Running posture monitoring method based on image recognition technology

Abstract

The invention relates to the technical field of motion monitoring, in particular to a running posture monitoring method based on an image recognition technology, which comprises the following steps: deploying cameras at different angles, synchronously recording whole body actions of a user during running, extracting a joint angle change sequence, and constructing a running three-dimensional action model. According to the method, the consistency and the difference of the actions are recognized by comparing the amplitude and the speed of the joint movement, so that the detection of the action defects is facilitated, and the stability and the deviation of the gait can be quantitatively evaluated by analyzing the foot landing angle and the stride change; analysis of the motion change frequency and amplitude allows coaches and athletes to adjust and optimize gaits in a targeted manner in the training process, the risk of sports injury is reduced, and through continuous simulation and motion symmetry evaluation, recognition of sports defects and adjustment of the camera angle and shooting speed, the training effect of the coaches and the athletes is improved. And the data capture and analysis process in the training process is more efficient and targeted.

Description

Technical Field

[0001] The present invention relates to the technical field of motion monitoring, and in particular to a running posture monitoring method based on image recognition technology. Background Art

[0002] The field of sports monitoring technology involves the use of sensors, cameras, and tracking devices to monitor and analyze athletes' performance, health, and athletic skills. Advanced image recognition, sensing technology, and data analysis methods are constantly being used to provide real-time feedback on athletes' posture, speed, heart rate, and more. Through these technologies, coaches and sports scientists can obtain precise data to optimize training plans, reduce the risk of injury, and improve sports performance. The development of modern sports monitoring technology, especially systems that integrate machine learning and artificial intelligence, provides unprecedented accuracy and convenience for individual training and team management.

[0003] Among them, the user running posture analysis method based on image recognition refers to the use of image recognition technology to analyze the runner's posture to identify the movement efficiency and movement defects when running. This analysis helps runners and coaches understand the bad posture that leads to injuries, optimize running techniques, improve performance and reduce the chance of injury. Analyzing the runner's movements in real time or through recorded videos can provide detailed information about various parts of the body, such as the foot landing angle, leg swing, etc., and is an important tool in modern sports training.

[0004] Although existing sports monitoring technology can provide feedback on athletes' posture, speed and heart rate, it is limited by the angle and resolution of the equipment, resulting in the captured data not being comprehensive or accurate enough. This limitation means that coaches and sports scientists lack sufficient data support when analyzing athletes' specific movements and improving training plans. Especially in terms of movement symmetry and subtle movement analysis, existing technologies have difficulty accurately capturing subtle changes in movements, which is particularly critical in fine-tuning sports performance and preventing injuries. Existing technologies have delays in real-time feedback and long-term tracking analysis, affecting real-time adjustments to training and evaluation of effects. These deficiencies prevent athletes from fully optimizing their performance in high-level sports training and competitions, and even increase the risk of injury. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a running posture monitoring method based on image recognition technology.

[0006] In order to achieve the above object, the present invention adopts the following technical solution, a running posture monitoring method based on image recognition technology, comprising the following steps: S1: Deploy cameras with different angles to synchronously record the user's full-body running movements, extract the joint angle change sequence, and construct a three-dimensional running movement model; S2: Based on the three-dimensional running motion model, analyze the time series data of the left and right leg movements, compare the joint movement amplitude and speed, identify the consistency and difference of the movements, and obtain the leg movement analysis results; S3: using the leg motion analysis result and image recognition technology, analyzing the foot landing angle and stride change, determining the gait stability and deviation through the motion change frequency and amplitude, and outputting the gait stability analysis result; S4: Based on the gait stability analysis result, adjusting the leg parameters of the three-dimensional motion model, simulating differentiated running motion sequences, and analyzing the leg motions through continuous motion simulation to obtain motion symmetry evaluation results; S5: According to the movement symmetry evaluation result, the user's running movement is compared with the preset ideal movement, and the movement deviation of the runner is detected through a set threshold value, potential movement defects are identified, and deviation analysis results are obtained; S6: Using the deviation analysis results, analyze the motion data of the differentiated training stage, adjust the angle and shooting speed of the camera, optimize the data capture process, extract the key change points of the stride length and the foot landing pressure, and obtain a gait adjustment plan.

[0007] As a further solution of the present invention, the three-dimensional running motion model includes three-dimensional angle data of joints, motion trajectories of multiple parts of the body and real-time motion rendering, the leg motion analysis results include motion synchronization scores, joint motion range differences and motion speed comparisons, the gait stability analysis results include step frequency stability metrics, statistical fluctuations in stride and landing angle analysis, the motion symmetry assessment results include left and right leg motion consistency indexes, quantitative assessments of symmetry imbalances and visual comparisons of simulated motions, the deviation analysis results include numerical values ​​of joint motion deviations, motion defect types and risk level assessments, and the gait adjustment plan includes adjusted camera angle settings, data collection optimization points and key motion correction measures.

