Athlete action performance analysis method based on big data analysis

Abstract

The invention discloses an athlete action performance analysis method based on big data analysis, and particularly relates to the technical field of action performance, and the method comprises the steps: monitoring and determining a penalty process of an athlete through a sensor, and analyzing the optimal track of each shooting action through clustering according to a time sequence of extracting key feature points of a forearm, and calculating a minimum cost path between the current penalty ball track and the optimal track through a dynamic time warping algorithm, calculating a success probability under the current forearm angle amplitude through a Bayesian theorem based on the forearm angle amplitude of each penalty ball of the athlete, and determining personalized expression information of the forearm action of the athlete. By collecting the speed time sequences of different feature points of the forearm of the athlete in the penalty process and fitting the relationship between the speed and the time by using the least square method, the actual performance information of the forearm action of the athlete is determined, and the method is beneficial to the personalized analysis of the influence of the forearm on the penalty hitting in the penalty process of the athlete.

Description

Technical Field

[0001] This invention relates to the field of motion performance technology, and more specifically, to a method for analyzing athlete motion performance based on big data analysis. Background Technology

[0002] Basketball shooting, especially free throws, is a process that highly relies on the coordinated work of various parts of the athlete's body. The success rate of a free throw depends on the coordination and power control of multiple parts of the body, including the arm, wrist, knee, and footwork. Among these factors, the forearm plays a crucial role in the entire free throw motion, directly affecting the shooting angle, the transfer of force, and the final shooting accuracy.

[0003] Traditional sports performance analysis methods often struggle to accurately separate and quantify the impact of individual body parts on movement performance, especially in high-intensity, multi-body coordination sports like basketball, where each part of the movement is affected by others. Most traditional analysis methods rely on video analysis or motion capture technology, but these technologies cannot effectively provide in-depth, personalized assessments of individual athlete movements. They ignore the impact of individual differences on movement, lack personalized analysis of every detail of the shooting motion, and cannot offer targeted and practical training recommendations. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a method for analyzing athlete performance based on big data analysis, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] The method for analyzing athlete performance based on big data analytics includes the following steps:

[0007] S1: By monitoring the angle changes of the athlete's forearm and wrist through sensors, the starting moment of the athlete's free throw is identified, and combined with computer vision technology, the ending moment of the basketball leaving the athlete's palm is identified, thus determining the athlete's free throw process.

[0008] S2: Extract the time series of key feature points of the forearm, use cluster analysis to analyze the optimal trajectory of each shooting action, calculate the minimum cost path between the current free throw trajectory and the optimal trajectory through the dynamic time warping algorithm, and calculate the success probability under the current forearm angle amplitude based on the forearm angle amplitude of each free throw by the athlete, thereby determining the personalized performance information of the athlete's forearm action.

[0009] S3: By collecting the velocity time series of different feature points of the athlete's forearm during the free throw process, and using the least squares method to fit the relationship between velocity and time, the actual performance information of the athlete's forearm movement is determined.

[0010] S4: Quantify the impact of the forearm on free throws through a comprehensive evaluation of the athlete's personalized forearm movement and actual performance information, and generate warning signals.

[0011] In a preferred embodiment, determining the athlete's personalized performance information and actual performance information for forearm movements includes:

[0012] The personalized performance information of the athlete's forearm movement is represented by the displacement feature regularization deviation coefficient and the angle amplitude Bayesian distribution coefficient, while the actual performance information of the athlete's forearm movement is represented by the feature velocity anomaly deviation coefficient. This is the displacement characteristic regularity deviation coefficient. The Bayesian distribution coefficient of the angular amplitude. This is the characteristic velocity anomaly deviation coefficient.

[0013] In a preferred embodiment, the logic for obtaining the displacement characteristic regularization deviation coefficient is as follows:

[0014] Based on athletes' historical free throw data, successful free throws are selected. For each successful free throw, feature points at key positions in the athlete's forearm before and after the free throw are extracted, and a time series of these feature points is generated. The time series of feature points is represented as follows: Where t is the time point, i is the feature point number, i = 1, 2, 3, ..., I, I is a positive integer, and x, y, z are the spatial coordinates of the feature point, respectively.

