An interactive ball teaching method and system

Abstract

This application provides an interactive ball game teaching method and system, applied in the field of interactive teaching technology. By acquiring learners' movement data within a single training cycle, and after meeting basic evaluation standards, it further extracts indicator data reflecting advanced athletic abilities. If the indicator data meets the upper limit allowed by the laws of motion, the learner's movement is deemed to have reached the standard and they proceed to the next training session; otherwise, the learner is considered to have entered a pseudo-plateau period. During the pseudo-plateau period, the method can calculate the gap between the indicator data and the preset advanced target value, and adjust the basic evaluation standards based on this gap until the indicator data meets the upper limit standard. Furthermore, after each adjustment, the method can also pinpoint the root cause of the learner's movement deviation based on the gap and generate personalized training instructions. Therefore, this application has the advantage of significantly improving the intelligence and personalization level of interactive ball game teaching.

Description

Technical Field

[0001] This application relates to the field of interactive teaching technology, and in particular to an interactive ball game teaching method and system. Background Technology

[0002] In the field of interactive ball sports instruction, to achieve real-time evaluation of movements and personalized training, existing technologies typically establish a standard set of movement evaluation benchmarks initially, such as arm angle, force sequence, or body center of gravity stability in basketball shooting training. During daily training, the system captures the learner's movements and compares them with this initial benchmark in real time, providing feedback on whether the performance is satisfactory or needs improvement, thereby generating subsequent training priorities. This fixed evaluation logic is reasonable and effective in the introductory stage of physical education instruction.

[0003] However, during long-term training, as learners' skill levels improve, their movements gradually become highly standardized and fluid, with motion parameters stabilizing within the system's initially set excellent range. At this point, the logical judgment step responsible for comparing real-time motion data with the evaluation benchmark begins to reveal its limitations. Because the evaluation benchmark remains static and does not evolve in sync with the dynamic improvement of the learner's abilities, the system is still using a beginner-oriented tolerance for motion deviations for evaluation. This results in learners receiving the highest rating almost every time they shoot, making it impossible to distinguish between barely acceptable standard movements and excellent standard movements in the evaluation results.

[0004] Because each evaluation is positive feedback, the personalized training module may mistakenly assume that the learner has already perfected the skill and requires no further improvement. The generated training plan then stops at simple repetitive practice, failing to guide the learner towards higher-order dimensions such as improving speed or maintaining performance under pressure. Existing conventional methods often lack sensitivity to changes in skill stages during long-term training, easily overlooking the match between the evaluation benchmark and the learner's current actual ability, and failing to identify pseudo-plateaus caused by the system's evaluation logic. The specific technical problem arising from this is that during long-term training, when the learner's movements consistently reach and exceed the fixed initial evaluation benchmark's upper limit, the lack of dynamic adaptation of the movement evaluation logic to changes in skill level leads to a loss of discriminatory power in the evaluation results. This causes personalized training content to fail to match the learner's actual skill stage, thus interrupting the continuous evolution of skill status.

[0005] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0006] In view of the shortcomings of the prior art, this application provides an interactive ball sports teaching method and system, which aims to solve the problem that in the long-term training process, when the learner's movements stably reach and exceed the upper limit of the fixed initial evaluation benchmark, the evaluation results lose their discriminatory power due to the lack of dynamic adaptation of the movement evaluation logic to changes in skill level, making it impossible for personalized training content to match the learner's actual skill stage, thereby interrupting the continuous evolution of skill state.

[0007] Firstly, an interactive ball game teaching method, the method comprising the following steps: S1: Obtain the learner's action data within a single training cycle, and determine whether the learner's actions meet the basic evaluation criteria based on the action data; S2: When the basic evaluation criteria are met, extract the indicator data reflecting advanced motor ability from the motion data; S3: If the indicator data meets the upper limit standard allowed by the movement law, it is determined that the learner's movement has reached the standard, and the learner is guided to the next training exercise; S4: If the indicator data does not meet the upper limit standard allowed by the law of motion, it is determined that the learner has entered a pseudo plateau period caused by the low basic evaluation standard; S5: During the pseudo-plateau period, calculate the difference between the indicator data and the preset advanced target value, and adjust the basic evaluation standard according to the difference until the indicator data meets the upper limit standard allowed by the motion law. S6: After each adjustment of the basic evaluation criteria, the root cause of the learner's deviation in action is located based on the gap, and corresponding personalized training instructions are generated based on the root cause.

[0008] Furthermore, step S1 includes: S11: Acquire the learner’s action data in a single training cycle, and split the action data into multiple consecutive action frames according to the action data sampling points; S12: Calculate the score for each action frame, count the number of excellent actions whose scores exceed a preset excellent score threshold, and calculate the excellent action rate based on the proportion of the number of excellent actions in the total number of action frames. S13: Calculate the score variance based on the scores of all the action frames, and determine whether the score variance is less than a preset stability threshold. S14: When the rate of excellent actions exceeds the preset proportion and the score variance is less than the preset stability threshold, the learner's actions are determined to meet the basic evaluation criteria; otherwise, the learner's actions are determined to not meet the basic evaluation criteria.

[0009] Furthermore, in step S12, calculating the score for each action frame includes the following steps: S121: Extract the limb joint angles and body center of gravity position from each action frame as static action features; S122: Calculate the joint angular velocity as a dynamic motion feature based on the change in joint angle and the time interval between at least two consecutive action frames; S123: Compare the static motion features and the dynamic motion features with the corresponding standard joint angles, standard body center of gravity positions, and standard joint angular velocities in the basic evaluation criteria to obtain a matching score for each motion feature relative to the standard motion. S124: Sum the matching scores of all action features to obtain the score of the action frame.

[0010] Furthermore, step S2 includes: S21: Identify the start and end times of each action execution cycle from the action data, calculate the time difference between the end and start times of each action execution cycle as action execution speed data, and calculate the average value of all the action execution speed data as the action execution speed index. S22: Extract the peak force exertion time of each joint of the lower limb, trunk and upper limb in each action execution cycle from the action data, calculate the time interval between two adjacent peak force exertion times according to the force exertion sequence, compare each time interval with its corresponding standard time interval, obtain the force exertion continuity data of the corresponding action execution cycle according to the degree of deviation obtained by comparison, and calculate the average value of the force exertion continuity data of all action cycles as the whole body force exertion continuity index. S23: Divide multiple consecutive action execution cycles into an early cycle group and a late cycle group, calculate the degree of deviation of the action execution speed data or force continuity data of the late cycle group relative to the early cycle group, and use it as an indicator of action quality change. The index data includes at least the action execution speed index, the whole-body force exertion continuity index, and the action quality change index.

[0011] Furthermore, in step S22, the peak force exertion times of each joint of the lower limb, trunk, and upper limb are respectively the first time corresponding to the peak value of the knee joint extension angular velocity, the second time corresponding to the peak value of the hip joint extension angular velocity, the third time corresponding to the peak value of the shoulder joint flexion angular velocity, and the fourth time corresponding to the peak value of the elbow joint extension angular velocity. Calculating the time interval between two adjacent peak force moments according to the force exertion sequence includes the following steps: S221: Calculate the first time interval between the first moment and the second moment according to the force exertion sequence; S222: Calculate the second time interval between the second time moment and the third time moment; S223: Calculate the third time interval between the third time moment and the fourth time moment.

[0012] Furthermore, step S3 includes: S31: Obtain the sequence of the learner's action execution speed index in multiple consecutive training cycles, calculate the first improvement range of the action execution speed index between adjacent training cycles, and when the first improvement range is lower than the preset first progress threshold for a preset number of consecutive times, and the action execution speed index reaches the preset progress target value in the current training cycle, determine that the action execution speed index meets the upper limit standard allowed by the law of motion. S32: Obtain the sequence of the learner's whole-body force coherence index in multiple consecutive training cycles, calculate the second improvement range of the whole-body force coherence index between adjacent training cycles, and when the second improvement range is lower than the preset second progress threshold for a preset number of consecutive times, and the whole-body force coherence index reaches the preset advancement target value in the current training cycle, determine that the whole-body force coherence index meets the upper limit standard allowed by the law of motion. S33: Obtain the sequence of movement quality change indicators of the learner in multiple consecutive training cycles. When the fluctuation value of the movement quality change indicator sequence is lower than the preset quality stability threshold, determine that the movement quality change indicator meets the upper limit standard allowed by the law of motion. S34: When the above three conditions are met simultaneously, the learner's action is determined to be up to standard, and the learner is guided to the next training exercise.

[0013] Furthermore, step S4 includes: S41: When the improvement rate of a preset number of consecutive first improvements is lower than the preset first progress threshold, but the action execution speed index in the current training cycle has not reached the preset advancement target value, Or, if the improvement rate of the second step is lower than the preset second progress threshold for a preset number of consecutive times, but the whole-body force coherence index does not reach the preset advancement target value in the current training cycle, Alternatively, if the fluctuation value of the sequence of action quality change indicators is lower than the preset quality stability threshold, but the action execution speed indicator or the whole-body force coherence indicator does not reach the corresponding preset advanced target value, it is determined that the learner has entered a pseudo-plateau period caused by the low basic evaluation standard.