[0008] As a further solution of the present invention, cameras with different angles are deployed to synchronously record the whole body movements of the user while running, extract the joint angle change sequence, and construct the three-dimensional running movement model in the following steps: S101: deploy multiple cameras on both sides and above the running track to capture the user's full-body running movements in all directions, set the camera resolution and shooting frequency to optimize the clarity and data richness of motion capture, and generate full-body motion capture data; S102: Based on the whole-body motion capture data, synchronously record the joint angles of the user at multiple moments while running, including angle changes of the knees, elbows, and shoulders, and set a data collection frequency to generate a joint angle change sequence; S103: Using the joint angle change sequence, three-dimensional modeling is performed to reflect the dynamic relationship and coordination of multiple parts of the body during running, and the model is optimized and adjusted to construct a three-dimensional running motion model.

[0009] As a further solution of the present invention, based on the three-dimensional running motion model, the time series data of the left and right leg motions are analyzed, the joint motion amplitude and speed are compared, the consistency and difference of the motions are identified, and the steps of obtaining the leg motion analysis results are specifically as follows: S201: extracting time series data of the left and right legs according to the three-dimensional running motion model, including position and speed information of each joint in motion, setting a time window to synchronously compare the data of the left and right legs, and generating time series comparison data of the left and right legs; S202: Based on the time series comparison data of the left and right legs, calculate and compare the movement amplitude and speed of each joint during the running process, identify and evaluate the movement consistency and difference of the legs on both sides, and generate joint movement analysis results; S203: Using the joint motion analysis results, evaluate the coordination and efficiency of the runner's left and right legs, use graphical tools to display the synchronization and difference of the left and right leg movements, and generate leg motion analysis results.

[0010] As a further solution of the present invention, the leg motion analysis results are used to analyze the foot contact angle and stride changes using image recognition technology, and the gait stability and deviation are determined by the motion change frequency and amplitude. The steps of outputting the gait stability analysis results are specifically as follows: S301: Based on the leg motion analysis result, a linear regression analysis method is used to quantitatively model the relationship between the foot landing angle and the stride, and a statistical relationship model between the angle and the stride is constructed; S302: Calculate the changing frequency and amplitude of the foot movement through the statistical relationship model between the angle and the stride length according to the image recognition technology, evaluate the periodicity and repeatability of the gait, adjust the sensitivity of the frequency analysis to capture the change of the gait, and generate a gait periodicity analysis result; S303: Using the gait periodicity analysis result, evaluate the stability and deviation of the gait, display the deviation of the gait through dynamic simulation, identify the influencing factors, and generate a gait stability analysis result.

[0011] As a further solution of the present invention, the formula of the linear regression analysis method is as follows: Where S is the stride length of each step, θ represents the foot angle when landing, k is the linear coefficient of the relationship between the landing angle and the stride length, b is the intercept of the linear model, Δθ is the difference between the landing angle of the current step and the landing angle of the previous step, and θ refis the reference base angle.

[0012] As a further solution of the present invention, based on the gait stability analysis result, the leg parameters of the three-dimensional motion model are adjusted to simulate a differentiated running motion sequence, and the leg motion is analyzed through continuous motion simulation to obtain the motion symmetry evaluation result. Specifically, the steps are as follows: S401: input the gait stability analysis result into the running three-dimensional motion model, adjust the leg motion parameters, including the angle limit and dynamic characteristics of the joints, check the simulated running motion matching the real-time motion characteristics according to the gait deviation data, and generate an adjusted three-dimensional motion model; S402: Using the adjusted three-dimensional motion model, by simulating differentiated running conditions, including speed changes and ground types, recording and analyzing responses of leg motions, adjusting simulation frequency and motions, and generating differentiated running motion data; S403: Analyze the differentiated running motion data, evaluate the symmetry and coordination of the left and right leg motions, and generate a motion symmetry evaluation result by comparing the synchronization of the original and adjusted motions.

[0013] As a further solution of the present invention, according to the movement symmetry evaluation result, the user's running movement is compared with the preset ideal movement, the runner's movement deviation is detected by a set threshold, and potential movement defects are identified. The steps of obtaining the deviation analysis result are specifically as follows: S501: extracting comparison data between the user's real-time running action and the preset ideal action through the action symmetry evaluation result, setting analysis thresholds to quantify the runner's deviations at key action points, including stride, foot angle, and knee bending degree, and generating action deviation quantification data; S502: Based on the action deviation quantification data, identify the volatility and deviation range in each action cycle, record the gap between the real-time action and the ideal model, and generate deviation identification information; S503: Using the deviation identification information, evaluate the runner's motion efficiency and movement quality, identify the user's potential motion defects in running through multi-dimensional deviation analysis, and generate deviation analysis results.

