Athlete action assessment method and system
By categorizing athletes based on body shape data and adjusting thresholds based on differences in motor function, keyframes are dynamically selected for evaluation. This solves the problem of evaluation bias caused by individual differences in existing technologies and achieves more accurate athlete movement evaluation.
Patent Information
- Application Number
- CN202511186292.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-23
- Publication Date
- 2025-12-05
- Estimated Expiration
- 2045-08-23
AI Technical Summary
Existing methods for evaluating athlete performance fail to effectively account for individual differences, resulting in insufficient accuracy, especially in evaluations based on pre-set uniform standards and random time points.
By classifying target athletes based on their body shape data, benchmark movement samples are determined. The threshold for the rate of change of movement is adjusted based on differences in motor function. Keyframes are dynamically selected for evaluation, and the evaluation results are determined by combining the movement data of benchmark athletes.
It enables individualized movement assessment, improves the accuracy and relevance of the assessment, ensures that the assessment results match the individual characteristics of the athletes, and reduces errors caused by individual differences.
Smart Images

Figure CN121074984A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application belongs to the technical field of data analysis, and more particularly relates to a method and system for evaluating the action of an athlete. BACKGROUND
[0002] The evaluation of the action of an athlete is a key technology in sports training, sports rehabilitation and the improvement of competitive level, which collects and analyzes the action parameters of an athlete to determine the standardization, effectiveness and potential injury risk of the action, and provides a basis for training program optimization.
[0003] With the improvement of competitive sports level and the growth of public fitness demand, the accuracy and objectivity of action evaluation are increasingly required. In the prior art, the action evaluation methods mainly include the following types: first, automatic evaluation based on a preset unified standard, which specifically analyzes the action of an athlete by setting fixed evaluation criteria, but this type of method does not consider individual differences of athletes, and has a large deviation when evaluating the action of different athletes; second, dynamic evaluation based on key frames, which specifically analyzes by selecting fixed time nodes or random time nodes as key frames, which lacks rationality and accuracy, affecting the pertinence and accuracy of evaluation.
[0004] Therefore, a scientific and reasonable action evaluation method for athletes that adapts to individual differences is needed to improve the accuracy of action evaluation of athletes. SUMMARY
[0005] The purpose of the present application is to provide a method and system for evaluating the action of an athlete to improve the accuracy of action evaluation of athletes.
[0006] The first aspect of the embodiment of the present application provides a method for evaluating the action of an athlete, comprising: determining a target athlete category and a target athlete category corresponding benchmark action sample based on target body type data of a target athlete, the target athlete category corresponding benchmark action sample including motion function parameters of a benchmark athlete and action data of the benchmark athlete; determining motion function difference data based on target motion function parameters of the target athlete and the motion function parameters of the benchmark athlete; the target athlete is an athlete to be evaluated for action; adjusting an initial action change rate threshold based on the motion function difference data to obtain a target action change rate threshold; the target action change rate threshold is a critical value for judging action deformation; obtaining target action data of the target athlete, calculating an action change rate based on the target action data; the target action data includes action parameters of a plurality of consecutive frames, and determining a target key frame from the plurality of consecutive frames based on the size relationship between the action change rate and the target action change rate threshold; The action evaluation result is determined based on the action parameter corresponding to the target key frame and the action data of the benchmark athlete.
[0007] In a second aspect, the embodiment of the present application provides an athlete action evaluation system, comprising: The athlete category division module is configured to determine a target athlete category and target athlete category corresponding benchmark action samples based on target athlete body shape data, wherein the target athlete category corresponding benchmark action samples comprise motion function parameters of the benchmark athletes and action data of the benchmark athletes; and determine motion function difference data based on target athlete motion function parameters and motion function parameters of the benchmark athletes; the target athlete is an athlete to be evaluated for action; The deformation threshold determination module is configured to adjust the initial action change rate threshold based on the motion function difference data to obtain a target action change rate threshold; the target action change rate threshold is a critical value for judging action deformation; The key frame extraction module is configured to obtain target action data of the target athlete, calculate an action change rate based on the target action data; the target action data comprises action parameters of a plurality of continuous frames, and determine a target key frame from the plurality of continuous frames based on a size relationship between the action change rate and the target action change rate threshold; The action evaluation module is configured to determine an action evaluation result based on the action parameter corresponding to the target key frame and the action data of the benchmark athlete.