[0015] Based on the time series of feature points, clustering analysis is performed on the time series of feature points using a clustering algorithm. Each shooting action of the athlete is divided into several clusters, and each cluster represents a specific shooting action pattern. The cluster center of each cluster is obtained as the optimal trajectory for each shooting action.

[0016] Based on the current athlete's free throw shooting motion pattern, determine the optimal trajectory belonging to the current athlete's free throw shooting motion pattern, and mark the optimal trajectory of the feature points in the optimal trajectory as: , And mark the characteristic trajectory of the current athlete's free throw as: , The dynamic time warping algorithm is used to determine the minimum cost path between the feature point trajectory of the current athlete's free throw and the optimal feature point trajectory, and the minimum cost path is marked as: ;

[0017] The displacement characteristic regularity deviation coefficient is calculated using the following formula: .

[0018] In a preferred embodiment, the logic for obtaining the Bayesian distribution coefficient of the angular amplitude is as follows:

[0019] Based on the athlete's historical free throw data, the forearm angle amplitude of each historical free throw is determined and marked as follows: ;in, , The angle of the athlete's forearm at the end of the free throw. The angle of the athlete's forearm at the start of the free throw;

[0020] Based on the forearm angle amplitude of each of the athlete's historical free throws, the mean and standard deviation of the historical forearm angle amplitude under successful free throws are determined, and the mean and standard deviation of the historical forearm angle amplitude under successful free throws are denoted as: and The probability density distribution of the historical forearm angle amplitude under successful free throws is calculated using the Gaussian probability density function. The calculation formula is as follows: ;in, The probability of an athlete successfully shooting a free throw with the amplitude of their forearm angle A.

[0021] Based on the athlete's historical free throw data, determine the number of successful free throws the athlete has under the current forearm angle amplitude, obtain the total number of successful free throws the athlete has, and calculate the probability of the athlete successfully making a free throw under the current forearm angle amplitude. The calculation formula is as follows: ;in, This represents the probability of a player successfully shooting a free throw given the current forearm angle amplitude. This represents the current amplitude of the athlete's forearm angle. This represents the number of successful free throws by the athlete under the current forearm angle amplitude. Given the total number of free throws by the athlete in historical free throw data, determine the number of free throws the athlete can attempt at the current forearm angle amplitude, and calculate the probability of the athlete attempting free throws at the current forearm angle amplitude using the following formula: ;in, This represents the probability of the athlete's current forearm angle amplitude. This represents the number of times a player has failed a free throw in historical free throw data.

[0022] The Bayesian distribution coefficient of angular amplitude is calculated based on Bayes' theorem, and the formula is as follows: .

[0023] In a preferred embodiment, the logic for obtaining the characteristic velocity anomaly deviation coefficient is as follows:

[0024] The velocities of different feature points in the athlete's forearm during a free throw were collected, resulting in velocity-time series of these points. The least squares method was used to fit the velocity and time of each feature point, yielding a function of velocity versus time for each feature point. This function is denoted as: ;

[0025] Set speed range thresholds for different feature points, obtain the time periods during which different feature points exceed the speed range thresholds during the free throw process, and calculate the abnormal deviation coefficient of the feature speed. The calculation formula is as follows: ;in, The influence weights of velocity deviation at different feature points in the athlete's forearm are determined. The time period during which feature point i exceeds the maximum speed threshold during the free throw process. This refers to the time period during which feature point i exceeds the minimum speed threshold during the penalty kick process.