[0014] As a further solution of the present invention, the steps of using the deviation analysis results to analyze the motion data of the differentiation training stage, adjusting the angle and shooting speed of the camera, optimizing the data capture process, extracting the key change points of the stride length and the foot landing pressure, and obtaining the gait adjustment solution are specifically as follows: S601: adjusting the angle and position of the camera based on the deviation analysis result to capture the running leg and foot movements of the user, adjusting the shooting speed to match the data capture requirements of the differentiated running speed, and generating an optimized camera configuration; S602: using the optimized camera configuration, capturing the motion data of the user during the differentiated training phase, focusing on analyzing the stride length and foot contact pressure, capturing key change points, including the maximum contact force and dynamic changes of the stride, and generating key motion data; S603: Based on the key action data, analyze the changing trends of the stride length and the ground pressure, predict the potential improvement effect after adjusting the action, adjust and optimize the action through real-time feedback, and generate a gait adjustment plan.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by deploying cameras with differentiated angles to synchronously record the whole body movements of runners, more detailed and comprehensive motion data can be captured, so that the extraction of joint angle change sequences is more accurate, and a more accurate three-dimensional motion model is constructed. The accuracy of the motion model is improved, so that the analysis of the time series data of the left and right leg movements is more detailed, and the amplitude and speed of joint movements can be effectively compared to identify the consistency and difference of the movements. This comparison not only helps to find motion defects, but also can quantitatively evaluate the stability and deviation of gait through the analysis of foot landing angle and stride changes. The detailed analysis of motion change frequency and amplitude further allows coaches and athletes to make targeted adjustments during training, optimize gait, and reduce the risk of sports injuries. Through continuous simulation and motion symmetry evaluation, motion defects can be more accurately identified, the camera angle and shooting speed can be adjusted, and the data capture and analysis process during training can be more efficient and targeted. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the workflow of the present invention.

[0017] Figure 2 This is a detailed flow chart of S1 of the present invention.

[0018] Figure 3 This is a detailed flow chart of S2 of the present invention.

[0019] Figure 4 This is a detailed flow chart of S3 of the present invention.

[0020] Figure 5 This is a detailed flow chart of S4 of the present invention.

[0021] Figure 6 This is a detailed flow chart of S5 of the present invention.

[0022] Figure 7 This is a detailed flow chart of S6 of the present invention.

[0023] Figure 8 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0025] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0026] See also Figure 1 The present invention provides a technical solution, a running posture monitoring method based on image recognition technology, comprising the following steps: S1: Deploy cameras with different angles to synchronously record the user's full-body running movements, extract the joint angle change sequence, and construct a three-dimensional running movement model; S2: Extract the joint data of leg motion from the three-dimensional running motion model, analyze the time series data of the left and right leg motions, compare the joint motion amplitude and speed, identify the consistency and difference of the motions, and obtain the leg motion analysis results; S3: Using the leg motion analysis results and image recognition technology, analyze the foot landing angle and stride changes, determine the gait stability and deviation through the movement change frequency and amplitude, and output the gait stability analysis results; S4: Based on the gait deviation analysis results, the leg parameters of the three-dimensional motion model are adjusted to simulate differentiated running motion sequences. The symmetry of the leg motions is analyzed through continuous motion simulations. The symmetry problems of each motion are identified and recorded to obtain the motion symmetry evaluation results. S5: According to the movement symmetry evaluation result, the user's running movement is compared with the preset ideal movement, the runner's movement deviation is detected through the set threshold, asymmetric leg movement and potential movement defects are identified, and the deviation analysis result is obtained; S6: Use the deviation analysis results to identify the cause of the deviation, analyze the motion data in the differentiated training phase, adjust the camera angle and shooting speed, optimize the data capture process, extract the key change points of stride length and foot landing pressure, and obtain the gait adjustment plan.

[0027] The three-dimensional running motion model includes three-dimensional angle data of joints, motion trajectories of multiple parts of the body and real-time motion rendering. The leg motion analysis results include motion synchronization scores, differences in joint motion ranges and motion speed comparisons. The gait stability analysis results include cadence stability measurements, statistical fluctuations in stride length and landing angle analysis. The motion symmetry assessment results include left and right leg motion consistency indexes, quantitative assessments of symmetry imbalances and visual comparisons of simulated motions. The deviation analysis results include numerical values ​​of joint motion deviations, motion defect types and risk level assessments. The gait adjustment plan includes adjusted camera angle settings, data collection optimization points and key motion correction measures.

[0028] See also Figure 2 , deploy cameras with different angles, synchronously record the user's whole body running movements, extract the joint angle change sequence, and build a running 3D motion model in the following steps: S101: deploy multiple cameras on both sides and above the running track to capture the user's full-body running movements in all directions, set the camera resolution and shooting frequency to optimize the clarity and data richness of motion capture, and generate the full-body motion capture data. The execution process is as follows; Deploy multiple cameras on both sides and above the runway, adjust the cameras to optimize the clarity and data richness of motion capture, and adjust the camera resolution and shooting frequency to ensure full-body capture of the user's running movements. The key is to set appropriate resolution and frame rate to ensure data quality and integrity. Use specialized software to synchronize and integrate data to ensure that data obtained from all angles can accurately reflect the athlete's movement characteristics. The collected data will be used for subsequent analysis and research to ensure that full-body motion capture data can be fully recorded and generated.

[0029] S102: Based on the whole body motion capture data, the joint angles of the user at multiple moments while running are synchronously recorded, including the angle changes of the knees, elbows and shoulders, and the data collection frequency is set to generate the execution flow of the joint angle change sequence as follows; Synchronously record the user's joint angles at multiple moments while running, and calculate the joint angle change sequence according to the following formula: Where θ(t) represents the joint angle, θ0 represents the initial joint angle, ω represents the initial angular velocity, α represents the angular acceleration, and t represents time.