[0014] Furthermore, step S5 includes: S51: Obtain the absolute difference between the action execution speed index, the whole body force coherence index and the corresponding preset advanced target value in the index data, as the gap; S52: Based on the size of the corresponding gap, the qualified judgment interval of the corresponding action feature in the basic evaluation standard is narrowed towards the direction of the corresponding advanced target value to obtain the adjusted basic evaluation standard; S53: The matching degree scores of the action execution speed index and the whole body force continuity index in the index data with the corresponding indicators in the preset advanced target values ​​are added as new scoring items and added to the scoring calculation process of the action frame. S54: Under the adjusted basic evaluation criteria, repeat steps S1 to S5 until the index data meets the upper limit standard allowed by the motion law.

[0015] Furthermore, step S6 includes: S61: After each adjustment of the basic evaluation criteria, the advanced motor ability with the largest gap is identified as the weak link; S62: Based on the causal relationship between preset action features, reverse the process from the weak link to find the fundamental action feature that causes the weak link; wherein, the fundamental action feature that causes the weak link is the root cause of the deviation in the learner's action. S63: Generate personalized training instructions for the fundamental motion characteristics.

[0016] Secondly, an interactive ball sports teaching system, the system being used to implement the steps of any of the methods described above, the system comprising: Acquisition module: Acquires the learner's action data within a single training cycle, and determines whether the learner's actions meet the basic evaluation criteria based on the action data; Extraction module: When the basic evaluation criteria are met, extract the indicator data reflecting advanced motor ability from the action data; Judgment module: If the indicator data meets the upper limit standard allowed by the movement law, the learner's movement is judged to have met the standard, and the learner is guided to the next training exercise; Judgment module: If the indicator data does not meet the upper limit standard allowed by the movement law, it is determined that the learner has entered a pseudo plateau period caused by the low basic evaluation standard; Adjustment module: During the pseudo-plateau period, calculate the difference between the indicator data and the preset advanced target value, and adjust the basic evaluation standard according to the difference until the indicator data meets the upper limit standard allowed by the motion law; Training guidance module: After each adjustment of the basic evaluation criteria, the module identifies the root cause of the learner's deviation in action based on the gap, and generates corresponding personalized training instructions based on the root cause.

[0017] Beneficial Effects: The interactive ball sports teaching method and system proposed in this application effectively solves the problem of fixed evaluation benchmarks in existing technologies, which cannot adapt to the dynamic improvement of learners' skill levels, by dynamically adjusting the basic evaluation criteria. Specifically, the method first acquires the learner's movement data within a single training cycle and determines whether it meets the basic evaluation criteria. After meeting the basic evaluation criteria, it further extracts indicator data reflecting advanced athletic abilities. If the indicator data meets the upper limit allowed by the laws of motion, the learner's movement is deemed to have reached the standard and proceeds to the next training exercise; if not, the learner is deemed to have entered a pseudo-plateau period caused by excessively low basic evaluation criteria. During the pseudo-plateau period, the method can calculate the gap between the indicator data and the preset advanced target value, and adjust the basic evaluation criteria according to the gap until the indicator data meets the upper limit standard. Furthermore, after each adjustment, the method can also locate the root cause of the learner's movement deviation based on the gap and generate personalized training instructions. Therefore, this application can effectively solve the problem that in the long-term training process, when the learner's movements stably reach and exceed the upper limit of the fixed initial evaluation benchmark, the lack of dynamic adaptation of the movement evaluation logic to changes in skill level leads to the loss of differentiation in the evaluation results, making it impossible for personalized training content to match the learner's actual skill stage, thus interrupting the continuous evolution of skill state, and significantly improving the intelligence and personalization level of interactive ball sports teaching. Attached Figure Description

[0018] Figure 1 This is a flowchart of an interactive ball game teaching method proposed in this application.

[0019] Figure 2 This is a structural diagram of an interactive ball sports teaching system proposed in this application.

[0020] Labeling Explanation: 201. Acquisition Module; 202. Extraction Module; 203. Judgment Module; 204. Judgment Module; 205. Adjustment Module; 206. Training Guidance Module. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and marked in the accompanying drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0022] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0023] Traditional interactive ball sports teaching methods suffer from a lack of differentiation in evaluation results as learners' skill levels improve due to the static nature of the evaluation benchmarks. This makes personalized training content unable to match the learner's actual skill stage, thus interrupting the continuous evolution of skill status. For example, in long-term ball sports training scenarios, when a learner's movements consistently reach and exceed a fixed initial evaluation benchmark upper limit, the system still uses the beginner-level tolerance for movement deviations for evaluation. This results in learners receiving the highest evaluation almost every time they perform training movements, making it impossible to distinguish between barely acceptable standard movements and excellent standard movements. If this problem is not addressed, the personalized training module may mistakenly assume that the learner has already perfected the skill and no longer needs improvement. The generated training plan then stops at simple repetitive exercises and fails to guide higher-order dimensions such as improving release speed or maintaining movements under pressure.

[0024] Please refer to Figure 1 An interactive ball game teaching method, the method includes the following steps: S1: Obtain the learner's motion data within a single training cycle, and determine whether the learner's motion has met the basic evaluation criteria based on the motion data; S2: When the basic evaluation criteria are met, extract the indicator data reflecting advanced motor ability from the motion data; S3: If the indicator data meets the upper limit standard allowed by the law of motion, the learner's movement is judged to have met the standard, and the learner is guided to the next training exercise. S4: If the indicator data does not meet the upper limit standard allowed by the law of motion, it is judged that the learner has entered a pseudo plateau period caused by the low basic evaluation standard. S5: During the pseudo-plateau period, calculate the difference between the indicator data and the preset advanced target value, and adjust the basic evaluation standard according to the difference until the indicator data meets the upper limit standard allowed by the law of motion. S6: After each adjustment to the basic evaluation criteria, identify the root cause of the learner's deviation in action based on the gap, and generate corresponding personalized training instructions based on the root cause.

[0025] This application effectively solves the problem caused by the static evaluation benchmark in the prior art by dynamically adjusting the basic evaluation criteria, so that the teaching system can continuously adapt to the improvement of learners' skill level, avoid the occurrence of pseudo-plateau period, and thus ensure the continuity and effectiveness of training.

[0026] This application provides an interactive ball game teaching method that aims to help learners break through skill bottlenecks and achieve continuous progress by dynamically adjusting evaluation criteria and personalized training instructions.

[0027] Among them, motion data refers to information such as the learner's body posture, movement trajectory, joint angles, and speed when performing ball sports, collected through devices such as sensors and cameras.

[0028] The basic evaluation criteria are action specifications set for learners in the introductory stage, used to determine whether learners have mastered the basic action techniques.

[0029] Advanced motor skills refer to the higher level of ability that learners demonstrate in terms of movement efficiency, strength, and coordination after mastering basic movements.

[0030] Indicator data are specific numerical values ​​that quantify advanced athletic abilities, such as speed of movement execution and continuity of force exertion.

[0031] The upper limit of the standard allowed by the laws of motion refers to the best or near-best performance that a learner can achieve after training in a specific ball sport, based on the laws of motion.

[0032] A pseudo-plateau refers to a stage where learners feel their progress has stalled during training, but this is actually due to the evaluation criteria not being updated in a timely manner, causing the system to be unable to recognize their true progress and thus unable to provide higher-level training guidance.

[0033] Advanced target values ​​are higher-level target values ​​set for specific advanced athletic abilities. They can be obtained from statistical data of the reference population based on the learner's age, gender, and years of training, or they can be manually set by the coach according to the training goals.

[0034] The gap refers to the quantitative difference between a learner's indicator data and their advanced target value.

[0035] The root cause refers to the fundamental movement characteristics that lead to deviations in the learner's movements, such as insufficient range of motion of a certain joint or improper force application sequence.

[0036] Personalized training instructions are customized training suggestions generated based on the learner's specific problems and weaknesses.

[0037] In step S1, it is necessary to acquire the learner's motion data within a single training cycle and determine whether the learner's movements meet the basic evaluation criteria based on this data. Motion data acquisition can be achieved in various ways. For example, a high-speed camera can be used to capture full-body images of the learner during training, and then image processing techniques can be used to extract the position and posture information of key skeletal points. Another method is to have the learner wear inertial measurement unit (IMU) sensors. These sensors can record data such as angular velocity and acceleration of various parts of the learner's body in real time, thereby constructing complete motion data. After acquiring the motion data, the system compares it with preset basic evaluation criteria. The basic evaluation criteria can be a series of thresholds or models regarding motion posture, trajectory, and timing.

[0038] For example, in basketball shooting training, basic evaluation criteria might include the range of elbow flexion angle, wrist rotation angle, and body center of gravity stability during the shot. The system analyzes the learner's motion data based on these criteria to determine whether they have met the basic requirements. If the learner's motion data falls within all basic evaluation criteria, their motion is considered to meet the basic evaluation criteria; otherwise, it is considered not to meet them. It is understood that this applies to other ball sports besides basketball shooting training, including but not limited to badminton, football, and tennis. For ease of understanding, this application uses basketball training as an example.