[0026] In a preferred embodiment, the comprehensive evaluation of the athlete's forearm movement, including personalized performance information and actual performance information, includes:

[0027] This study comprehensively analyzes the personalized and actual performance information of athletes' forearm movements. By weighting the normalized displacement characteristic regularization deviation coefficient, angular amplitude Bayesian distribution coefficient, and characteristic velocity anomaly deviation coefficient, a performance evaluation model is constructed, generating performance evaluation coefficients. The formula for calculating the performance evaluation coefficients is as follows: ;in, For performance evaluation coefficients, , , These are the proportional coefficients for the displacement characteristic regularity deviation coefficient, the Bayesian distribution coefficient of the angular amplitude, and the characteristic velocity anomaly deviation coefficient, respectively. , , All are greater than 0.

[0028] In a preferred embodiment, generating a warning signal includes:

[0029] Set a performance evaluation coefficient threshold, obtain the athlete's current free throw performance evaluation coefficient, compare the performance evaluation coefficient with the performance evaluation coefficient threshold, if the performance evaluation coefficient is greater than the performance evaluation coefficient threshold, a warning signal is generated, if the performance evaluation coefficient is less than the performance evaluation coefficient threshold, no warning signal is generated.

[0030] The technical effects and advantages of this invention are as follows:

[0031] This invention monitors the angle changes of an athlete's forearm and wrist using sensors, and combines this with computer vision technology to accurately identify the free throw process. Based on each athlete's historical free throw data, it uses methods such as dynamic time warping and Bayes' theorem to provide personalized movement patterns and success probability assessments, determining the personalized performance information and actual performance information of the athlete's forearm movements. This invention helps to analyze the impact of the athlete's forearm on the accuracy of the free throw, and by promptly identifying and correcting unstable factors in the shooting motion, it ensures that each free throw motion is as consistent as possible, thereby improving the free throw accuracy. Attached Figure Description

[0032] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings;

[0033] Figure 1 This is a flowchart illustrating the athlete performance analysis method based on big data analysis according to the present invention. Detailed Implementation

[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] Example 1

[0036] Figure 1 This is a flowchart illustrating the athlete performance analysis method based on big data analysis according to the present invention, which specifically includes the following steps:

[0037] S1: By monitoring the angle changes of the athlete's forearm and wrist through sensors, the starting moment of the athlete's free throw is identified, and combined with computer vision technology, the ending moment of the basketball leaving the athlete's palm is identified, thus determining the athlete's free throw process.

[0038] S2: Extract the time series of key feature points of the forearm, use cluster analysis to analyze the optimal trajectory of each shooting action, calculate the minimum cost path between the current free throw trajectory and the optimal trajectory through the dynamic time warping algorithm, and calculate the success probability under the current forearm angle amplitude based on the forearm angle amplitude of each free throw by the athlete, thereby determining the personalized performance information of the athlete's forearm action.

[0039] S3: By collecting the velocity time series of different feature points of the athlete's forearm during the free throw process, and using the least squares method to fit the relationship between velocity and time, the actual performance information of the athlete's forearm movement is determined.

[0040] S4: Quantify the impact of the forearm on free throws through a comprehensive evaluation of the athlete's personalized forearm movement and actual performance information, and generate warning signals.

[0041] Free throws are a type of shot in basketball that is taken from a fixed spot and is usually free from external interference. Therefore, the athlete's shooting motion requires extremely high precision and consistency. Any slight deviation can lead to the success or failure of the shot. In this case, accurately identifying each key stage of the shooting process, especially the start, process and end of the shooting motion, is crucial for analyzing and optimizing shooting performance.

[0042] Different athletes exhibit different movement patterns when shooting free throws, and even their shooting methods during training may vary significantly. Individual differences in an athlete's height, body type, arm length, strength, flexibility, and coordination can lead to variations in arm angle, range of motion, and wrist power during the shooting process. The power generation phase of the free throw is mainly completed by the arm and wrist, and the coordination between the forearm and wrist determines the angle and power transfer of the shot. Therefore, the athlete's forearm plays a core role in the power generation, stability, and accuracy of the free throw.