[0030] Detailed explanation of the formula and the process of formula calculation: Considering that the change of joint angle during running can be regarded as uniform acceleration motion, the initial joint angle θ0=45°, the initial angular velocity ω=10° / s, and the angular acceleration α=0.5° / s are obtained through monitoring. 2 , calculate the value of the joint angle at time t=2s: The results show that from the initial moment to two seconds later, the joint angle increased from 45° to 55°.

[0031] S103: Using the joint angle change sequence, three-dimensional modeling is performed to reflect the dynamic relationship and coordination of multiple parts of the body during running, and the model is optimized and adjusted. The execution process of constructing a three-dimensional running motion model is as follows; 3D modeling is performed using the sequence of joint angle changes. The model is optimized and adjusted by accurately simulating the dynamic relationship and coordination of multiple parts of the body during running. The model is adjusted in detail using professional software to ensure that the movements of each joint accurately reflect the actual situation. The simulation software will generate a motion model based on the input joint angle data. This data includes angle changes of key joints such as the knees, elbows and shoulders. Such motion simulation not only helps athletes in training, but also provides important data support for sports science research and constructs a 3D running motion model.

[0032] See also Figure 3 ,Based on the running three-dimensional motion model, the time series data of the left and right leg movements are analyzed, the joint movement amplitude and speed are compared, and the consistency and difference of the movements are identified. The specific steps to obtain the leg movement analysis results are as follows: S201: extracting the time series data of the left and right legs according to the three-dimensional running motion model, including the position and speed information of each joint in motion, setting a time window to synchronously compare the data of the left and right legs, and generating the execution flow of the time series comparison data of the left and right legs is as follows; Based on the three-dimensional running motion model, the time series data of the left and right legs are extracted, including the position and speed information of each joint in motion. The data of the left and right legs are synchronously compared by setting a time window to ensure the synchronization and comparability of the data. The coordination and efficiency of the left and right legs in motion can be observed through comparative analysis. The extraction of time series data involves advanced data processing technology to ensure that the data extracted from the three-dimensional model is accurate. These data include the specific position and movement speed of the joints. During the comparison process, the data window is used to synchronously analyze every movement moment of the left and right legs to ensure that the movement differences of the left and right legs can be accurately compared and evaluated, and to generate time series comparison data for the left and right legs.

[0033] S202: Based on the time series comparison data of the left and right legs, the movement range and speed of each joint during running are calculated and compared, the movement consistency and difference of the legs on both sides are identified and evaluated, and the execution process of generating the joint movement analysis results is as follows; Calculate and compare the range of motion and speed of each joint during running, and calculate the joint motion analysis results according to the following formula: In the formula, V represents speed, x 11 ,y 11 、z 11 and x 22 ,y 22 、z 22 They represent the spatial coordinates of the joint at two moments, and Δt represents the time interval.

[0034] Detailed explanation of the formula and the process of formula calculation: Assume that the coordinates of a joint at time t1 are (2, 3, 5) and at time t2 are (5, 6, 2). The time interval Δt=2s, calculate the speed: The results indicate that the joint velocity during this period was 2.25 m / s, which was used to assess the efficiency and coordination of the movement.

[0035] S203: Using the joint motion analysis results, evaluate the coordination and efficiency of the runner's left and right legs, and use graphical tools to display the synchronization and difference of the left and right leg movements. The execution process of generating the leg motion analysis results is as follows; The results of joint motion analysis are used to evaluate the coordination and efficiency of the runner's left and right legs. Graphical tools are used to show the synchronization and differences in the left and right leg movements. In this way, the coordination and efficiency of each leg movement can be intuitively seen. The graphical analysis tool processes the joint motion data through an algorithm and converts the data into graphs and charts to show the synchronization and differences in the movements between the left and right legs. This analysis helps athletes and coaches better understand athletic performance and guide training and adjustments. It not only provides specific directions for movement improvement, but also reveals potential injury risks, which helps to develop more effective training plans and generate leg motion analysis results.

[0036] See also Figure 4 , using the leg motion analysis results, using image recognition technology, analyzing the foot landing angle and stride changes, and determining the gait stability and deviation through the frequency and amplitude of the motion changes. The specific steps for outputting the gait stability analysis results are: S301: Based on the leg motion analysis results, a linear regression analysis method is used to quantitatively model the relationship between the foot landing angle and the stride length, and the execution process of constructing a statistical relationship model between the angle and the stride length is as follows; The formula for the linear regression analysis method is as follows: ; Where S is the stride length of each step, θ represents the foot angle when landing, k is the linear coefficient of the relationship between the landing angle and the stride length, b is the intercept of the linear model, Δθ is the difference between the landing angle of the current step and the landing angle of the previous step, and θ refis the reference base angle.

