together, and the data reliability weight coefficient is generated by combining the importance of each data point in the motion assessment.

[0123] In this embodiment of the invention, after obtaining the stability index of the data acquisition environment and the consistency index of user feedback for each data point, the two indices are further weighted and fused to form a comprehensive data reliability assessment index. The platform first sets weighting factors, such as 0.6 for data acquisition environment stability and 0.4 for user feedback consistency, reflecting the platform's overall trust in the external data acquisition environment and user subjective responses. Then, for each data point in the initial user motion pattern map, its importance in the motion assessment is determined based on its position and role in the map (e.g., whether it is a key node or path node, or whether it participates in the capability assessment model training). This importance is then used as an adjustment factor in the weighted calculation, ultimately assigning a data reliability weight coefficient to each data point. This coefficient quantifies whether the data point is qualified to be trusted and retained in the analysis model and is a key mediating variable for the platform to construct a high-quality assessment system.

[0124] Step S45: Based on the data reliability weight coefficient, set a screening threshold to identify and extract high-confidence data points, form a high-quality data subset, and build a user motion evaluation model;

[0125] This invention uses a data reliability weighting coefficient as the screening criterion, sets a reliability threshold (e.g., 0.75), identifies and extracts high-confidence data points, forming a high-quality data subset. This high-quality data subset excludes data points from abnormal collection environments or those generated by users with inconsistent feedback, retaining data sources with high platform credibility, good training environments, and positive user feedback. Based on this, the platform retrains the user motion assessment model using this subset. The assessment model employs a multi-input multi-output neural network structure, using user technical characteristics, learning characteristics, behavioral performance, and external interference as input nodes. Through training, it extracts the user's current motion ability level score and outputs a predicted trend. The high-quality data subset ensures that the model's training results are more accurate and robust, thereby improving the platform's overall reliability in user ability analysis and personalized recommendation capabilities.

[0126] Step S46: Based on the user motion assessment model and interactive incentive data analysis, analyze the operation fluency and usage frequency of each functional module to generate functional mastery assessment data; implement a low-threshold progressive functional unlocking mechanism based on the functional mastery assessment data and interactive incentive data.

[0127] This invention, based on an updated user motion assessment model, further analyzes user usage patterns and interaction records across various functional modules (such as motion correction, health plan recommendations, feedback records, and interactive incentive modules). It assesses operational fluency (e.g., number of clicks required to complete an operation, average dwell time, error frequency) and usage frequency (e.g., daily visits, single usage duration, operation coverage), generating functional mastery assessment data for each module. The platform then correlates this assessment data with user interactive incentive data (e.g., number of incentive points earned, number of task achievements completed, frequency of challenge participation) to determine the user's current level of mastery of platform functions. Based on this, the platform implements a low-threshold, progressive function unlocking mechanism. This automatically unlocks more training modules or analysis reports once the user's mastery reaches a basic requirement. For example, when the mastery score for the "motion assessment function" exceeds 60 points and meets the standard for three consecutive days, the "goal prediction function" is automatically unlocked. This mechanism reduces the initial cognitive burden on users while enhancing the sense of accomplishment in later function exploration, thereby improving overall platform stickiness and user satisfaction.

[0128] Of particular importance, step S46 includes the following steps:

[0129] Based on the user activity assessment model, the usage time, access frequency and operation sequence of users in each functional module of the platform are recorded, and time distribution analysis is performed to identify peak user usage periods and changes in functional preferences. In combination with user ability characteristics, a mapping relationship of user functional needs is constructed.

[0130] Based on the mapping relationship of user functional needs, we extract operation path features and decision node selection tendencies, combine them with user response features in interactive incentive data, identify user functional exploration methods and learning paths, and generate user behavior pattern analysis results.

[0131] Based on the analysis results of user behavior patterns, the system counts the time it takes for users to complete operations in each functional module, the number of retries, and the frequency of querying auxiliary information. Combined with the frequency and duration of function usage, the system calculates the operation fluency index and function familiarity score for each function, thus forming the function mastery assessment data.

[0132] A low-threshold, gradual feature unlocking mechanism is implemented based on feature mastery assessment data and interaction incentive data.