[0043] To accurately identify the start and end of an athlete's free throw shooting motion, multiple methods are combined, including sensor data, motion capture technology, and computer vision. When an athlete is preparing to shoot, there is usually a fixed motion pattern. The start of the free throw is determined by identifying the key moment when the athlete is preparing to start the motion, and the end of the free throw is determined by identifying the key moment when the basketball leaves the athlete's hand.

[0044] It should be noted that the angle changes of the arm and forearm are monitored by sensors, and a threshold is set. When the angle change exceeds the threshold, it is considered that the athlete is ready to start the action. The starting moment of this change is taken as the starting moment of the shot. The system also combines computer vision to identify the critical moment when the basketball leaves the athlete's hand.

[0045] Based on the trajectory of the athlete's forearm during the free throw, and combined with big data analysis of the athlete's free throw history, personalized performance information and actual performance information of the athlete's forearm movement are collected. The personalized performance information of the athlete's forearm movement is represented by the displacement feature regularization deviation coefficient and the angle amplitude Bayesian distribution coefficient, while the actual performance information of the athlete's forearm movement is represented by the characteristic velocity anomaly deviation coefficient.

[0046] The advantage of the displacement characteristic regularity deviation coefficient is that:

[0047] The displacement feature regularization deviation coefficient calculates the difference between the current athlete's free throw action and its historical successful free throw actions through dynamic time regularization. It provides a quantitative indicator for action analysis, which helps to quantify the deviation and inconsistency of the athlete's action, and thus make more targeted training adjustments.

[0048] The displacement feature regularization deviation coefficient uses K-means clustering to divide the athlete's successful free throw action into several clusters. Each cluster represents a specific shooting action pattern. This clustering method can adapt to the individual differences of each athlete, find commonalities and individualities among different athletes, and thus generate a personalized optimal trajectory for each athlete.

[0049] The displacement feature regularization deviation coefficient reflects the consistency and stability of an athlete's free throw motion. If the displacement feature regularization deviation coefficient is small, it indicates that the athlete's motion pattern is relatively stable and the motion is consistent. If the displacement feature regularization deviation coefficient is large, it indicates that the athlete's motion fluctuates greatly and there may be unstable shooting motions.

[0050] The logic for obtaining the displacement feature regularization deviation coefficient is as follows: Based on the athlete's historical free throw data, the data of successful free throws are filtered. For each successful free throw action, feature points at key positions in the athlete's forearm before and after the free throw are extracted, and a time series of feature points is formed. The time series of feature points is represented as follows: Where t is the time point, i is the feature point number, i = 1, 2, 3, ..., I, I is a positive integer, and x, y, z are the spatial coordinates of the feature point, respectively.

[0051] It should be noted that the filtered data only includes the athlete's successful free throws. This data reflects the athlete's shooting motion in their best condition, which helps to determine the dataset of the athlete's successful shooting motion. The key feature points in the athlete's forearm are usually the athlete's wrist and elbow positions. The time series of the feature points reflects the relationship between the coordinate position of the feature points and the change over time in the athlete's free throw motion. The coordinates of each time point record the three-dimensional spatial position of the wrist and elbow. Over time, they describe the trajectory of the athlete's arm movement during the shooting process.

[0052] Based on the time series of feature points, clustering analysis is performed on the time series of feature points using a clustering algorithm. Each shooting action of the athlete is divided into several clusters, and each cluster represents a specific shooting action pattern. The cluster center of each cluster is obtained as the optimal trajectory for each shooting action.

[0053] It should be noted that K-means clustering is used to group similar actions into one class by calculating the mean trajectory of each time series. The spatial coordinate sequence of each feature point is regarded as a vector and their Euclidean distance is calculated. The cluster center of each cluster is the average trajectory of the actions of that class, which can be regarded as the representative trajectory of the cluster.