[0008] In a third aspect, the embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and running on the processor; when the processor executes the computer program, the steps of the above-mentioned athlete action evaluation method are implemented.
[0009] In a fourth aspect, the embodiment of the present application provides a computer readable storage medium, which stores a computer program; when the computer program is executed by a processor, the steps of the above-mentioned athlete action evaluation method are implemented.
[0010] The athlete action evaluation method and system provided by the embodiment of the present application have the following beneficial effects: The embodiment of the present application determines the athlete category and corresponding benchmark action samples based on the body shape data of the athlete, and calculates the motion function difference data between the athlete and the benchmark, so that the evaluation benchmark matches the individual characteristics of the athlete, solves the evaluation deviation problem caused by the unified standard, and improves the individual adaptability of the evaluation.
[0011] The embodiment of the application adjusts the initial motion change rate threshold value based on the motion function difference data to obtain a target threshold value, so that the critical value for judging motion deformation is adapted to the individual motion function characteristics of the athlete, and the limitations of the fixed threshold value are avoided, thereby providing a reasonable basis for accurate selection of key frames. The motion data of the athlete in the embodiment of the application is calculated to determine the target key frame based on the motion change rate and the target threshold value, and compared with the fixed or random time node selection point, the method can dynamically capture the key moment related to individual motion deformation, so that the selection of key frames is more targeted and accurate.
[0012] The embodiment of the application determines the evaluation result based on the target key frame parameter and the benchmark data. Since the previous steps provide an individual-adapted reference, a reasonable judgment threshold value and accurate key frames for evaluation, the final evaluation result is more scientific and accurate, and the problem of insufficient evaluation accuracy in the prior art is effectively solved. BRIEF DESCRIPTION OF DRAWINGS
[0013] In order to more clearly illustrate the technical solutions in the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only some embodiments of the application, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of these drawings.
[0014] Figure 1 A flowchart of a motion evaluation method for an athlete provided by an embodiment of the application is shown in the figure. Figure 2 A structural block diagram of a motion evaluation system for an athlete provided by an embodiment of the application is shown in the figure. Figure 3 A schematic block diagram of an electronic device provided by an embodiment of the application is shown in the figure. DETAILED DESCRIPTION
[0015] In the following description, specific details such as specific system structures, techniques, etc. are presented in order to thoroughly understand the embodiments of the application. However, it should be clear to those skilled in the art that the application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits and methods are omitted to avoid unnecessary details that hinder the description of the application.
[0016] In order to make the purpose, technical solutions and advantages of the application more clear, the following will be described by specific embodiments in conjunction with the drawings.
[0017] Please refer to Figure 1 , Figure 1A flowchart of an athlete action evaluation method provided by an embodiment of the present application is shown. The method can be executed by an electronic device. Specifically, the method can include S101-S104.
[0018] S101: Determine a target athlete category and target athlete category corresponding benchmark action samples based on target athlete body type data, the target athlete category corresponding benchmark action samples including a benchmark athlete's motor function parameters and the benchmark athlete's action data; determine motor function difference data based on the target athlete's target motor function parameters and the benchmark athlete's motor function parameters; the target athlete is an athlete to be evaluated.
[0019] In this embodiment, before determining the target athlete category and the target athlete category corresponding benchmark action samples based on the target athlete body type data, the method further includes: Obtain body type data of multiple athletes, and cluster the body type data of the multiple athletes to obtain multiple athlete category subsets, each athlete category subset corresponding to a cluster center body type data and an athlete category; Obtain body type data, motor function parameters and action data of multiple benchmark athletes; For each benchmark athlete: calculate the matching degree of the benchmark athlete's body type data and the cluster center body type data of each athlete category subset, and take the athlete category corresponding to the athlete category subset to which the cluster center body type data with the highest matching degree of the benchmark athlete's body type data belongs as the athlete category corresponding to the benchmark athlete; take the benchmark athlete's motor function parameters and action data as a benchmark action sample, and the benchmark action sample and the athlete category corresponding to the benchmark athlete have a corresponding relationship.