[0133] In this embodiment of the invention, after constructing a high-quality motion ability profile based on a user motion assessment model, the platform further records detailed behavioral data of users in various functional modules of the platform in real time, including the start and end times of each function usage, access frequency, click sequence, and page jump path. By parsing and reconstructing the sequence of this data using timestamps, the platform can extract the usage duration and access frequency of each user in specific functions (such as "action guidance," "feedback submission," and "training task browsing"), and establish an operation event timeline. The platform uses time distribution analysis methods (such as hourly-granular heatmap generation or normal distribution fitting) to identify the peak periods of user function usage (such as 19:00–21:00 daily) and their changing trends over time. Simultaneously, it performs correlation analysis between frequently used functions and the ability characteristics reflected in the motion assessment model (such as users with higher core stability scores preferring dynamic balance training modules), thereby constructing a mapping relationship of user function needs and clarifying the function preferences and usage habits of each user group at different ability levels. Based on the mapping relationship of user functional needs, the platform further extracts the common operation path characteristics of each user within the platform, i.e., the jump logic between modules when using the platform, such as the sequence "training plan → action guidance → feedback record → personal summary". The platform uses path sequence mining algorithms (such as PrefixSpan sequence pattern recognition) to obtain high-frequency operation paths and statistically analyzes the selection tendencies at key nodes (i.e., "decision nodes") in each path. For example, in the "training recommendation" module, users tend to click on "basic difficulty" rather than "advanced challenge". On this basis, combined with the user's reaction characteristics in the interactive incentive data (such as response time, completion rate, and skip rate to incentive tasks), the platform can identify the user's functional exploration methods (such as linear progression, skip attempts, and repeated review) and overall learning path preferences (such as a preference for understanding the content before practicing, or watching and doing simultaneously). Through the above analysis, the platform generates user behavior pattern analysis results, forming a structured profile of the user's functional usage logic, selection habits, and exploration depth within the platform, providing behavioral basis for subsequent functional proficiency judgment. By leveraging user behavior pattern analysis results, the platform quantitatively evaluates each user's performance across various functional modules, primarily focusing on three key dimensions: operation completion time (the average time required for a user to successfully complete a task from entering a functional module), number of retries (the average number of times an operation fails, reverts, or resubmits during the completion of the same task), and frequency of auxiliary information queries (such as clicking the "Help" button, viewing "Operation Examples," or "Frequently Asked Questions"). After statistically summarizing this behavioral data, the platform combines the frequency of use and duration of use of each functional module to derive a weighted calculation method for each module's "Operation Fluency Index" (measuring the user's proficiency and ease of use when using the function) and "Function Familiarity Score" (measuring the user's ability to independently complete the function's operation).For example, in the "Personalized Training Recommendation" module, if a user has a short average operation completion time, low retry count, rarely consults help documentation, and uses the service frequently, the system will give them a higher fluency and familiarity score; conversely, a lower score will be given for lower scores. These scores will be used to further construct an evaluation framework for users' mastery of platform functions. Based on each user's operation fluency index and function familiarity score in different functional modules, combined with their interaction incentive data (such as earned points levels, incentive task completion cycles, and continuous usage days), a "low-threshold progressive function unlocking mechanism" will be dynamically implemented. The core of this mechanism is to automatically release more advanced functions based on the user's current level of mastery, thereby increasing the user's exploration enthusiasm and lowering the threshold for understanding functions. For example, the platform sets the unlocking threshold for the "Training Load Adjustment" module as follows: if the user's function familiarity score in the "Action Evaluation" module is ≥70 and the incentive task completion rate is ≥60%, the automatic unlocking process will be triggered, and a notification will be sent to remind the user to "unlock new functions". Furthermore, to encourage users to continuously improve their skills, the platform adopts a "low threshold + time-driven" strategy. Even if a user's rating is slightly low, they can gradually unlock the next level of functionality, such as "training prediction" and "action comparison and review," by continuously using a basic functional module for more than 5 days and accumulating a certain amount of behavioral records. Through this gradual unlocking strategy, the platform ensures controllable user experience while helping users gradually master the entire functional chain, thereby enhancing the overall depth and stickiness of platform usage.