[0039] In step S2, when the learner's movements meet the basic evaluation criteria, the system extracts indicator data reflecting advanced motor abilities from the movement data. The extraction methods for advanced motor ability indicators can be diverse. For example, the time required for the learner to complete one full movement can be calculated and used as an indicator of movement execution speed. Shorter time generally indicates more efficient and fluid movements. Another approach is to analyze the sequence and peak force of each joint during the movement to assess the continuity of force exertion. For example, in shooting, if the force exertion of the lower limbs, torso, and upper limbs forms a smooth chain, the continuity of force exertion is good. Furthermore, the stability or trend of movement quality can be assessed by comparing the learner's performance over multiple consecutive movement cycles. For example, the fluctuation range of the release angle in consecutive shooting movements can be calculated; smaller fluctuations indicate more stable movements. These indicator data can more precisely reflect the learner's higher-level motor abilities beyond basic movements.

[0040] In step S3, the system determines whether the extracted indicator data meets the upper limit standard allowed by the movement laws. If it does, the learner's movement is deemed to have met the standard, and the learner is guided to the next training exercise. Determining whether the indicator data meets the upper limit standard can be done through preset thresholds or models. For example, for the movement execution speed indicator, a theoretically fastest completion time can be set as the upper limit standard. If the learner's average movement execution speed reaches this time, they are considered to have met the standard for that indicator. For the force application continuity indicator, an ideal force application timing model can be set. If the learner's force application timing closely matches the model, their force application continuity is considered to have met the standard. When all key advanced movement ability indicators have reached their respective upper limit standards, the system considers the learner to have reached a high level in that skill and can proceed to higher-level training content, such as transitioning from stationary shooting to moving shooting, or from non-contact training to contact training.

[0041] In step S4, if the extracted indicator data does not meet the upper limit allowed by the laws of motion, the system will determine whether the learner has entered a pseudo-plateau period caused by excessively low basic evaluation standards. Determining whether a pseudo-plateau period has occurred typically requires comprehensive consideration of the learner's training performance over a period of time. For example, if, over several consecutive training cycles, the learner's improvement in advanced motor ability indicators (such as movement execution speed or force continuity) continues to decrease or even stagnates, while their pass rate for basic movements remains consistently high, this may indicate that the learner has entered a pseudo-plateau period. In this case, the system will identify this contradictory phenomenon—that is, the learner performs well in basic movements but their advanced abilities improve slowly—and thus determine that they are in a pseudo-plateau period.

[0042] In step S5, when the learner is determined to have entered a pseudo-plateau, the system calculates the difference between the indicator data and the preset advanced target value, and adjusts the basic evaluation criteria based on this difference until the indicator data meets the upper limit allowed by the laws of motion. The difference can be calculated by simple numerical subtraction; for example, if the advanced target value is a movement execution speed of 0.5 seconds, and the learner's current speed is 0.7 seconds, the difference is 0.2 seconds. Based on this difference, the system dynamically narrows the acceptable range for the corresponding movement characteristics in the basic evaluation criteria. For example, if the acceptable range for the elbow flexion angle was previously 90-120 degrees, to encourage the learner to improve movement efficiency, the system might adjust the acceptable range to 95-115 degrees, thereby increasing the requirement for movement precision. This adjustment means that the learner will need to exert more effort to reach the acceptable level under the new, more stringent basic evaluation criteria. The system repeats steps S1 to S5 until the learner's advanced motor ability indicators meet the upper limit allowed by the laws of motion.

[0043] In step S6, after each adjustment to the basic evaluation criteria, the system identifies the root cause of the learner's movement deviation based on the calculated gap and generates corresponding personalized training instructions. Identifying the root cause can be achieved by analyzing the advanced motor ability indicators with the largest gaps. For example, if the difference in movement execution speed is the largest, the system will further analyze which aspect is causing the speed deficiency; it could be insufficient lower limb strength or poor upper limb coordination. Through the causal relationships between preset movement characteristics, the system can trace back to the fundamental movement characteristic causing the weakness. For example, if insufficient shoulder flexion speed is found to cause slow shooting speed, then insufficient shoulder flexion speed is the root cause. Once the root cause is found, the system generates targeted personalized training instructions. For example, for insufficient shoulder flexion speed, the system may suggest specific shoulder strength or flexibility training, or provide more refined movement guidance to help learners improve their shoulder joint movement patterns.

[0044] The interactive ball sports teaching method provided in this application works by constructing a dynamic evaluation and training loop that adapts to the learner's skill level, effectively solving the pseudo-plateau problem caused by the static evaluation benchmark in traditional teaching methods.

[0045] In summary, this application establishes a closed-loop, adaptive teaching system by dynamically adjusting basic evaluation criteria, identifying pseudo-plateau periods, pinpointing the root causes of movements, and generating personalized training instructions. This system continuously adapts to the learner's skill level improvement, effectively avoiding the pseudo-plateau problem caused by static evaluation benchmarks in traditional teaching methods. It ensures the continuity and effectiveness of training, thereby significantly improving the effectiveness of interactive ball sports instruction.

[0046] This application presents an interactive ball sports teaching method that addresses the problems in existing technologies, such as static evaluation benchmarks leading to a loss of discriminative power in evaluation results and the inability of personalized training to match learners' actual skill levels. The core innovation lies in the introduction of a "pseudo-plateau" judgment mechanism and a method for dynamically adjusting the basic evaluation criteria.

[0047] Compared to the closest existing technology, the advantages of this application are: First, existing technologies typically employ a fixed initial evaluation benchmark. When a learner's skill level reaches or exceeds the upper limit of this benchmark, the evaluation results lose their discriminatory power and fail to identify the learner's progress in higher-order dimensions. This application, through steps S1 to S3, not only determines whether the learner meets the basic evaluation criteria but also further extracts indicator data reflecting advanced motor abilities and determines whether it conforms to the upper limit standards allowed by the laws of motor development. This hierarchical evaluation mechanism enables the system to more comprehensively and precisely assess the learner's true skill level.

[0048] Secondly, existing technologies lack sensitivity to skill stage changes during long-term training, easily overlooking the matching relationship between evaluation benchmarks and current actual abilities, and struggling to identify pseudo-plateau periods caused by system evaluation logic. This application, through step S4, explicitly proposes the crucial step of "determining whether the learner has entered a pseudo-plateau period caused by excessively low basic evaluation standards." This mechanism enables the system to proactively identify the contradictory phenomenon of learners performing well in basic movements but showing slow progress in advanced abilities, thereby avoiding training stagnation caused by evaluation system failure.

[0049] Furthermore, existing technologies often fail to provide effective solutions after identifying a pseudo-plateau, resulting in personalized training content that cannot match the learner's actual skill level. This application provides a complete solution through steps S5 and S6. Step S5 calculates the gap between the indicator data and the preset advancement target value, and dynamically adjusts the basic evaluation criteria based on this gap, allowing the evaluation criteria to evolve synchronously with the learner's ability improvement. This dynamic adjustment mechanism forces learners to continuously improve under stricter standards, thereby breaking through skill bottlenecks. Step S6 further identifies the root cause of the learner's deviation in action based on the adjusted gap, and generates corresponding personalized training instructions based on the root cause, ensuring the targeting and effectiveness of the training.

[0050] Through the aforementioned innovations, the method of this application can continuously adapt to the improvement of learners' skill levels, effectively avoiding the occurrence of pseudo-plateaus and thus ensuring the continuity and effectiveness of training. This enables the teaching system to provide learners with truly personalized, progressive training guidance, significantly improving the effectiveness of interactive ball sports teaching and achieving continuous evolution of learners' skill status.

[0051] In some embodiments of this application, the specific implementation process of step S1, which involves acquiring the learner's action data within a single training cycle and determining whether the learner's actions meet the basic evaluation criteria based on the action data, can be further refined.

[0052] Furthermore, step S1 includes: S11: Acquire the learner's action data within a single training cycle and split the action data into multiple consecutive action frames according to the action data sampling points; S12: Calculate the score for each action frame, count the number of excellent actions whose scores exceed the preset excellent score threshold, and calculate the excellent action rate based on the proportion of excellent actions in the total number of action frames. S13: Calculate the score variance based on the scores of all action frames, and determine whether the score variance is less than the preset stability threshold. S14: When the rate of excellent actions exceeds the preset proportion and the score variance is less than the preset stability threshold, the learner's actions are judged to meet the basic evaluation criteria; otherwise, the learner's actions are judged not to meet the basic evaluation criteria.

[0053] Specifically, in step S11, the learner's motion data within a single training cycle can be acquired using various sensors (e.g., inertial measurement units, optical motion capture systems, depth cameras, etc.). This motion data is typically time-series data, containing information such as limb joint positions, angles, and speeds. Dividing the motion data into multiple consecutive motion frames based on motion data sampling points means segmenting the continuous motion data stream into discrete time segments, where each motion frame represents the learner's posture or state within a very short period. For example, it can be divided at a frequency of 30 or 60 frames per second.