[0020] In this embodiment, the target athlete category is determined based on the target athlete body type data, specifically including: calculating the matching degree of the target athlete's target body type data and the cluster center body type data of each athlete category subset, respectively; taking the athlete category corresponding to the athlete category subset to which the cluster center body type data with the highest matching degree of the target athlete's target body type data belongs as the target athlete category.
[0021] In the embodiment, the target body type data refers to the body type characteristic data of the athlete to be evaluated, which is used for category division, and can include height, limb length (upper arm length, thigh length, etc.), muscle girth (biceps girth, quadriceps girth, etc.), and the like. The movement function parameter refers to the function characteristic data related to action execution exhibited by the athlete during movement, and can include joint range of motion (such as shoulder joint flexion angle range, knee joint extension angle range), muscle force exertion characteristic parameter (such as muscle activation intensity, force exertion peak value, and force exertion duration), and the like. The action data refers to dynamic action characteristic data generated by the athlete during execution of a specific action, and can include joint angle change sequence over time (such as elbow joint angle change in a throwing action), body part movement trajectory (such as ankle joint spatial path during running), and duration of each stage of the action (such as length of time during the approach phase of a high jump), and the like. The action data is a dynamic record reflecting the action execution process, and can be obtained in real time by means of an action capture system such as an optical action capture camera or an inertial measurement unit.
[0022] The athlete category subset refers to each grouping set obtained by clustering the body type data of a plurality of athletes, each subset containing athletes with similar body type characteristics, and is used to determine the standard range of the same body type. Each athlete category subset corresponds to an athlete category, such as any one of categories 1 to 4. The clustering center body type data refers to the central characteristic data of each athlete category subset, which is a representative value of the body type data of all athletes in the subset, and can be obtained based on the average height and average limb length of all athletes in the subset, for example. The benchmark athlete refers to an athlete with high action standards and competitive level, and the data of the benchmark athlete can be used as a reference standard for evaluation. The matching degree refers to the degree of similarity between two body type data, which can be calculated based on the degree of difference between the body type parameters. The benchmark action sample refers to a set of standard action related data of the benchmark athlete, including movement function parameters (such as joint range of motion and muscle force exertion intensity) and action data (such as joint angle change data and action timing data) of the benchmark athlete, and is used as a benchmark for action evaluation of athletes with the same body type.
[0023] In the embodiment, the body type data of a plurality of athletes is grouped by a clustering algorithm to obtain a plurality of athlete category subsets and corresponding clustering center body type data, thereby achieving body type classification of athletes. The body type data of the benchmark athlete is compared with the clustering center body type data to determine the category to which each benchmark athlete belongs, and a benchmark action sample corresponding to the category is constructed, thereby establishing a mapping between the sample and the category. The category to which the target athlete belongs can be determined by matching the body type data of the target athlete with the clustering center body type data, and the corresponding benchmark action sample can be obtained, thereby providing an adaptive benchmark for subsequent evaluation.
[0024] The consideration behind the present embodiment is that the differences in body types of different athletes will cause differences in the reasonable range of movements, and the use of a unified standard for evaluation will easily produce deviations. The present embodiment classifies athletes with similar body types into a category through clustering, which can make the movement standards of the same category more targeted; the present embodiment takes the movement data of the benchmark athletes as the standard sample of the same category, which can ensure the professionalism and reasonableness of the evaluation benchmark; the present embodiment establishes a mapping relationship to realize the rapid matching of the target athlete category and the benchmark movement sample, thereby improving the evaluation efficiency; at the same time, the present embodiment can further refine the differences in individual functional characteristics under the same body type through the calculation of the movement function difference data, thereby laying a foundation for subsequent threshold adjustment and accurate evaluation.