[0134] The present invention also includes a system for constructing a sports information intelligent service platform, used to execute the above-described method for constructing a sports information intelligent service platform, wherein the system for constructing the sports information intelligent service platform includes:

[0135] The motion trajectory analysis module is used to collect user motion image data and user basic health data; extract key body node position information during the user's movement; generate user motion trajectory curve based on key node position information; and calculate the deviation angle and distance from the standard trajectory template to obtain action correctness data.

[0136] The fitness risk assessment module is used to build a user fitness risk assessment model based on movement correctness data and user health baseline data, analyze the difference from the preset safety threshold, and generate tiered health guidance data.

[0137] The progress tracking and interaction matching module is used to decompose the hierarchical health guidance data into quantifiable progress units to generate progress tracking data; to perform anonymous user matching based on similar fitness characteristics and progress patterns on the progress tracking data to form interaction incentive data; and to build an initial user exercise pattern map based on movement correctness data, hierarchical health guidance data, and progress tracking data.

[0138] The user motion assessment module is used to calculate the stability index of the data collection environment and the consistency index of user feedback for each data point in the initial user motion pattern map, and to screen high-confidence data points to build a user motion assessment model. Based on the user motion assessment model and interactive incentive data, the module analyzes the smoothness of operation and frequency of use of each functional module, and implements a low-threshold progressive function unlocking mechanism in combination with interactive incentive data.

[0139] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0140] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for constructing a sports information intelligent service platform, characterized in that, Includes the following steps: Step S1: Collect user motion image data and user basic health data; extract key body node location information during the user's movement process; The user's motion trajectory curve is generated based on the location information of key nodes, and the deviation angle and distance from the standard trajectory template are calculated to obtain motion correctness data. Step S2: Construct a user fitness risk assessment model based on movement correctness data and user health baseline data, analyze the difference from the preset safety threshold, and generate tiered health guidance data. Step S2 includes: Step S21: Extract motion deviation characteristic indicators from motion correctness data to form motion risk basic assessment data, where motion deviation characteristic indicators include the maximum deviation value, average deviation value, and deviation duration of key nodes; Step S22: Extract user physiological characteristic parameters from user basic health data to generate user physiological risk factor data; Step S23: Establish a correlation model between the basic assessment data of movement risk and the user's physiological risk factor data, calculate the degree of injury risk to the user for various movement deviations, and generate individualized risk assessment values; Step S24: Based on the individualized risk assessment value and the preset safety threshold, risk level stratification is performed to generate risk classification data; Step S25: Perform differentiated processing on action deviations of different risk levels based on risk classification data, and record action deviation patterns and guidance responses to obtain historical user feedback data; Step S26: Visualize health guidance content based on risk grading data and user historical feedback data to generate tiered health guidance data; Step S3: Decompose the tiered health guidance data into quantifiable progress units to generate progress tracking data; perform anonymous user matching based on similar fitness characteristics and progress patterns on the progress tracking data to form interactive incentive data; establish an initial user exercise pattern map based on movement correctness data, tiered health guidance data, and progress tracking data; establishing the initial user exercise pattern map based on movement correctness data, tiered health guidance data, and progress tracking data includes: A user data fusion preprocessing mechanism is established based on action correctness data, tiered health guidance data, and progress tracking data to generate a multi-source data integration table; Perform time synchronization processing on the multi-source data integration table, establish a unified time axis reference system, and form a time-series synchronized dataset; Extract common user action deviation patterns and technical defect characteristics from action correctness data to form user technical characteristic profiles; Based on the analysis of hierarchical health guidance data and progress tracking data, user acceptance and execution effectiveness of different types of guidance, as well as the characteristics of ability development pace, user learning characteristic data are constructed. Based on user technical feature profiles and user learning characteristic data, identify mutual influence relationships and dependency patterns, and establish feature association networks; Based on the feature association network, a predictive model for the development of user motor ability is constructed, and short-term, medium-term and long-term ability predictions are made to generate a user potential assessment report. The user potential assessment report is used to generate a visualization scheme that includes node definition, connection rules and graph hierarchy. Visual optimization and interaction design are carried out based on highlighting important nodes, coloring critical paths and implementing interactive functions, thereby forming an initial user motion pattern graph. Step S4: Calculate the stability index of the data collection environment and the consistency index of user feedback for each data point in the initial user motion pattern map, and select high-confidence data points to construct a user motion evaluation model; based on the user motion evaluation model and interactive incentive data, analyze the operational fluency and usage frequency of each functional module, and implement a low-threshold progressive function unlocking mechanism in conjunction with the interactive incentive data. Step S4 includes: Step S41: Extract environmental feature parameters from each data point in the initial user motion pattern map during data acquisition, and combine them with image clarity score, skeleton point recognition confidence and tracking stability score to form a data acquisition environment quality assessment table, where environmental feature parameters include light intensity, background complexity and device stability. Step S42: Based on the consistency performance of each data point in the data acquisition environment quality assessment table, assess the degree of environmental influence on each data point and generate the acquisition environment stability index. Step S43: Record and analyze user feedback behavior on the health guidance data provided by the system, including acceptance, completion status and subjective evaluation, assess the time stability and content consistency of user feedback, and establish a user feedback consistency index; Step S44: The stability index of the collected environment and the consistency index of user feedback are weighted and fused together, and the data reliability weight coefficient is generated by combining the importance of each data point in the motion assessment. Step S45: Based on the data reliability weight coefficient, set a screening threshold to identify and extract high-confidence data points, form a high-quality data subset, and build a user motion evaluation model; Step S46: Based on the user motion assessment model and interactive incentive data analysis, analyze the operational fluency and usage frequency of each functional module to generate functional mastery assessment data; implement a low-threshold progressive functional unlocking mechanism based on the functional mastery assessment data and interactive incentive data.