[0054] In step S12, a score is calculated for each action frame to quantify the degree of matching between each instantaneous action and a standard action. The score can be calculated based on a preset scoring model or expert experience. Counting the number of excellent actions whose scores exceed a preset excellent score threshold refers to identifying those action frames with high execution quality. The excellent score threshold can be set according to teaching objectives and action difficulty. Based on the proportion of excellent actions in the total number of action frames, an excellent action rate is calculated, which reflects the overall quality level of the learner's action execution throughout the entire training period.

[0055] In practical applications, step S13 calculates the score variance based on the scores of all the action frames. The purpose of this calculation is to assess the stability of the learner's action execution. A smaller score variance indicates less fluctuation in the learner's action quality across different action frames, and more stable action execution. The preset stability threshold can be set according to the characteristics of different ball sports and the requirements of the teaching stage. For example, for beginners, the stability threshold can be relatively lenient, while for advanced learners, a stricter stability threshold is required.

[0056] Furthermore, in step S14, determining whether the learner's actions meet the basic evaluation criteria is achieved by comprehensively considering the rate of excellent actions and the variance of the rating. A rate of excellent actions exceeding a preset proportion indicates that the learner can perform high-quality actions most of the time; a variance of the rating is less than a preset stability threshold indicates that the learner's action execution has good stability. Only when both conditions are met simultaneously can the learner's actions be considered to have met the basic evaluation criteria; otherwise, their actions are considered not to meet the basic evaluation criteria.

[0057] This application's solution refines the learner's continuous motion data into discrete motion frames and scores each frame, enabling a precise assessment of the learner's performance within a single training cycle. By calculating the excellent motion rate, the overall quality of the learner's movements can be quantified, avoiding the bias of judging based solely on a single movement or localized performance. Simultaneously, introducing scoring variance as an indicator of motion stability effectively identifies fluctuations in the learner's motion execution, ensuring that their movements are not only of high quality but also consistently output. Therefore, by combining the excellent motion rate and scoring variance as a dual judgment mechanism, this application can more comprehensively and accurately determine whether the learner's movements meet the basic evaluation criteria, providing a reliable basis for subsequent advanced training or problem diagnosis.

[0058] The above technical solution overcomes the potential crudeness and inaccuracy of traditional methods in judging whether learners' movements meet basic evaluation standards. By performing frame-level segmentation and scoring of movement data, and comprehensively considering the rate of excellent movements and the variance of scores, this application provides a more refined, objective, and comprehensive basic movement assessment mechanism. This not only improves the accuracy and reliability of the assessment but also enables the earlier detection of potential problems in learners' movements, laying a solid foundation for subsequent personalized guidance and training adjustments, thereby effectively improving teaching efficiency and learning outcomes.

[0059] Further, in step S12, calculating the score for each action frame includes the following steps: S121: Extract the limb joint angles and body center of gravity position in each action frame as static action features; S122: Calculate the joint angular velocity as a dynamic motion feature based on the change in joint angle and the time interval between at least two consecutive action frames; S123: Compare the static and dynamic motion features with the corresponding standard joint angles, standard body center of gravity positions, and standard joint angular velocities in the basic evaluation criteria to obtain a matching score for each motion feature relative to the standard motion. S124: Sum the matching scores of all action features to obtain the score of the action frame.

[0060] Specifically, in step S121, limb joint angles can be understood as the degree of flexion or extension of key joints such as the knee, elbow, and shoulder joints of the learner in a specific action frame. The body's center of gravity position can refer to the learner's center of mass coordinates in three-dimensional space. These features collectively constitute the learner's posture information at a given moment, forming the basis for evaluating the static accuracy of the movement. For example, during a racket swing, the angles of the wrist, elbow, and shoulder, as well as the distribution of the body's center of gravity, are all important static movement features.

[0061] In step S122, joint angular velocity refers to the change in angle of a joint's rotation per unit time. It is calculated by acquiring angle data of the same joint in at least two consecutive action frames, calculating the change between these angles, and dividing it by the corresponding time interval. As a dynamic movement characteristic, joint angular velocity reflects the fluidity, explosiveness, and efficiency of force transmission in a learner's movements. For example, in ball sports, the joint angular velocity at the moment of impact has a decisive influence on ball speed and impact quality.

[0062] Further, in step S123, the extracted static and dynamic action features are compared with the corresponding standard values ​​in the preset basic evaluation criteria. This comparison can employ various mathematical methods, such as calculating the Euclidean distance between the feature value and the standard value, cosine similarity, or setting an allowable error range. Through this comparison, the degree of conformity between each action feature and the standard action can be quantified, thereby obtaining a matching score for each action feature relative to the standard action. The higher the matching score, the closer the learner's action is to the standard action for that specific feature.

[0063] Therefore, in step S124, the comprehensive score of the action frame can be obtained by summing the matching scores of all action features. This comprehensive score fully reflects the overall action quality of the learner within a single action frame, providing a quantitative basis for the subsequent calculation of the excellent action rate and score variance.

[0064] This application's solution refines the scoring of each action frame, decomposing action features into static and dynamic features, and comparing them separately with basic evaluation criteria to obtain a matching score. This meticulous scoring mechanism makes the assessment of learner actions more comprehensive and accurate. By comprehensively considering static features such as limb joint angles and body center of gravity position, as well as dynamic features such as joint angular velocity, it can more accurately capture the details and quality of learner actions, avoiding errors that may result from judging based on a single or coarse indicator. This precise scoring is the basis for calculating the excellent action rate and score variance, thereby improving the accuracy and reliability of judging whether learner actions meet the basic evaluation criteria.

[0065] Through the above technical solution, this application provides a more refined and comprehensive motion frame scoring method. This method considers not only the learner's static posture but also the dynamic process of the movement, making the assessment of learner movement quality more objective and accurate. Therefore, when judging whether a learner's movement meets the basic evaluation criteria, it can be based on more reliable scoring data, thereby improving the accuracy of the judgment, effectively avoiding misjudgments, and providing stronger data support for subsequent training phases. This precise assessment helps coaches or systems identify potential problems in learner movements earlier and provides a more accurate basis for generating subsequent personalized training instructions.

[0066] Furthermore, step S2 includes: S21: Identify the start and end times of each action execution cycle from the action data, calculate the time difference between the end and start times of each action execution cycle as action execution speed data, and calculate the average value of all action execution speed data as the action execution speed index. S22: Extract the peak force exertion time of each joint of the lower limb, trunk and upper limb in each action execution cycle from the action data, calculate the time interval between two adjacent peak force exertion times according to the force exertion sequence, compare each time interval with its corresponding standard time interval, obtain the force exertion continuity data of the corresponding action execution cycle according to the degree of deviation obtained by comparison, and calculate the average value of the force exertion continuity data of all action cycles as the whole body force exertion continuity index. S23: Divide multiple consecutive action execution cycles into an early cycle group and a late cycle group, calculate the degree of deviation of the action execution speed data or force continuity data of the late cycle group relative to the early cycle group, and use it as an indicator of action quality change. The indicators should include at least the speed of movement execution, the continuity of whole-body force exertion, and the change in movement quality.

[0067] In step S21, the start and finish times of the action execution cycle can be identified based on when a learner's specific limb joints (e.g., the wrist joint in a racket swing, the shoulder joint in a throwing motion) reach a preset position or speed threshold during ball sports. For example, in a tennis serve, the start time can be considered when the racket reaches its highest point, and the finish time can be considered when the racket hits the ball. By calculating the time difference between the finish and start times, the time required for the learner to complete one action can be quantified, thus obtaining action execution speed data. Averaging all action execution speed data yields the learner's average action execution speed index within the current training cycle, which directly reflects the learner's explosive power and efficiency.

[0068] Further, in step S22, the peak force moment refers to the instant during which the joints of the lower limbs, trunk, and upper limbs generate maximum force or maximum angular velocity during movement within one action cycle. The selection of these four moments is based on the typical bottom-up kinetic chain transmission principle in ball sports. These peak moments can be obtained by analyzing motion data captured by wearable sensors (such as inertial measurement units, IMUs) or high-speed cameras. For example, in throwing sports, specific peak force moments exist for the lower limbs' push-off, hip rotation, trunk twisting, and upper limb arm swing. Following a preset force sequence (e.g., kinetic chain transmission from the lower limbs to the trunk and then to the upper limbs), the time interval between two adjacent peak force moments is calculated and compared with the force interval of the standard movement. The degree of deviation obtained through comparison can assess the smoothness and coordination of the learner's force sequence, thus obtaining force continuity data. Averaging the force continuity data of all movement cycles yields a whole-body force continuity index, which reflects the learner's overall movement coordination and kinetic chain transmission efficiency.

[0069] Furthermore, in step S23, multiple consecutive action execution cycles can be divided into an early cycle group and a late cycle group. For example, the first 5 action cycles can be defined as the early cycle group, and the subsequent 5 action cycles as the late cycle group. By calculating the degree of deviation of the action execution speed data or force continuity data of the late cycle group relative to the early cycle group, the stability of the learner's movement quality under continuous training or fatigue can be assessed. For example, if the action execution speed of the late cycle group decreases significantly, or the deviation in force continuity increases, it indicates that the learner's movement quality has declined during long-term training, which can serve as an indicator of movement quality change. This indicator reflects the learner's endurance, movement stability, and ability to maintain technical movements under fatigue.