[0025] For example, the specific implementation process of the present embodiment can include: The present embodiment can obtain body type data of multiple athletes and perform clustering. Specifically, the present embodiment can collect body type data of 100 athletes with different body types through a three-dimensional human body scanner, including height, upper arm length, thigh length, biceps circumference, and quadriceps circumference, etc.; the present embodiment can process these data using a clustering algorithm (such as K-means algorithm), set the number of clusters (such as determined as 4 categories through the contour coefficient), and obtain 4 athlete category subsets; the present embodiment can calculate the mean value of the body type parameters of all athletes in each subset as the clustering center body type data of the subset, and assign a unique athlete category identifier to each subset, such as category 1 to category 4.
[0026] The present embodiment can process the data of the benchmark athletes and establish a mapping relationship. Specifically, the present embodiment can select 10 benchmark athletes of the sports project, and collect their body type data in the same way. Then, an optical motion capture system is used to collect movement function parameters in cooperation with a wireless electromyography sensor. When collecting the movement function parameters, reflective marker points can be pasted on the key joints (shoulder, elbow, hip, knee, and ankle) of the benchmark athletes, and the electromyography sensor is attached to the target muscle (such as the midpoint of the quadriceps muscle) according to the anatomy. During the movement function parameter collection process, the benchmark athletes perform a specific movement, assuming that the specific movement is a knee range of motion test, which requires the benchmark athletes to slowly flex from the fully extended position to the maximum angle, and then slowly stretch back to the original position, repeating 3 times; during the movement function parameter collection process, the electromyography sensor synchronously records the muscle activation signals of the benchmark athletes, and the motion capture system records the joint angle changes. The present embodiment can calculate the joint range of motion (such as the maximum flexion angle of the knee) and the muscle force peak value (such as the maximum value of the root mean square value) through the collected data, and take the average value of 3 tests as the final movement function parameter.
[0027] The embodiment can collect action data of the pole athletes by using an inertial motion capture system. During the action data collection, the sensors are fixed to the corresponding positions of the athletes' head, torso and limbs (e.g., feet, shanks, thighs, pelvis, etc.), and after calibration, the pole athletes are required to complete specific movement actions, and 3 groups of effective actions are continuously collected. The device synchronously records the joint angle (e.g., the hip joint flexion angle, the ankle joint plantar flexion angle), the body center of gravity trajectory (three-dimensional coordinates) and the action stage time length (e.g., the time from the start to the take-off in the run-up stage) and other data at each time point, and finally the action parameters containing continuous frames are obtained after preprocessing.
[0028] For each pole athlete, the embodiment can calculate the matching degree of the body shape data of the pole athlete and the body shape data of the 4 cluster centers, such as calculating the sum of the absolute differences of each body shape parameter, and the smaller the difference is, the higher the matching degree is. The embodiment can take the category to which the cluster center with the highest matching degree belongs as the category of the pole athlete, such as a certain pole athlete matching category 2. The embodiment can integrate the movement function parameters and the action data of all the pole athletes in the category, take the reasonable range (such as the 25%-75% quantile of the joint angle) of each parameter as the pole action sample of the category, and establish a mapping relationship between the sample and category 2 and store it in the database.
[0029] The embodiment can determine the category of the target athlete and the corresponding pole action sample. Specifically, the embodiment can collect the body shape data of the target athlete, such as a height of 176 cm, an upper arm length of 59 cm, etc. The embodiment can calculate the difference between the body shape data and the body shape data of the 4 cluster centers respectively, and the smaller the difference is, the greater the matching degree is. For example, the sum of the differences of each data in the body shape data and the body shape data of the cluster center of category 1 is 8, and the sum of the differences of each data in the body shape data and the body shape data of the cluster center of category 2 is 3. The embodiment can take category 2 with the smallest sum of differences and the highest matching degree as the category of the target athlete. According to the mapping relationship, the embodiment can call the corresponding pole action sample of category 2 from the database. At the same time, the embodiment can collect the movement function parameters of the target athlete, and compare them with the movement function parameters of the pole athletes in the pole action sample, such as the knee joint range of motion of the target athlete being 130° and that of the pole athletes being 135°. The embodiment can calculate the movement function difference data, such as 5°, to provide a basis for subsequent threshold adjustment.
[0030] S102: adjusting the initial action change rate threshold based on the movement function difference data to obtain a target action change rate threshold; the target action change rate threshold is a critical value for judging action deformation.