2. The method for constructing a sports information intelligent service platform according to claim 1, characterized in that, Step S1 includes the following steps: Step S11: Collect multiple frames of continuous images of the user performing fitness movements through the camera of the smart terminal device to generate user motion image data; Step S12: Collect the user's height, weight, age, and exercise habits to form the user's basic health data; Step S13: Perform image preprocessing on the user motion image data and identify skeletal points to locate key body nodes of the user's shoulder, elbow, wrist, hip, knee, and ankle, and generate key body node position information; Step S14: Generate the user's motion trajectory curve based on the key node location information, and calculate the deviation angle and distance from the standard trajectory template to obtain motion correctness data.

3. The method for constructing a sports information intelligent service platform according to claim 2, characterized in that, Step S14 includes the following steps: Step S141: Perform time-series analysis on the position information of key body nodes, record the position changes of each key node in continuous time frames, and generate raw data of node motion trajectory; Step S142: Smooth the original data of the node motion trajectory to obtain the user motion trajectory curve data; Step S143: Calculate the motion angle, velocity, and acceleration of the human motion chain segment formed between adjacent key nodes based on the user motion trajectory curve data, and construct a dynamic parameter set for the user motion chain segment; Step S144: Match the user's dynamic parameter set of motion chain segments with the standard parameter set in the action library to determine the type of fitness action currently being performed by the user and generate action type recognition results; Step S145: Extract the corresponding standard trajectory template data from the action standard library based on the action type recognition result, and perform spatiotemporal alignment between the user motion trajectory curve data and the standard trajectory template data to obtain normalized trajectory comparison data; Step S146: Calculate the Euclidean distance deviation between the user trajectory and the standard trajectory at each key time point in the normalized trajectory comparison data, generate a point deviation sequence, and calculate the angle deviation between the user motion chain segment and the standard motion chain segment at each key time point, generate an angle deviation sequence. Step S147: Calculate the overall motion trajectory deviation score based on the position deviation sequence and angle deviation sequence to obtain motion correctness data.