[0070] This application's solution extracts indicators such as movement execution speed, whole-body power continuity, and movement quality variation from motion data, enabling a comprehensive and multi-dimensional assessment of learners' advanced athletic abilities in ball sports. Movement execution speed directly reflects the learner's explosive power and efficiency, a key factor in measuring athletic performance. Whole-body power continuity reveals the degree of coordination among different body parts when performing complex movements, significantly contributing to improved efficiency and reduced injury. Movement quality variation further evaluates the learner's ability to maintain high-level movement quality during sustained training or fatigue, crucial for consistent performance in competitive sports. By comprehensively analyzing these indicators, it's possible to more accurately determine whether learners have truly reached an advanced level, rather than merely achieving basic movement proficiency.

[0071] The aforementioned technical solution provides a more refined and comprehensive assessment of advanced motor skills, overcoming the limitations of relying solely on basic evaluation standards. This solution effectively identifies learners' strengths and weaknesses in movement speed, force coordination, and movement stability, providing more precise data support for subsequent personalized training. This helps learners overcome technical bottlenecks, achieve continuous improvement in athletic performance, and effectively avoid pseudo-plateaus caused by excessively low basic evaluation standards, thereby enhancing the scientific rigor and effectiveness of teaching.

[0072] Furthermore, in step S22, the peak force exertion times of each joint of the lower limb, trunk, and upper limb are respectively the first time corresponding to the peak value of the knee joint extension angular velocity, the second time corresponding to the peak value of the hip joint extension angular velocity, the third time corresponding to the peak value of the shoulder joint flexion angular velocity, and the fourth time corresponding to the peak value of the elbow joint extension angular velocity. Calculating the time interval between two adjacent peak force moments according to the force exertion sequence includes the following steps: S221: Calculate the first time interval between the first moment and the second moment according to the force exertion sequence; S222: Calculate the second time interval between the second and third time points; S223: Calculate the third time interval between the third and fourth time points.

[0073] Specifically, in ball sports, such as tennis serves or golf swings, the generation of force typically follows a chain-like transmission process from the lower limbs to the trunk and then to the upper limbs. To accurately capture the continuity of this force chain, this application defines the peak force moment as the point in time when the angular velocity of a specific joint reaches its peak. Specifically, the first moment corresponds to the peak angular velocity of the knee joint extension, representing the critical moment of lower limb push-off; the second moment corresponds to the peak angular velocity of the hip joint extension, representing the critical moment of hip rotation; the third moment corresponds to the peak angular velocity of the shoulder joint flexion, representing the critical moment of trunk-driven shoulder force generation; and the fourth moment corresponds to the peak angular velocity of the elbow joint extension, representing the critical moment of acceleration at the end of the arm. The selection of these moments is based on a deep understanding of the biomechanical analysis of ball sports and effectively reflects the energy transfer efficiency from the ground to the racket / club.

[0074] Furthermore, when calculating the time interval between two adjacent peak force moments, the first time interval between the first and second moments is calculated first, reflecting the smoothness of the connection between lower limb and hip force generation. Next, the second time interval between the second and third moments is calculated, reflecting the smoothness of the connection between hip and shoulder force generation. Finally, the third time interval between the third and fourth moments is calculated, reflecting the smoothness of the connection between shoulder and elbow force generation. Through precise calculation of these time intervals, the coordination and continuity of each link in the force chain can be quantified, thus providing detailed data support for evaluating the overall force generation continuity index.

[0075] The above technical solution effectively improves the calculation accuracy and reliability of the whole-body force continuity index. Compared to simply assessing force continuity in a general way, this application refines the peak force moments and their time intervals down to specific joints, enabling more precise identification of weak links in the learner's force chain. This helps the teaching system more accurately judge the learner's advanced motor abilities and provides a more solid data foundation for subsequent pseudo-plateau assessment and personalized training instruction generation, thereby enhancing the relevance and effectiveness of teaching.

[0076] Furthermore, step S3 includes: S31: Obtain the sequence of the learner's action execution speed index in multiple consecutive training cycles, calculate the first improvement of the action execution speed index between adjacent training cycles, and when the first improvement is lower than the preset first progress threshold for a preset number of consecutive times, and the action execution speed index reaches the preset progress target value in the current training cycle, determine that the action execution speed index meets the upper limit standard allowed by the law of motion. S32: Obtain the sequence of whole-body force coherence indicators of the learner in multiple consecutive training cycles, calculate the second improvement range of the whole-body force coherence indicators between adjacent training cycles, and determine that the whole-body force coherence indicators meet the upper limit standard allowed by the law of movement when the second improvement range is lower than the preset second progress threshold for a preset number of consecutive times, and the whole-body force coherence indicators in the current training cycle reach the preset advancement target value. S33: Obtain the sequence of movement quality change indicators of the learner in multiple consecutive training cycles. When the fluctuation value of the movement quality change indicator sequence is lower than the preset quality stability threshold, it is determined that the movement quality change indicator meets the upper limit standard allowed by the law of motion. S34: When the above three conditions are met simultaneously, the learner is judged to have met the standard and is guided to the next training exercise.

[0077] Specifically, the motion execution speed index sequence refers to the set of records of the learner's motion execution speed index in each of multiple consecutive training cycles. The first improvement margin refers to the amount or rate of change of the motion execution speed index between two adjacent training cycles, used to measure the learner's progress in motion execution speed. The first improvement threshold is a preset value; when the first improvement margin is lower than this threshold, it indicates that the learner's progress in motion execution speed is stagnating. The advanced target value is an ideal or advanced level value set for motion execution speed; reaching this value means that the learner has achieved a high level of motor ability in this area.

[0078] Similarly, the whole-body coherence index sequence refers to the set of records of the learner's whole-body coherence index in each training cycle across multiple consecutive training cycles. The second improvement magnitude refers to the amount or rate of change of the whole-body coherence index between two adjacent training cycles, used to measure the learner's progress in whole-body coherence. The second improvement threshold is a preset value; when the second improvement magnitude is lower than this threshold, it indicates that the learner's progress in whole-body coherence is stagnating.

[0079] Furthermore, the sequence of movement quality change indicators refers to the set of records of movement quality change indicators for the learner in each of multiple consecutive training cycles. The fluctuation value can be understood as the standard deviation or coefficient of variation of the movement quality change indicator sequence, used to measure the stability of the learner's movement quality. The quality stability threshold is a preset value; when the fluctuation value is below this threshold, it indicates that the learner's movement quality has reached a stable high level.

[0080] This application's solution, by comprehensively evaluating learners' movement execution speed, whole-body force coherence, and movement quality change indicators over multiple consecutive training cycles, can more accurately determine whether learners have truly reached the upper limit allowed by the laws of movement. Specifically, by monitoring the relationship between the first improvement magnitude and the first progress threshold, and the relationship between the second improvement magnitude and the second progress threshold, it can be determined whether the learner's progress in speed and force coherence has plateaued, i.e., whether a bottleneck has been reached. Simultaneously, by assessing whether the fluctuation value of the movement quality change indicator sequence is below the quality stability threshold, it can be ensured that while learners reach a high level, their movement quality also maintains a high degree of stability. Only when all three conditions are met simultaneously is the learner's movement deemed to have met the standard, thus avoiding errors that may arise from judging based on a single indicator or short-term performance.

[0081] Through the above technical solution, this application provides a more comprehensive, objective, and accurate mechanism for judging learners' motor skill attainment. This mechanism not only considers learners' improvements in advanced motor abilities (such as speed of movement execution and continuity of whole-body force exertion) but also takes into account the stability of movement quality. Therefore, it effectively avoids misjudgments caused by errors in judging a single indicator or short-term performance fluctuations, ensuring that learners only move on to the next training exercise after truly reaching a high level and achieving stable movement. This improves the scientific rigor and effectiveness of teaching and helps learners build a solid and stable foundation in motor skills.

[0082] Furthermore, step S4 includes: S41: When the improvement rate of a consecutive preset number of times is lower than the preset first progress threshold, but the action execution speed index in the current training cycle does not reach the preset advancement target value, Or, if the improvement rate of the second step is lower than the preset second progress threshold for a preset number of consecutive times, but the full-body force coherence index does not reach the preset advancement target value in the current training cycle, Alternatively, if the fluctuation value of the movement quality change index sequence is lower than the preset quality stability threshold, but the movement execution speed index or the whole body force continuity index does not reach the corresponding preset advanced target value, it is judged that the learner has entered a pseudo plateau period caused by the basic evaluation standard being too low.

[0083] Specifically, "the improvement rate for each consecutive preset number of training cycles is lower than the preset first progress threshold" means that within multiple consecutive training cycles, the learner's action execution speed indicator shows consistently small improvements, failing to reach the expected level of progress. This indicates that the learner may have plateaued in terms of action execution speed.