[0037] Detailed explanation of the formula and the process of formula calculation: In the formula, S represents the stride, which is the distance of each step taken by the athlete during the movement. The parameter θ represents the angle of the foot landing, which is obtained by actual measurement, for example, 30 degrees; k and b are the coefficient and intercept obtained by the linear regression model based on historical data, and the specific values ​​need to be determined through statistical analysis of historical sports data. Set k=0.5 and b=0.2 based on historical data analysis. The parameter Δθ represents the change in the landing angle between two adjacent steps, which is continuously recorded by the angle sensor during the movement, and this change is set to 5 degrees. θ ref It is the reference angle, which is used for normalizing angle changes. It is set to the average landing angle or a standard motion angle, which is 30 degrees here.

[0038] Using the above numerical values, a specific example is derived: 1. Calculation 2. Determine 3. Calculation 4. Calculation 5. Calculate the stride S: The results show that under the given foot landing angle and stride model parameters, the stride length of each step is about 15 meters. Through the numerical value, the stride efficiency and sports performance of the athlete can be further evaluated. By analyzing the actual sports data, this model helps to guide athletes to adjust their pace and improve their landing technique to optimize their sports performance.

[0039] S302: Calculate the frequency and amplitude of foot movement changes through the statistical relationship model between angle and stride length according to image recognition technology, evaluate the periodicity and repeatability of gait, adjust the sensitivity of frequency analysis to capture the change of gait, and generate the execution process of gait periodicity analysis results as follows; Calculate the changing frequency and amplitude of the foot movements, and calculate the gait periodicity analysis results according to the following formula: In the formula, f represents frequency, T represents cycle time, A represents amplitude, and x represents 13 ,y 13 and x 23 ,y 23 Represents the position coordinates of the foot movement at the beginning and end of the cycle. Detailed explanation of the formula and the derivation process of the formula calculation: Suppose the foot moves from (0,0) to (3, 4) in one cycle, the cycle time T=1s, and calculate the frequency and amplitude: The results showed that the amplitude of the foot motion was 5 units and the frequency was 1 Hz, which was used to evaluate the periodicity and repetitiveness of gait.

[0040] S303: using the gait periodicity analysis results, evaluating the stability and deviation of the gait, displaying the deviation of the gait through dynamic simulation, and identifying the influencing factors, the execution process of generating the gait stability analysis results is as follows; The results of gait periodic analysis are used to evaluate the stability and deviation of gait, and the deviation of gait is displayed through dynamic simulation, and the influencing factors are identified. During the analysis, graphical tools are used to dynamically display the gait simulation to intuitively reveal the stability and deviation of gait. This technical processing can simulate the performance of gait under different conditions through algorithms. The dynamic display can include the changes in gait at different speeds and under different ground conditions. The analysis also takes into account the impact of external factors such as uneven ground and sole wear on gait stability. Through this comprehensive evaluation, gait stability analysis results are generated.

[0041] See also Figure 5 Based on the gait stability analysis results, the leg parameters of the three-dimensional motion model are adjusted to simulate differentiated running motion sequences. The leg motions are analyzed through continuous motion simulation to obtain the motion symmetry evaluation results. The specific steps are as follows: S401: Input the gait stability analysis results into the running three-dimensional motion model, adjust the leg motion parameters, including the angle limit and dynamic characteristics of the joints, check the simulated running motion matching the real-time motion characteristics according to the gait deviation data, and generate the adjusted three-dimensional motion model. The execution process is as follows; Input the gait stability analysis results into the running 3D motion model, adjust the leg motion parameters, including the angle limit and dynamic characteristics of the joints, and conduct detailed checks based on the gait deviation data to ensure that the simulated running motion can match the real-time motion characteristics. The process includes analyzing the angle and dynamic data of each joint, adjusting the parameters to simulate various motion conditions that occur during actual running, and performing advanced simulations through software to ensure that each adjustment is based on actual gait analysis data. The optimization of motion parameters also considers the effects of muscle strength and joint flexibility in actual motion. These adjustments help the model more accurately reflect the actual performance of the athlete and generate an adjusted 3D motion model.

[0042] S402: Using the adjusted three-dimensional motion model, by simulating differentiated running conditions, including speed changes and ground types, recording and analyzing responses of leg motions, adjusting simulation frequency and motions, and generating differentiated running motion data, the execution process is as follows; By simulating differentiated running conditions, recording and analyzing the response of leg movements, differentiated running movement data is calculated according to the following formula: Where ΔR represents the spatial variation of the leg action response, and x1, y1, z1 and x2, y2, z2 represent the spatial coordinates of the leg action at the beginning and end of the simulation, respectively.

[0043] Detailed explanation of the formula and the process of formula calculation: Assume that the initial leg coordinates are (1, 2, 3), and after a certain running condition simulation, the coordinates become (4, 6, 2). Calculate the response change: The results showed that the variation in leg action response under different running conditions was 5.1 units, which was used to evaluate the accuracy of the simulation effect and the need for adjustment.