4. The method for constructing a sports information intelligent service platform according to claim 3, characterized in that, Step S25 includes the following steps: Step S251: Extract the motion chain segment angle deviation value, speed control deviation value and motion stability index from the low-risk level motion deviation in the risk classification data, and quantitatively compare them with the preset standard motion model to identify the optimizable parameters and obtain the low-risk optimizable parameters. Step S252: Develop a progressive adjustment plan for low-risk optimizable parameters to generate general optimization recommendation data; Step S253: Analyze the proximity of the motion deviation points to the human biomechanical safety limits for the medium-risk level motion deviations in the risk classification data, and calculate the potential stress concentration areas. Step S254: Based on potential stress concentration areas and user physiological risk factor data, assess the risk of injury to the user's joints, muscles and ligaments caused by the deviation, and generate basic safety guidance data including descriptions of prohibited movements, definitions of safe movement ranges and recommendations for alternative movements; Step S255: Extract the degree and duration of the movement deviations exceeding the safety threshold from the high-risk level movement deviations in the risk classification data, and classify the errors based on the high-risk movement deviations to obtain error type data, which includes technical errors, physical limitation errors, or errors with potential health problems. Step S256: Generate professional consultation guidance data based on error type data; Step S257: Record the user's behavioral response after receiving general optimization suggestion data, basic safety guidance data, or professional consultation guidance data, including the time delay in accepting guidance, the completion rate of implementing guidance, and the persistence in continuously applying guidance, and generate user historical feedback data.

5. The method for constructing a sports information intelligent service platform according to claim 4, characterized in that, Step S3, which involves decomposing the stratified health guidance data into quantifiable progress units, includes: The hierarchical health guidance data is decomposed into a structured form to identify the core improvement goals in the health guidance content, and the improvement goal extraction data is obtained. Extract motion angle control, muscle stability, movement rhythm, and joint range of motion from the data of the improvement target to form the structural data of the improvement target; Based on the improved target structure data, a phased progress standard is designed, and the long-term improvement target is decomposed into three levels of progress targets: short-term, medium-term and long-term, to obtain multi-time domain progress target mapping data. The multi-time-domain progress target mapping data is refined into tiny progress units that can be directly measured in daily training, including the reduction value of angle deviation, the percentage improvement of stability, and the index of improvement of movement smoothness, forming a micro-progress evaluation module; The micro-progress assessment module is transformed into specific numerical indicators and visual representations, thereby generating quantifiable progress unit data. The data collection frequency and feedback cycle of the progress unit are set, and the micro-progress achieved during the user's training process is recorded to obtain progress tracking data.

6. The method for constructing a sports information intelligent service platform according to claim 5, characterized in that, Step S3, which involves anonymously matching progress tracking data based on similar fitness characteristics and progress patterns, includes: Time series analysis of progress tracking data is performed to identify user progress rate, stagnation period and breakthrough point characteristics, and to generate progress pattern characteristic data. Based on the progress pattern feature data, an anonymous user group with similar fitness goals, physical conditions and initial ability levels is extracted from the pre-set user database to form a candidate matching user pool. The similarity between the progress pattern feature data and each user in the candidate matching user pool is calculated to identify anonymous users with high similarity to the progress trajectory, thereby generating user matching results; Based on the user matching results, extract the successful breakthrough strategies and incentive factors of the matched users, and combine them with the current user's progress pattern feature data to generate personalized incentive strategies; Transform personalized incentive strategies into interactive incentive data that includes success story sharing, breakthrough point prompts, and virtual competition mechanisms.

7. A system for constructing a sports information intelligent service platform, characterized in that, The system for constructing the intelligent sports information service platform as described in claim 1, for executing the method of constructing the intelligent sports information service platform, comprises: The motion trajectory analysis module is used to collect user motion image data and user basic health data; extract key body node position information during the user's movement; generate user motion trajectory curve based on key node position information; and calculate the deviation angle and distance from the standard trajectory template to obtain action correctness data. The fitness risk assessment module is used to build a user fitness risk assessment model based on movement correctness data and user health baseline data, analyze the difference from the preset safety threshold, and generate tiered health guidance data. The progress tracking and interaction matching module is used to decompose the hierarchical health guidance data into quantifiable progress units to generate progress tracking data; to perform anonymous user matching based on similar fitness characteristics and progress patterns on the progress tracking data to form interaction incentive data; and to build an initial user exercise pattern map based on movement correctness data, hierarchical health guidance data, and progress tracking data. The user motion assessment module is used to calculate the stability index of the data collection environment and the consistency index of user feedback for each data point in the initial user motion pattern map, and to screen high-confidence data points to build a user motion assessment model. Based on the user motion assessment model and interactive incentive data, the module analyzes the smoothness of operation and frequency of use of each functional module, and implements a low-threshold progressive function unlocking mechanism in combination with interactive incentive data.

Images