[0084] The fact that the speed of movement execution has not reached the preset advanced target value in the current training cycle means that although the learner shows stagnation in the speed of movement execution, he has not yet reached a higher level of advanced goal, which is different from truly reaching the upper limit standard allowed by the laws of movement.

[0085] Similarly, if the improvement rate of the second step is lower than the preset second progress threshold for a certain number of consecutive training cycles, it means that the learner's improvement rate of the whole-body force coherence index is consistently small and fails to reach the expected level of progress, indicating that there may be a bottleneck in the learner's whole-body force coordination.

[0086] The failure of the full-body power exertion continuity index to reach the preset advanced target value during the current training cycle means that the learner has not yet reached a higher level of advanced target in terms of full-body power exertion continuity.

[0087] Furthermore, a fluctuation value below the preset quality stability threshold for the movement quality change index sequence indicates that the learner's movement quality change index exhibits low volatility within consecutive training cycles, meaning the movement quality tends to stabilize. When this stability is accompanied by the movement execution speed index or the whole-body force coherence index failing to reach the corresponding preset advanced target value, it further confirms that the learner may be in a pseudo-plateau period. Specifically, the fluctuation value can be either the standard deviation or the range of the sequence, which can be preset according to the training scenario.

[0088] A pseudo-plateau can be understood as a period during training when, due to overly lenient basic evaluation standards, learners, after reaching the current basic standard, experience a stagnant improvement in their advanced motor abilities (such as speed of movement or continuity of whole-body power generation), failing to reach higher-level goals. At this point, the learner has not truly reached their maximum potential, but is limited by the current basic evaluation standards, failing to fully unleash their potential.

[0089] This application's solution, by comprehensively considering the learner's improvement in various advanced motor ability indicators (movement execution speed, whole-body force coherence, and movement quality change indicators) over a continuous training cycle, the relationship between current values ​​and advanced target values, and the stability of movement quality, can accurately identify whether a learner has entered a pseudo-plateau period caused by excessively low basic evaluation standards. Specifically, when the improvement in one or more of the learner's advanced motor ability indicators is repeatedly lower than a preset progress threshold, it indicates that their progress has significantly slowed down or stagnated; simultaneously, if these indicators have not yet reached the preset advanced target values, it excludes the possibility that they have reached their upper limit of motor ability. Furthermore, when the fluctuation value of the movement quality change indicator sequence is low, it means that the learner's movement performance has tended to stabilize, but this stability is not the stability after reaching a high level, but rather a stagnation at a relatively low level. Through the combination of the above multi-dimensional conditions, this solution can avoid confusing the true upper limit of ability plateau with a pseudo-plateau caused by unreasonable basic standards, thus providing an accurate basis for subsequent personalized intervention.

[0090] Through the aforementioned technical solution, this application enables refined identification of learners' training states. Compared to judging a pseudo-plateau solely based on indicators failing to reach the upper limit, this solution introduces multiple judgment conditions, including the magnitude of improvement, advanced target values, and the stability of movement quality, significantly improving the accuracy of pseudo-plateau identification. This accurate identification helps avoid misjudging learners' training states, allowing for timely and effective adjustments to basic evaluation standards, breaking through training bottlenecks, and enabling learners to overcome pseudo-plateaus and continuously improve athletic performance. Therefore, this solution provides learners with more precise and personalized training guidance, improving teaching efficiency and learning outcomes.

[0091] Furthermore, step S5 includes: S51: Obtain the absolute difference between the action execution speed index and the whole body force coherence index in the indicator data and the corresponding preset advanced target value as the gap; S52: Based on the size of the corresponding gap, the qualified judgment interval of the corresponding action feature in the basic evaluation standard is narrowed towards the corresponding advanced target value to obtain the adjusted basic evaluation standard. S53: The matching degree scores of the action execution speed index and the whole body force continuity index in the indicator data with the corresponding indicators in the preset advanced target values ​​are added as new scoring items and added to the scoring calculation process of the action frame. S54: Under the adjusted basic evaluation criteria, repeat steps S1 to S5 until the indicator data meets the upper limit standard allowed by the motion law.

[0092] The aforementioned indicator data can be understood as various quantitative data reflecting the learner's advanced motor abilities, such as movement execution speed indicators, whole-body force exertion continuity indicators, and movement quality change indicators. The preset advanced target values ​​refer to the ideal or expected values ​​set for these advanced motor ability indicators, representing a higher level of motor performance. Step S51 aims to quantify the gap between the learner's current advanced motor ability and the target, so as to make targeted adjustments later.

[0093] Specifically, in step S52, narrowing the pass / fail judgment interval for the corresponding action feature in the basic evaluation criteria toward the advanced target value means that the pass / fail requirements for the learner's actions become more stringent, thereby prompting the learner to control the actions more precisely during training to achieve a higher standard. For example, if the pass / fail interval for a certain action feature is [A, B], and the advanced target value is C (C>B), then the narrowing operation may adjust the pass / fail interval to [A', B'], where A'>= A, B'<= B, and B' is closer to C.

[0094] Furthermore, in step S53, the matching scores of the action execution speed index and the whole-body force coherence index with the corresponding indicators in the preset advanced target values ​​are added as new scoring items and incorporated into the scoring calculation process of the action frame. The purpose is to include the learner's achievement of the advanced target values ​​in the evaluation system of basic movements while adjusting the basic evaluation criteria. This means that even if the learner's basic movements meet the adjusted standards, their performance in advanced abilities will still affect their overall movement score, thus providing a more comprehensive assessment of the learner's progress.

[0095] The matching score can be calculated using a similar method to the scores for each action in the action frame, but the comparison is between the current indicator value and the advanced target value. Specifically, the matching score can be calculated using the following formula: Matching Score = max(0, 1 - |Current Value - Advanced Target Value| / Advanced Target Value) × 100. The current value includes the action execution speed indicator and the whole-body force exertion continuity indicator, while the advanced target value consists of the pre-set standard indicators for action execution speed and whole-body force exertion continuity.

[0096] The repeated execution mechanism of step S54 ensures the iterative nature of the adjustment process. After each adjustment of the basic evaluation criteria, the system re-evaluates the learner's movements until their advanced motor ability indicators reach the upper limit allowed by the laws of motor development, that is, the learner has truly broken through the pseudo-plateau period.

[0097] The solution proposed in this application systematically addresses the problem of learners' stagnation in progress during the pseudo-plateau period due to excessively low basic evaluation standards, through the detailed steps described above.

[0098] Specifically, step S51 quantifies the gap between the learner's current advanced motor ability and the preset advanced target value, providing a clear basis for subsequent adjustments. Then, step S52 directly raises the requirements for the learner's basic movements by narrowing the pass / fail range of the basic evaluation criteria, forcing learners to focus more on the precision and efficiency of their movements during training, thereby indirectly promoting the improvement of their advanced motor abilities. Simultaneously, step S53 incorporates the matching score of the advanced target value into the scoring calculation of the movement frames, ensuring that the evaluation of basic movements is no longer limited to the minimum pass line but is closely linked to higher-level advanced goals. This helps learners establish an understanding and pursuit of high-level movements from the early stages of training. Finally, the iterative adjustment mechanism in step S54 ensures that the basic evaluation criteria can dynamically adapt to the learner's progress, gradually guiding the learner to break through the pseudo-plateau period until their advanced motor ability reaches the upper limit allowed by the laws of motor development.

[0099] Furthermore, step S6 includes: S61: After each adjustment to the basic evaluation criteria, the advanced motor skills with the largest gap are identified as the weak link. S62: Based on the pre-defined causal relationship between action features, reverse the process from the weak point to find the fundamental action feature that causes the weak point; whereby the fundamental action feature that causes the weak point is the root cause of the learner's action deviation. S63: Generate personalized training instructions targeting fundamental movement characteristics.

[0100] Specifically, after adjusting the basic evaluation criteria, the system will reassess the gap between the learner's various indicator data and the preset advanced target values. Among them, the advanced motor ability with the largest gap will be identified as the learner's current weakness. For example, if the gap between the movement execution speed indicator and the preset advanced target value is the largest, then movement execution speed will be identified as a weakness.

[0101] Furthermore, to identify the root cause of this weakness, this application utilizes a reverse analysis based on the causal relationships between pre-defined movement features. For example, if movement execution speed is a weak point, the system may trace back to more fundamental movement features based on causal relationships (e.g., the whole-body force coherence index affects the movement execution speed index, and body posture affects the whole-body force coherence index), such as a specific angle of a joint or the timing of force application. This most fundamental movement feature, traced back to the point of origin, is identified as the root cause of the learner's movement deviation.

[0102] The causal relationships between pre-defined action characteristics can be established in advance based on the biomechanical principles of different ball sports and teaching experience. The following are examples of causal relationships in three different ball sports scenarios (this is to be understood that in actual application, the examples are not limited to these three): Basketball shooting training: Insufficient lower limb extension strength leads to excessive forward tilt of the torso, which in turn results in a low release point, leading to a decrease in shooting accuracy. The system starts with the weakness of slow movement execution speed and traces it back to insufficient lower limb extension strength as the fundamental movement characteristic.