[0044] S403: Analyze the differentiated running motion data, evaluate the symmetry and coordination of the left and right leg motions, and generate the motion symmetry evaluation result by comparing the synchronization of the original and adjusted motions. The execution process is as follows; Analyze differentiated running action data, evaluate the symmetry and coordination of left and right leg movements, compare the synchronization of original and adjusted movements, record the changes of each leg movement in detail, analyze symmetry and coordination, use advanced analysis tools to quantify the synchronization of movements, and ensure that each adjusted movement can achieve the best exercise effect. This comparative analysis helps reveal the imbalance of left and right legs in movement. Through continuous adjustment and optimization, it can effectively improve exercise efficiency and reduce the risk of injury, provide valuable feedback to athletes, and guide them to pay attention to adjusting movements in training to achieve better sports performance and generate movement symmetry evaluation results.

[0045] See also Figure 6 ,According to the results of the movement symmetry evaluation, the user's running movement is compared with the preset ideal movement, the runner's movement deviation is detected through the set threshold, and potential movement defects are identified. The specific steps for obtaining the deviation analysis results are as follows: S501: extracting comparison data between the user's real-time running action and the preset ideal action through the action symmetry evaluation result, setting analysis thresholds to quantify the runner's deviations at key action points, including stride, foot angle, and knee bending degree, and generating action deviation quantification data in the following execution flow; Through the results of movement symmetry assessment, we extract the comparison data between the user's real-time running movement and the preset ideal movement, and set the analysis threshold to quantify the runner's deviation at key movement points. In the process, the setting of the analysis threshold is crucial, which determines the sensitivity and accuracy of identifying deviations, including key indicators such as stride, foot angle and knee bending degree. Through high-precision sensing technology and image analysis technology, we ensure that the data of each key movement point can be accurately captured and analyzed. Movement deviation quantification involves not only static measurement, but also dynamic changes. Through real-time tracking and historical data comparison, it provides a scientific basis for subsequent training, helping athletes and coaches understand and optimize running movements, so as to improve exercise efficiency and reduce the risk of injury, and generate movement deviation quantification data.

[0046] S502: Based on the action deviation quantification data, the volatility and deviation range in each action cycle are identified, the gap between the real-time action and the ideal model is recorded, and the execution process of generating deviation identification information is as follows; Based on the quantitative data of movement deviation, the volatility and deviation range in each movement cycle are identified, and the gap between the real-time movement and the ideal model is recorded. In the process, data analysis tools are used to conduct a detailed review of each link of the movement cycle. By comparing the real-time data with the ideal model, it can be clearly seen which movements deviate from the standard. Identifying these deviations can help athletes adjust their training plans to specifically improve movements that frequently deviate. Volatility analysis reveals the consistency and repeatability of movement execution, helps coaches and athletes understand the stability and optimization space of sports movements, and generates deviation identification information.

[0047] S503: using the deviation identification information to evaluate the runner's exercise efficiency and movement quality, identifying the user's potential running movement defects through multi-dimensional deviation analysis, and generating the deviation analysis results. The execution process is as follows; Deviation recognition information is used to evaluate the runner's movement efficiency and movement quality. The deviation analysis results are calculated according to the following formula: In the formula, E represents the motion efficiency, D a Represents the actual action deviation, D i Represents the ideal action standard. Detailed explanation of the formula and the process of formula calculation: Setting the ideal action standard D i is 0.1, while the actual action deviation D a is 0.2, then the motion efficiency is calculated as: The results show that exercise efficiency is directly affected by the deviation of movement quality. Through multi-dimensional deviation analysis, the user's potential running defects can be accurately identified, providing a scientific basis for further improving training methods and improving sports performance.

[0048] See also Figure 7 ,Using the deviation analysis results, analyzing the action data of the differentiation training phase, adjusting the camera angle and shooting speed, optimizing the data capture process, extracting the key change points of stride length and foot landing pressure, and obtaining the gait adjustment plan, the specific steps are: S601: Based on the deviation analysis results, the angle and position of the camera are adjusted to capture the running leg and foot movements of the user, and the shooting speed is adjusted to match the data capture requirements of the differentiated running speeds. The execution process of generating the optimized camera configuration is as follows; Based on the results of the deviation analysis, the angle and position of the camera are adjusted to capture the user's running leg and foot movements. During the process, the camera is adjusted to ensure that it can match the data capture requirements of differentiated running speeds. By optimizing the shooting speed and angle, every detail of the movement process can be recorded more accurately, which not only improves the accuracy of the data, but also provides higher quality visual data for subsequent motion analysis and training. These adjustments are based on the specific findings of the previous deviation analysis to ensure that the camera settings can maximize the capture of key movements, thereby helping athletes and coaches to conduct more effective training and improvements, and generate optimized camera configurations.

[0049] S602: Using the optimized camera configuration, the user's running motion data in the differentiated training stage is captured, the stride length and foot contact pressure are analyzed, and key change points are captured, including the maximum contact force and dynamic changes of the stride length. The execution process of generating key motion data is as follows; Utilizing the optimized camera configuration, the motion data of users running in the differentiated training phases is captured, with a focus on analyzing stride length and foot landing pressure. High-precision camera technology and pressure sensors are used to accurately record every foot landing and stride change during exercise. These data help identify key change points in exercise, including maximum landing force and dynamic changes in stride. Data collection and analysis are essential for understanding athletes' movement patterns and potential risks. The optimized equipment settings ensure high data quality and high temporal resolution, making it possible to track subtle changes in movements in real time for further analysis and training optimization, providing athletes with specific improvement guidance and generating key motion data.