[0054] Based on the current athlete's free throw shooting motion pattern, determine the optimal trajectory belonging to the current athlete's free throw shooting motion pattern, and mark the optimal trajectory of the feature points in the optimal trajectory as: , And mark the characteristic trajectory of the current athlete's free throw as: , The dynamic time warping algorithm is used to determine the minimum cost path between the feature point trajectory of the current athlete's free throw and the optimal feature point trajectory, and the minimum cost path is marked as: ;

[0055] It should be noted that the minimum cost path refers to finding the path from the first time point to the last time point through the cumulative distance matrix, which minimizes the total distance between the two time series. It represents the minimum alignment path and difference between the current athlete's feature point trajectory and the optimal trajectory, reflecting the difference between the athlete's current free throw action and its historical successful free throw actions. That is, if the cost of the minimum cost path is small, it means that the current athlete's free throw trajectory is very similar to the optimal trajectory, which may indicate that the athlete's shooting action is at a high level. If the cost of the minimum cost path is large, it means that the current athlete's free throw trajectory is significantly different from the optimal trajectory, which may require action adjustment or training.

[0056] The displacement characteristic regularity deviation coefficient is calculated using the following formula: ;in, This is the displacement characteristic regularity deviation coefficient.

[0057] As can be seen from the formula, the smaller the displacement characteristic regularity deviation coefficient, the greater the difference between the current athlete's free throw action and the historical best trajectory. There may be inconsistencies in the shooting action, and the athlete needs to be reminded to adjust the shooting posture and informed that the action is deformed.

[0058] The advantage of the Bayesian distribution coefficient of angular amplitude is that:

[0059] The angle amplitude Bayesian distribution coefficient can provide personalized assessments based on an athlete's historical free throw data and current forearm angle amplitude. Each athlete's forearm angle amplitude has a unique pattern. By calculating the angle amplitude Bayesian distribution coefficient, targeted feedback and suggestions can be provided to different athletes, thereby improving the targeting and effectiveness of training.

[0060] Bayesian methods can dynamically update prior and posterior distributions based on new data. As athletes' shooting data accumulates, the system can adjust the probability of success in real time, providing athletes with more accurate action feedback. This allows the training process to be flexibly adjusted according to changes in the athlete's state, thereby helping athletes achieve their best performance under different conditions.

[0061] By using the Bayesian distribution coefficient of angle amplitude, coaches and athletes can more clearly understand the impact of different forearm angles on free throw success rate. If a certain angle is highly correlated with successful free throws, athletes can optimize their free throw performance by adjusting their movements. For unsuccessful angles, data can be used to identify and improve them, thereby increasing the overall success rate.

[0062] The logic for obtaining the Bayesian distribution coefficient of the angle amplitude is as follows: Based on the athlete's historical free throw data, determine the forearm angle amplitude of each historical free throw, and mark the forearm angle amplitude of each historical free throw as: ;in, , The angle of the athlete's forearm at the end of the free throw. The angle of the athlete's forearm at the start of the free throw;

[0063] It should be noted that the forearm angle amplitude can be recorded using high-precision motion capture equipment to capture the changes in the forearm angle during each shooting motion.

[0064] Based on the forearm angle amplitude of each of the athlete's historical free throws, the mean and standard deviation of the historical forearm angle amplitude under successful free throws are determined, and the mean and standard deviation of the historical forearm angle amplitude under successful free throws are denoted as: and The probability density distribution of the historical forearm angle amplitude under successful free throws is calculated using the Gaussian probability density function. The calculation formula is as follows: ;in, The probability of an athlete successfully shooting a free throw with the amplitude of their forearm angle A.

[0065] It should be noted that the probability density distribution of the forearm angle amplitude of each of the athlete's historical free throws represents the athlete's habitual forearm angle during free throws, specifically the common patterns and range of variation of the athlete's forearm angle during free throws.

[0066] Based on the athlete's historical free throw data, determine the number of successful free throws the athlete has under the current forearm angle amplitude, obtain the total number of successful free throws the athlete has, and calculate the probability of the athlete successfully making a free throw under the current forearm angle amplitude. The calculation formula is as follows: ;in, This represents the probability of a player successfully shooting a free throw given the current forearm angle amplitude. This represents the current amplitude of the athlete's forearm angle. This represents the number of successful free throws by the athlete under the current forearm angle amplitude. Given the total number of free throws by the athlete in historical free throw data, determine the number of free throws the athlete can attempt at the current forearm angle amplitude, and calculate the probability of the athlete attempting free throws at the current forearm angle amplitude using the following formula: ;in, This represents the probability of the athlete's current forearm angle amplitude. This represents the number of times a player has failed a free throw in historical free throw data.