[0103] Badminton smash training: Insufficient wrist pronation speed leads to a deviation in the racket face angle at the moment of impact, causing the smash to land outside the target area, thus reducing the smash's threat. Simultaneously, insufficient lower limb rotational power results in slow hip joint rotation, which in turn leads to insufficient trunk rotation, resulting in insufficient initial upper limb swing velocity, and consequently, a slow smash speed. The system, starting with the weakness of slow smash speed, can trace back to insufficient lower limb rotational power as the fundamental movement characteristic; similarly, starting with the weakness of misplaced smashes, it can trace back to insufficient wrist pronation speed as the fundamental movement characteristic.

[0104] In soccer kicking training: A stance with the supporting foot too far out results in insufficient swing amplitude of the kicking leg, leading to inadequate acceleration of the lower leg at the moment of impact, resulting in slow ball speed. Simultaneously, excessive hip abduction during the kicking leg swing causes the striking point to deviate from the center of the ball, resulting in an unstable ball trajectory and decreased passing or shooting accuracy. Furthermore, leaning the body backward after impact causes the body to lean backward, shifting the force direction upward, leading to an excessively high kick arc, which results in reduced ball speed or the ball going over the crossbar. The system addresses the weakness of slow ball speed by tracing back to an excessively far-out stance of the supporting foot; the weakness of decreased passing accuracy by tracing back to excessive hip abduction during the kicking leg swing; and the weakness of an excessively high kick arc by tracing back to a leaning body center of gravity.

[0105] The above causal relationships are merely examples. In practical applications, corresponding causal chains can be established based on the characteristics of specific sports and teaching needs. By mapping the currently identified weaknesses to a preset causal chain and tracing back along the causal relationship, the system can locate the fundamental movement characteristics that lead to the weakness.

[0106] Therefore, once the root cause of a learner's movement deviation is identified, the system generates personalized training instructions targeting that fundamental movement characteristic. These instructions aim to directly correct or strengthen that fundamental movement characteristic, thereby addressing the learner's weaknesses at their root. For example, if the root cause is insufficient hip extension, specific training instructions for hip extension will be generated.

[0107] This application's solution refines step S6, which identifies the root cause of a learner's motor deviations, into three sub-steps: identifying weaknesses, tracing back to fundamental motor characteristics, and generating personalized training instructions. This enables precise diagnosis and targeted intervention. Specifically, step S61 quantifies the gap between various advanced motor abilities and preset advanced target values, objectively identifying the learner's most pressing areas for improvement and avoiding biases from subjective judgment. Step S62 utilizes preset causal relationships of motor characteristics to delve into the underlying fundamental motor characteristics behind apparent weaknesses, ensuring that training goes beyond symptom-based approaches and directly addresses the core of the problem. Finally, step S63, based on the precise identification of fundamental motor characteristics, generates highly personalized and effective training instructions, ensuring that each training session maximizes the learner's motor performance.

[0108] Through the aforementioned technical solution, this application can significantly improve the accuracy and effectiveness of personalized training. Traditional training methods may only identify a learner's deficiencies in a certain advanced ability, but struggle to pinpoint the underlying causes, leading to poor training results or low efficiency. However, this application systematically identifies the weakest link with the greatest gap and further utilizes the causal relationship of movement characteristics to trace back to the fundamental movement characteristics, enabling the generated personalized training instructions to directly target the root cause of movement deviations. This not only helps learners overcome pseudo-plateaus more quickly and effectively, achieving continuous improvement in motor skills, but also avoids unnecessary generalization training, thereby improving training efficiency and reducing learning costs.

[0109] Please refer to Figure 2 This application also provides an interactive ball sports teaching system, which implements the steps of any of the above methods. The system includes: Acquisition Module 201: Acquires the learner's action data within a single training cycle and determines whether the learner's actions meet the basic evaluation criteria based on the action data; Extraction module 202: When the basic evaluation criteria are met, extract the indicator data reflecting advanced motor ability from the motion data; Judgment Module 203: If the indicator data meets the upper limit standard allowed by the law of motion, the learner's movement is judged to have met the standard, and the learner is guided to the next training exercise. Judgment Module 204: If the indicator data does not meet the upper limit standard allowed by the law of motion, it is determined that the learner has entered a pseudo plateau period caused by the low basic evaluation standard; Adjustment module 205: During the pseudo-plateau period, calculate the difference between the indicator data and the preset advanced target value, and adjust the basic evaluation standard according to the difference until the indicator data meets the upper limit standard allowed by the movement law; Training guidance module 206: After each adjustment of the basic evaluation criteria, the module identifies the root cause of the learner's deviation in action based on the gap, and generates corresponding personalized training instructions based on the root cause.

[0110] This interactive ball sports teaching system aims to construct a dynamic, adaptive evaluation and training loop that adapts to learners' skill levels through the collaborative work of its various functional modules. The system acquires learners' movement data in real time through module 201, which then determines whether the learner has reached the basic evaluation standard. Once the basic movement is achieved, module 202 further analyzes advanced movement ability indicators. Subsequently, module 203 assesses whether these indicators have reached the upper limit allowed by the laws of movement to determine whether to proceed to the next stage of training. If the upper limit is not reached and progress stagnates, module 204 identifies a pseudo-plateau. In this case, module 205 dynamically narrows the basic evaluation standard based on the gap between the learner's performance and the advanced target value, prompting the learner to improve under stricter requirements. Finally, training guidance module 206 accurately locates the root cause of movement deviations based on the adjusted gap and generates personalized training instructions, effectively solving the problem of static evaluation benchmarks in traditional teaching and ensuring the continuity and effectiveness of training.

[0111] The interactive ball sports teaching system provided in this application can be implemented in various ways, usually through software programs, hardware circuits, or a combination of both.

[0112] Specifically, the acquisition module 201 can be configured to collect learner's motion data in real time during training using various sensor devices, such as high-speed cameras, inertial measurement unit (IMU) sensors, or pressure sensors. This data can include limb joint angles, body center of gravity position, motion trajectory, velocity, acceleration, and ground reaction force. The acquisition module 201 preprocesses the collected raw data, such as denoising and filtering, and transforms it into structured data that can be analyzed by subsequent modules. The specific process of acquiring learner motion data and determining whether the learner's actions meet the basic evaluation criteria based on the motion data has been described in detail in the above embodiments and will not be repeated here. It is important to emphasize that the acquisition module 201, as the system's input interface, is designed to ensure the comprehensiveness and accuracy of the motion data, providing a reliable data foundation for subsequent evaluation and training.

[0113] After receiving motion data conforming to the basic evaluation criteria provided by the acquisition module 201, the extraction module 202 is designed to identify and quantify indicator data reflecting advanced motor abilities from this data. This module may include a series of data processing and calculation methods, such as methods for calculating motion execution speed, force continuity, or changes in motion quality. The specific process of extracting indicator data reflecting advanced motor abilities from the motion data when it meets the basic evaluation criteria has been described in detail in the above embodiments and will not be repeated here. The effectiveness of the extraction module 202 lies in its ability to accurately identify quantitative indicators crucial to advanced abilities from massive amounts of motion data.

[0114] The judgment module 203 is configured to evaluate the advanced motor ability index data output by the extraction module. This module may include a logical judgment unit and a preset threshold or model library, used to compare the current index data with the upper limit standard allowed by the movement laws. The specific process of determining that the learner's movement has reached the standard if the index data meets the upper limit standard allowed by the movement laws, and guiding the learner to the next training exercise, has been described in detail in the above embodiments and will not be repeated here. The judgment module 203 ensures that the system can accurately identify whether the learner has reached the highest level of the current training stage.

[0115] The judgment module 204 is designed to identify whether a learner has entered a pseudo-plateau period caused by excessively low basic evaluation standards. This module may include trend analysis algorithms and state machine logic to continuously monitor the relationship between the improvement trend of the learner's advanced motor ability indicators and the pass rate of basic movements over multiple consecutive training cycles. The specific process for determining whether a learner has entered a pseudo-plateau period caused by excessively low basic evaluation standards if the indicator data does not meet the upper limit allowed by the laws of motor skills has been described in detail in the above embodiments and will not be repeated here. The judgment module 204 is key to the system's ability to dynamically adapt to the learner's skill level, avoiding misjudgments caused by static evaluation in traditional systems.

[0116] The adjustment module 205 is activated after the judgment module 204 identifies a pseudo-plateau period. This module may include a parameter adjustment algorithm and a standard update mechanism to calculate the gap between the current indicator data and the preset advanced target value, and dynamically modify the pass / fail judgment interval of the corresponding action feature in the basic evaluation standard based on this gap. The specific process of calculating the gap between the indicator data and the preset advanced target value during the pseudo-plateau period, and adjusting the basic evaluation standard according to the gap until the indicator data meets the upper limit standard allowed by the movement law, has been described in detail in the above embodiments and will not be repeated here. By dynamically narrowing the evaluation standard, the adjustment module effectively drives learners to break through skill bottlenecks.