[0050] S603: Based on the key action data, the changing trends of the stride length and the ground pressure are analyzed, the potential improvement effect after the action adjustment is predicted, and the action is adjusted and optimized through real-time feedback. The execution process of generating a gait adjustment plan is as follows; Analyze the changing trends of stride length and ground pressure, and predict the potential improvement effect after adjusting the action according to the following formula: In the formula, ΔS represents the change in stride length, S f and S i represent the stride length before and after adjustment, ΔP represents the change in ground pressure, and P f and P i Represent the landing pressure after and before adjustment respectively.

[0051] Detailed explanation of the formula and the process of formula calculation: Set the stride before adjustment to S i = 1.5 meters, the adjusted stride is S f =1.6 m, the ground pressure before adjustment is P i = 300 units, adjusted to P f = 280 units, calculate the change in stride length and ground pressure: These calculation results indicate that the stride length increases and the landing pressure decreases after adjustment, indicating that movement optimization improves running efficiency and reduces impact force, alleviating the burden on joints. These data provide a basis for formulating specific gait adjustment plans.

[0052] See also Figure 8 A running posture monitoring system based on image recognition technology is provided. The running posture monitoring system based on image recognition technology is used to execute the above-mentioned running posture monitoring method based on image recognition technology. The system includes: The multi-angle shooting module installs cameras at different angles to simultaneously capture the runner's full-body movements, record the action videos, and obtain full-body movement data; The motion model building module extracts joint angle changes from the whole body motion data, compares the movement amplitude and speed of the left and right leg joints, identifies the consistency of the motion, and obtains the leg motion analysis results; The stability assessment module uses the leg motion analysis results to analyze the changes in the foot landing angle and stride length, determine the stability and deviation of the gait, and output the gait stability analysis results; The motion simulation adjustment module simulates differentiated running motion sequences based on the gait stability analysis results, analyzes the symmetry of continuous motions, and obtains motion symmetry evaluation results; The training program optimization module compares the results of movement symmetry assessment with preset ideal movements, identifies movement deviations, adjusts the camera angle and shooting speed, and formulates a gait adjustment plan.

[0053] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A running posture monitoring method based on image recognition technology, characterized in that: The following steps are involved: Deploy cameras with different angles to synchronously record the user's full-body running movements, extract the joint angle change sequence, and build a three-dimensional running movement model; Based on the three-dimensional running motion model, the time series data of the left and right leg motions are analyzed, the joint motion amplitudes and speeds are compared, the consistency and differences of the motions are identified, and the leg motion analysis results are obtained; Utilizing the leg motion analysis results and image recognition technology, analyzing the foot contact angle and stride changes, determining the gait stability and deviation through the motion change frequency and amplitude, and outputting the gait stability analysis results; Based on the gait stability analysis results, the leg parameters of the three-dimensional motion model are adjusted to simulate differentiated running motion sequences, and the leg motions are analyzed through continuous motion simulation to obtain motion symmetry evaluation results; According to the movement symmetry evaluation result, the user's running movement is compared with the preset ideal movement, the runner's movement deviation is detected through a set threshold, potential movement defects are identified, and deviation analysis results are obtained; The deviation analysis results are used to analyze the motion data of the differentiation training stage, adjust the camera angle and shooting speed, optimize the data capture process, extract the key change points of stride length and foot landing pressure, and obtain a gait adjustment plan.

2. The running posture monitoring method based on image recognition technology according to claim 1, characterized in that: The three-dimensional running motion model includes three-dimensional angle data of joints, motion trajectories of multiple parts of the body and real-time motion rendering; the leg motion analysis results include motion synchronization scores, joint motion range differences and motion speed comparisons; the gait stability analysis results include step frequency stability metrics, statistical fluctuations in stride length and landing angle analysis; the motion symmetry assessment results include left and right leg motion consistency indexes, quantitative assessments of symmetry imbalances and visual comparisons of simulated motions; the deviation analysis results include numerical values ​​of joint motion deviations, motion defect types and risk level assessments; and the gait adjustment plan includes adjusted camera angle settings, data collection optimization points and key motion correction measures.

3. The running posture monitoring method based on image recognition technology according to claim 1, characterized in that: The steps of deploying cameras with different angles, synchronously recording the user's whole body running movements, extracting the joint angle change sequence, and constructing a 3D running movement model are as follows: Deploy multiple cameras on both sides and above the track to capture the user's full-body movements while running. Set the camera resolution and shooting frequency to optimize the clarity and data richness of motion capture, and generate full-body motion capture data. Based on the whole-body motion capture data, synchronously record the joint angles of the user at multiple moments while running, including angle changes of the knees, elbows, and shoulders, and set a data collection frequency to generate a joint angle change sequence; The joint angle change sequence is used to perform three-dimensional modeling to reflect the dynamic relationship and coordination of multiple parts of the body during running, optimize and adjust the model, and construct a three-dimensional running action model.