[0067] The Bayesian distribution coefficient of angular amplitude is calculated based on Bayes' theorem, and the formula is as follows: ;in, is the Bayesian distribution coefficient of the angular amplitude.

[0068] As can be seen from the formula, the larger the Bayesian distribution coefficient of the angle amplitude, the closer the current athlete's forearm angle amplitude is to the angle pattern of historical successful free throws, the higher the probability of success, and the higher the athlete's free throw action.

[0069] The advantage of the characteristic velocity anomaly deviation coefficient is that it can quantitatively reflect the velocity fluctuations of athletes at various key characteristic points during free throws, helping to identify subtle and hard-to-detect movement anomalies. By monitoring the characteristic velocity anomaly deviation coefficient, athletes can better control the movement speed of key parts during free throws.

[0070] The logic for obtaining the characteristic velocity anomaly deviation coefficient is as follows: The velocities of different feature points in the athlete's forearm during the free throw process are collected to obtain the velocity time series of these different feature points. The velocity and time of different feature points are then fitted using the least squares method to obtain a function of velocity versus time for each feature point. This function is then labeled as follows: ;

[0071] It should be noted that during the free throw, motion capture technology (such as IMU sensors, optical tracking devices, etc.) is used to record the positional changes of different key feature points in the athlete's forearm, thereby determining the velocity of the feature points. By nonlinearly fitting the relationship between velocity and time, the function of each feature point versus time is obtained.

[0072] Set speed range thresholds for different feature points, obtain the time periods during which different feature points exceed the speed range thresholds during the free throw process, and calculate the abnormal deviation coefficient of the feature speed. The calculation formula is as follows: ;in, This is the characteristic velocity anomaly deviation coefficient. The influence weights of velocity deviation at different feature points in the athlete's forearm are determined. The time period during which feature point i exceeds the maximum speed threshold during the free throw process. This refers to the time period during which feature point i exceeds the minimum speed threshold during the penalty kick process.

[0073] It should be noted that the upper and lower limits of the speed thresholds are different for different feature points. Each feature point (such as the wrist, elbow, etc.) will have different motion characteristics and speed patterns during the free throw process, so their speed ranges will be different. Setting the upper and lower limits for each feature point is to more accurately describe the range of motion of each key part during the free throw process, thereby improving the accuracy and effectiveness of the analysis. The speed range thresholds for different feature points are set by professionals in the field.

[0074] As can be seen from the formula, the larger the abnormal deviation coefficient of the characteristic velocity, the more likely the athlete has a problem with the control of the forearm during the free throw process, resulting in abnormal changes in the characteristic velocity and potentially a worse performance in the free throw.

[0075] This study comprehensively analyzes the personalized and actual performance information of athletes' forearm movements. By weighting the normalized displacement characteristic regularization deviation coefficient, angular amplitude Bayesian distribution coefficient, and characteristic velocity anomaly deviation coefficient, a performance evaluation model is constructed, generating performance evaluation coefficients. The formula for calculating the performance evaluation coefficients is as follows: ;in, For performance evaluation coefficients, , , These are the proportional coefficients for the displacement characteristic regularity deviation coefficient, the Bayesian distribution coefficient of the angular amplitude, and the characteristic velocity anomaly deviation coefficient, respectively. , , All are greater than 0.