[0117] The training guidance module 206 is triggered after each adjustment of the basic evaluation criteria by the adjustment module. This module may include a root cause analysis engine and an instruction generator, used to locate the root causes of learner deviations based on the gaps calculated by the adjustment module 205, and generate targeted personalized training instructions. The specific process of locating the root causes of learner deviations based on the gaps after each adjustment of the basic evaluation criteria, and generating corresponding personalized training instructions based on the root causes, has been described in detail in the above embodiments and will not be repeated here. The training guidance module ensures that the system can provide highly customized, accurate, and effective training suggestions, thereby achieving efficient skill improvement.

[0118] This application provides an interactive ball sports teaching system, which offers an innovative solution to the problems in existing technologies where static evaluation benchmarks lead to a loss of discriminative power in evaluation results and personalized training fails to match learners' actual skill levels.

[0119] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. An interactive ball teaching method, characterized by, The method includes the following steps: S1: Obtain the learner's action data within a single training cycle, and determine whether the learner's actions meet the basic evaluation criteria based on the action data; S2: When the basic evaluation criteria are met, extract the indicator data reflecting advanced motor ability from the motion data; S3: If the indicator data meets the upper limit standard allowed by the movement law, it is determined that the learner's movement has reached the standard, and the learner is guided to the next training exercise; S4: If the indicator data does not meet the upper limit standard allowed by the law of motion, it is determined that the learner has entered a pseudo plateau period caused by the low basic evaluation standard; S5: During the pseudo-plateau period, calculate the difference between the indicator data and the preset advanced target value, and adjust the basic evaluation standard according to the difference until the indicator data meets the upper limit standard allowed by the motion law. S6: After each adjustment of the basic evaluation criteria, the root cause of the learner's deviation in action is located based on the gap, and corresponding personalized training instructions are generated based on the root cause.

2. The interactive ball teaching method according to claim 1, wherein, Step S1 includes: S11: Acquire the learner’s action data in a single training cycle, and split the action data into multiple consecutive action frames according to the action data sampling points; S12: Calculate the score for each action frame, count the number of excellent actions whose scores exceed a preset excellent score threshold, and calculate the excellent action rate based on the proportion of the number of excellent actions in the total number of action frames. S13: Calculate the score variance based on the scores of all the action frames, and determine whether the score variance is less than a preset stability threshold. S14: When the rate of excellent actions exceeds the preset proportion and the score variance is less than the preset stability threshold, the learner's actions are determined to meet the basic evaluation criteria; otherwise, the learner's actions are determined to not meet the basic evaluation criteria.

3. The interactive ball game teaching method according to claim 2, characterized in that, In step S12, calculating the score for each action frame includes the following steps: S121: Extract the limb joint angles and body center of gravity position from each action frame as static action features; S122: Calculate the joint angular velocity as a dynamic motion feature based on the change in joint angle and the time interval between at least two consecutive action frames; S123: Compare the static motion features and the dynamic motion features with the corresponding standard joint angles, standard body center of gravity positions, and standard joint angular velocities in the basic evaluation criteria to obtain a matching score for each motion feature relative to the standard motion. S124: Sum the matching scores of all action features to obtain the score of the action frame.

4. The interactive ball game teaching method according to claim 1, characterized in that, Step S2 includes: S21: Identify the start and end times of each action execution cycle from the action data, calculate the time difference between the end and start times of each action execution cycle as action execution speed data, and calculate the average value of all the action execution speed data as the action execution speed index. S22: Extract the peak force exertion time of each joint of the lower limb, trunk and upper limb in each action execution cycle from the action data, calculate the time interval between two adjacent peak force exertion times according to the force exertion sequence, compare each time interval with its corresponding standard time interval, obtain the force exertion continuity data of the corresponding action execution cycle according to the degree of deviation obtained by comparison, and calculate the average value of the force exertion continuity data of all action cycles as the whole body force exertion continuity index. S23: Divide multiple consecutive action execution cycles into an early cycle group and a late cycle group, calculate the degree of deviation of the action execution speed data or force continuity data of the late cycle group relative to the early cycle group, and use it as an indicator of action quality change. The index data includes at least the action execution speed index, the whole-body force exertion continuity index, and the action quality change index.

5. The interactive ball game teaching method according to claim 4, characterized in that, In step S22, the peak force exertion times of each joint of the lower limb, trunk and upper limb are respectively the first time corresponding to the peak value of the knee joint extension angular velocity, the second time corresponding to the peak value of the hip joint extension angular velocity, the third time corresponding to the peak value of the shoulder joint flexion angular velocity, and the fourth time corresponding to the peak value of the elbow joint extension angular velocity. Calculating the time interval between two adjacent peak force moments according to the force exertion sequence includes the following steps: S221: Calculate the first time interval between the first moment and the second moment according to the force exertion sequence; S222: Calculate the second time interval between the second time moment and the third time moment; S223: Calculate the third time interval between the third time moment and the fourth time moment.

6. The interactive ball game teaching method according to claim 5, characterized in that, Step S3 includes: S31: Obtain the sequence of the learner's action execution speed index in multiple consecutive training cycles, calculate the first improvement range of the action execution speed index between adjacent training cycles, and when the first improvement range is lower than the preset first progress threshold for a preset number of consecutive times, and the action execution speed index reaches the preset progress target value in the current training cycle, determine that the action execution speed index meets the upper limit standard allowed by the law of motion. S32: Obtain the sequence of the learner's whole-body force coherence index in multiple consecutive training cycles, calculate the second improvement range of the whole-body force coherence index between adjacent training cycles, and when the second improvement range is lower than the preset second progress threshold for a preset number of consecutive times, and the whole-body force coherence index reaches the preset advancement target value in the current training cycle, determine that the whole-body force coherence index meets the upper limit standard allowed by the law of motion. S33: Obtain the sequence of movement quality change indicators of the learner in multiple consecutive training cycles. When the fluctuation value of the movement quality change indicator sequence is lower than the preset quality stability threshold, determine that the movement quality change indicator meets the upper limit standard allowed by the law of motion. S34: When the above three conditions are met simultaneously, the learner's action is determined to be up to standard, and the learner is guided to the next training exercise.

7. The interactive ball game teaching method according to claim 6, characterized in that, Step S4 includes: S41: When the improvement rate of a preset number of consecutive first improvements is lower than the preset first progress threshold, but the action execution speed index in the current training cycle has not reached the preset advancement target value, Or, if the improvement rate of the second step is lower than the preset second progress threshold for a preset number of consecutive times, but the whole-body force coherence index does not reach the preset advancement target value in the current training cycle, Alternatively, if the fluctuation value of the sequence of action quality change indicators is lower than the preset quality stability threshold, but the action execution speed indicator or the whole-body force coherence indicator does not reach the corresponding preset advanced target value, it is determined that the learner has entered a pseudo-plateau period caused by the low basic evaluation standard.

8. The interactive ball game teaching method according to claim 1, characterized in that, Step S5 includes: S51: Obtain the absolute difference between the action execution speed index, the whole body force coherence index and the corresponding preset advanced target value in the index data, as the gap; S52: Based on the size of the corresponding gap, the qualified judgment interval of the corresponding action feature in the basic evaluation standard is narrowed towards the direction of the corresponding advanced target value to obtain the adjusted basic evaluation standard; S53: The matching degree scores of the action execution speed index and the whole body force continuity index in the index data with the corresponding indicators in the preset advanced target values ​​are added as new scoring items and added to the scoring calculation process of the action frame. S54: Under the adjusted basic evaluation criteria, repeat steps S1 to S5 until the index data meets the upper limit standard allowed by the motion law.

9. The interactive ball game teaching method according to claim 1, characterized in that, Step S6 includes: S61: After each adjustment of the basic evaluation criteria, the advanced motor ability with the largest gap is identified as the weak link; S62: Based on the causal relationship between preset action features, reverse the process from the weak link to find the fundamental action feature that causes the weak link; wherein, the fundamental action feature that causes the weak link is the root cause of the deviation in the learner's action. S63: Generate personalized training instructions for the fundamental motion characteristics.

10. An interactive ball game teaching system, characterized in that, The system is used to implement the steps of the method according to any one of claims 1-9 above, and the system includes: Acquisition module: Acquires the learner's action data within a single training cycle, and determines whether the learner's actions meet the basic evaluation criteria based on the action data; Extraction module: When the basic evaluation criteria are met, extract the indicator data reflecting advanced motor ability from the action data; Judgment module: If the indicator data meets the upper limit standard allowed by the movement law, the learner's movement is judged to have met the standard, and the learner is guided to the next training exercise; Judgment module: If the indicator data does not meet the upper limit standard allowed by the movement law, it is determined that the learner has entered a pseudo plateau period caused by the low basic evaluation standard; Adjustment module: During the pseudo-plateau period, calculate the difference between the indicator data and the preset advanced target value, and adjust the basic evaluation standard according to the difference until the indicator data meets the upper limit standard allowed by the motion law; Training guidance module: After each adjustment of the basic evaluation criteria, the module identifies the root cause of the learner's deviation in action based on the gap, and generates corresponding personalized training instructions based on the root cause.

Images