4. The running posture monitoring method based on image recognition technology according to claim 1, characterized in that: Based on the three-dimensional running motion model, the time series data of the left and right leg motions are analyzed, the joint motion amplitude and speed are compared, and the consistency and difference of the motions are identified to obtain the leg motion analysis results. Specifically, the steps are as follows: According to the three-dimensional running motion model, extract the time series data of the left and right legs, including the position and speed information of each joint in motion, set a time window to synchronously compare the data of the left and right legs, and generate the time series comparison data of the left and right legs; Based on the time series comparison data of the left and right legs, the movement amplitude and speed of each joint during running are calculated and compared, the movement consistency and difference of the legs on both sides are identified and evaluated, and the joint movement analysis results are generated; The joint motion analysis results are used to evaluate the coordination and efficiency of the runner's left and right legs, and graphical tools are used to display the synchronization and differences of the left and right leg movements to generate leg motion analysis results.

5. The running posture monitoring method based on image recognition technology according to claim 1, characterized in that: The leg motion analysis results are used to analyze the foot contact angle and stride changes using image recognition technology, and the gait stability and deviation are determined through the motion change frequency and amplitude. The steps of outputting the gait stability analysis results are specifically as follows: Based on the leg motion analysis results, a linear regression analysis method is used to quantitatively model the relationship between the foot landing angle and the stride length, and a statistical relationship model between the angle and the stride length is constructed; By using the statistical relationship model between the angle and the stride length and according to the image recognition technology, the changing frequency and amplitude of the foot movement are calculated, the periodicity and repeatability of the gait are evaluated, the sensitivity of the frequency analysis is adjusted to capture the changes in the gait, and the gait periodicity analysis results are generated; The gait periodicity analysis results are used to evaluate the stability and deviation of the gait, the deviation of the gait is displayed through dynamic simulation, and the influencing factors are identified to generate a gait stability analysis result.

6. The running posture monitoring method based on image recognition technology according to claim 5 is characterized in that: The formula of the linear regression analysis method is as follows: Where S is the stride length of each step, θ represents the foot angle when landing, k is the linear coefficient of the relationship between the landing angle and the stride length, b is the intercept of the linear model, Δθ is the difference between the landing angle of the current step and the landing angle of the previous step, and θ ref is the reference base angle.

7. The running posture monitoring method based on image recognition technology according to claim 1, characterized in that: Based on the gait stability analysis results, the leg parameters of the three-dimensional motion model are adjusted to simulate differentiated running motion sequences, and the leg motions are analyzed through continuous motion simulation to obtain the motion symmetry evaluation results. Specifically, the steps are as follows: Input the gait stability analysis results into the running three-dimensional motion model, adjust the leg motion parameters, including the angle limit and dynamic characteristics of the joints, check the simulated running motion matching the real-time motion characteristics according to the gait deviation data, and generate the adjusted three-dimensional motion model; Using the adjusted three-dimensional motion model, by simulating differentiated running conditions, including speed changes and ground types, recording and analyzing responses of leg motions, adjusting simulation frequency and motions, and generating differentiated running motion data; The differentiated running motion data is analyzed to evaluate the symmetry and coordination of the left and right leg motions, and a motion symmetry evaluation result is generated by comparing the synchronization of the original and adjusted motions.

8. The running posture monitoring method based on image recognition technology according to claim 1, characterized in that: According to the movement symmetry evaluation result, the user's running movement is compared with the preset ideal movement, the runner's movement deviation is detected by a set threshold, and potential movement defects are identified. The specific steps of obtaining the deviation analysis result are as follows: Extracting comparison data between the user's real-time running action and the preset ideal action through the action symmetry evaluation result, setting analysis thresholds to quantify the runner's deviations at key action points, including stride, foot angle, and knee flexion, and generating action deviation quantification data; Based on the action deviation quantification data, the volatility and deviation range in each action cycle are identified, the gap between the real-time action and the ideal model is recorded, and deviation identification information is generated; The deviation identification information is used to evaluate the runner's motion efficiency and movement quality, and through multi-dimensional deviation analysis, the user's potential motion defects in running are identified to generate deviation analysis results.

9. The running posture monitoring method based on image recognition technology according to claim 1, characterized in that: The steps of using the deviation analysis results to analyze the motion data in the differentiation training phase, adjusting the camera angle and shooting speed, optimizing the data capture process, extracting the key change points of stride length and foot landing pressure, and obtaining the gait adjustment plan are as follows: According to the deviation analysis results, the angle and position of the camera are adjusted to capture the movements of the user's running legs and feet, and the shooting speed is adjusted to match the data capture requirements of the differentiated running speed, so as to generate an optimized camera configuration; Using the optimized camera configuration, the user's running motion data in the differentiated training stage is captured, the stride length and foot landing pressure are analyzed intensively, key change points are captured, including the maximum landing force and dynamic changes of the stride length, and key motion data is generated; Based on the key action data, the changing trends of stride length and ground pressure are analyzed, the potential improvement effect after adjusting the action is predicted, the action is adjusted and optimized through real-time feedback, and a gait adjustment plan is generated.

Images