[0076] As can be seen from the formula, the larger the displacement characteristic regularity deviation coefficient, the angle amplitude Bayesian distribution coefficient, and the characteristic velocity abnormal deviation coefficient, the smaller the performance evaluation coefficient. This indicates that there is a certain difference between the athlete's forearm and the normal pattern during the free throw process, that is, the athlete's shooting action is not stable enough, which affects the accuracy of the free throw and ultimately affects the free throw percentage. Coaches need to remind athletes to maintain the stability and coordination of their forearms and ensure that the forearm action is as consistent as possible during each free throw.

[0077] Set a performance evaluation coefficient threshold, obtain the athlete's current free throw performance evaluation coefficient, and compare the performance evaluation coefficient with the performance evaluation coefficient threshold. If the performance evaluation coefficient is greater than the performance evaluation coefficient threshold, a warning signal is generated, indicating that the current free throw performance may be far from the best state and needs to be adjusted or improved, reminding the athlete that there may be a problem with the current free throw action and it needs to be adjusted. If the performance evaluation coefficient is less than the performance evaluation coefficient threshold, no warning signal is generated.

[0078] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0079] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0080] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0081] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0082] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0083] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0084] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for analyzing the performance of an athlete's action based on big data analysis, characterized in that, Specifically comprising the following steps: S1: monitor the angle change of the player's forearm and wrist through the sensor, identify the starting moment of the player's free throw, and identify the ending moment of the basketball leaving the player's palm in combination with computer vision technology to determine the player's free throw process; S2: extract the time sequence of the key feature points of the forearm, use clustering analysis to analyze the best trajectory of each shooting action, calculate the minimum cost path between the current free throw trajectory and the best trajectory through the dynamic time warping algorithm, and based on the amplitude of the forearm angle of the player's each free throw, calculate the success probability of the current forearm angle amplitude through the Bayes theorem to determine the individualized performance information of the player's forearm action; S3: by collecting the speed time sequence of different feature points of the player's forearm during the free throw process, and using the least square method to fit the relationship between speed and time, the actual performance information of the player's forearm action is determined; S4: through the comprehensive evaluation of the individualized performance information and the actual performance information of the player's forearm action, the influence of the forearm on the free throw is quantified, and a warning signal is generated.

2. The method of claim 1, wherein the method is characterized by, The individualized performance information and the actual performance information of the player's forearm action are determined, comprising: The personalized performance information of the small arm movement of the athlete is expressed by a displacement characteristic regular deviation coefficient and an angle amplitude Bayesian distribution coefficient, and the actual performance information of the small arm movement of the athlete is expressed by a characteristic speed abnormal deviation coefficient, wherein, is the displacement characteristic regular deviation coefficient, is the angle amplitude Bayesian distribution coefficient, is the characteristic speed abnormal deviation coefficient.

3. The method of claim 2, wherein the method further comprises: The displacement feature regularity deviation coefficient acquisition logic is: Based on the historical penalty kick data of the player, the successful penalty kick data of the player is screened, for each successful penalty kick action of the player, the feature points of the key positions of the small arm of the player before and after the penalty kick are extracted, and a time sequence of the feature points is formed, and the time sequence of the feature points is expressed as: ; wherein t is a time point, i is a number of the feature point, i = 1, 2, 3, …, I, I is a positive integer, x, y and z are space coordinates of the feature point respectively. According to the time sequence of the feature points, the time sequence of the feature points is analyzed by clustering algorithm, each shooting action of the player is divided into several clusters, each cluster represents a specific shooting action mode, and the clustering center of each cluster is obtained as the best trajectory of each shooting action. According to the current player's free throw shooting action mode, the best trajectory belonging to the current player's free throw shooting action mode is determined, the best trajectory of the feature point in the best trajectory is marked as: , , and the feature point trajectory of the current player's free throw is marked as: , , the minimum cost path between the feature point trajectory of the current player's free throw and the best trajectory of the feature point is determined using dynamic time warping algorithm, and the minimum cost path is marked as: ; The displacement feature regularity correction coefficient is calculated, and the calculation formula is: .

4. The method of claim 3, wherein the method further comprises: The angle amplitude Bayes distribution coefficient acquisition logic is: Based on historical penalty kick data of the player, determine a historical elbow angle amplitude of the player for each penalty kick, mark the historical elbow angle amplitude of the player for each penalty kick as: ; wherein, , is an angle of the elbow of the player at the end of the penalty kick, is an angle of the elbow of the player at the beginning of the penalty kick; Based on the amplitude of the elbow angle of each free throw in the history of the player, the average value and the standard deviation of the amplitude of the elbow angle of the player in the history of the successful free throw are determined, and the average value and the standard deviation of the amplitude of the elbow angle of the player in the history of the successful free throw are marked as: and The probability density distribution of the amplitude of the elbow angle of the player in the history of the successful free throw is calculated by the Gaussian probability density function, and the calculation formula is: ; wherein, is the probability of the amplitude A of the elbow angle of the player in the successful free throw. Based on the historical penalty kick data of the player, the number of times that the player successfully kicks the penalty kick under the current small arm angle amplitude is determined, and the total number of times that the player successfully kicks the penalty kick is obtained, the probability that the player successfully kicks the penalty kick under the current small arm angle amplitude is calculated, and the calculation formula is: ; wherein, is the probability that the player successfully kicks the penalty kick under the current small arm angle amplitude, is the current small arm angle amplitude of the player, is the number of times that the player successfully kicks the penalty kick under the current small arm angle amplitude, is the total number of times that the player kicks the penalty kick in the historical penalty kick data, the number of times that the player kicks the penalty kick under the current small arm angle amplitude is determined, and the probability that the player kicks the penalty kick under the current small arm angle amplitude is calculated, and the calculation formula is: ; wherein, is the probability that the player kicks the penalty kick under the current small arm angle amplitude, is the number of times that the player fails to kick the penalty kick in the historical penalty kick data; The angle amplitude Bayesian distribution coefficient is calculated based on the Bayesian theorem, and the calculation formula is: .

5. The method of claim 4, wherein the method further comprises: The feature speed abnormal deviation coefficient acquisition logic is: Collect the speed of different feature points in the small arm of the player in the process of the penalty kick, obtain the speed time sequence of different feature points in the small arm of the player in the process of the penalty kick, fit the speed of different feature points and time through the least square method, obtain the function of the speed of different feature points and time, mark the function of the speed of different feature points and time as: ; A speed range threshold of different feature points is set, a time period in which different feature points exceed the speed range threshold in the penalty kick process is obtained, a feature speed abnormal deviation coefficient is calculated, and a calculation formula is as follows: ; wherein, is an influence weight of speed deviation of different feature points in the small arm of the athlete, is a time period in which the feature point i exceeds the maximum speed threshold in the penalty kick process, is a time period in which the feature point i exceeds the minimum speed threshold in the penalty kick process.

6. The method of claim 5, wherein the method further comprises: The comprehensive evaluation of the individualized performance information and the actual performance information of the player's forearm action comprises: The personalized performance information and the actual performance information of the small arm movement of the athlete are comprehensively analyzed, a performance evaluation model is constructed by weighted calculation on the normalized displacement characteristic regular deviation coefficient, the angle amplitude Bayesian distribution coefficient and the characteristic velocity abnormal deviation coefficient, and a performance evaluation coefficient is generated, and the calculation formula of the performance evaluation coefficient is: ; wherein, is the performance evaluation coefficient, , , respectively are the proportional coefficients of the displacement characteristic regular deviation coefficient, the angle amplitude Bayesian distribution coefficient and the characteristic velocity abnormal deviation coefficient, , , all are greater than 0.

7. The method of claim 6, wherein the method further comprises: The warning signal is generated, comprising: Set the performance evaluation coefficient threshold, obtain the performance evaluation coefficient of the player's current free throw, compare the performance evaluation coefficient with the performance evaluation coefficient threshold, if the performance evaluation coefficient is greater than the performance evaluation coefficient threshold, a warning signal is generated, if the performance evaluation coefficient is less than the performance evaluation coefficient threshold, no warning signal is generated.

Images