[0031] In the embodiment, the initial action change rate threshold includes an initial joint change rate threshold and an initial muscle change rate threshold; the initial joint change rate threshold and the initial muscle change rate threshold are obtained by the following way: extracting a joint angle change rate and a muscle force change rate based on the action data of the benchmark athlete corresponding to the target athlete category; performing probability density estimation on the joint angle change rate to obtain a joint change rate density curve, and determining an initial joint change rate threshold based on the joint change rate density curve; performing probability density estimation on the muscle force change rate to obtain a muscle change rate density curve, and determining an initial muscle change rate threshold based on the muscle change rate density curve.
[0032] In this embodiment, the initial action change rate threshold refers to a basic critical value for judging whether the action is deformed without adjustment, including the initial joint change rate threshold and the initial muscle change rate threshold. The initial joint change rate threshold refers to an initial critical value for judging whether the joint action is deformed, which is determined based on the joint angle change characteristics of the benchmark athlete, and can include, for example, a knee joint angle change rate threshold, an elbow joint angle change rate threshold, etc. The initial muscle change rate threshold refers to an initial critical value for judging whether the muscle force is abnormal, which is determined based on the muscle force intensity change characteristics of the benchmark athlete, and can include, for example, a quadriceps muscle force intensity change rate threshold, a biceps muscle force intensity change rate threshold, etc. The joint angle change rate refers to the rate of change of the joint angle with time, which is used to characterize the fast and slow changes of joint movement, and can include, for example, the number of degrees per second of change of the knee joint angle during flexion. The muscle force intensity change rate refers to the rate of change of muscle force intensity with time, which is used to characterize the dynamic adjustment speed of muscle force, and can include, for example, the amount of change per second of the quadriceps muscle force intensity during the action of kicking the ground. The joint change rate density curve is used to reflect the distribution characteristics of the joint angle change rate. The muscle change rate density curve is used to reflect the distribution characteristics of the muscle force intensity change rate.
[0033] In this embodiment, considering that the action change rate distribution of the benchmark athlete can reflect the dynamic characteristics of normal action under the same body type, the initial threshold determined based on the density curve thereof can be used as a basic standard; however, there are individual differences in motor function (such as joint range of motion, muscle strength) among athletes in the same category, and directly using the initial threshold can lead to judgment deviation; and the embodiment adjusts the threshold through motor function difference data, which can adapt the critical value to the actual functional level of the target athlete, for example, the reasonable action change rate of an athlete with a smaller joint range of motion is lower, and the adjusted threshold is more likely to accurately capture the action deformation; the final target threshold can improve the accuracy of action deformation judgment and provide a reliable basis for subsequent key frame determination.
[0034] For example, the specific implementation process of the embodiment can include: The embodiment can extract the action change rate of the benchmark athlete. Specifically, the embodiment can call the action data of the benchmark athlete stored in the benchmark action sample corresponding to the target athlete category (such as category 2), which includes the joint angle (such as the angle value of the knee joint every 10 ms) and muscle strength (such as the strength value of the quadriceps muscle every 10 ms) of the continuous frames; the embodiment can calculate the joint angle change rate, that is, the difference value of the joint angle of the adjacent two frames divided by the time interval, such as the knee joint angle of the nth frame is 30°, the knee joint angle of the n+1th frame is 35°, and the time interval is 0.01 s, and the change rate is 500° / s; the embodiment can calculate the muscle strength change rate, that is, the difference value of the muscle strength of the adjacent two frames divided by the time interval, such as the muscle strength of the quadriceps muscle of the nth frame is 200 μV, the muscle strength of the quadriceps muscle of the n+1th frame is 250 μV, and the change rate is 5000 μV / s; the embodiment can collect the above change rate data of all benchmark athletes in this category to form a joint angle change rate set and a muscle strength change rate set.
[0035] The embodiment can determine the initial action change rate threshold. Specifically, the embodiment can perform probability density estimation on the joint angle change rate set to generate a joint change rate density curve, and the peak value of the curve corresponds to the joint angle change rate value with the highest occurrence frequency; the embodiment can select a critical value that can cover 95% of the normal action change rate as the initial joint change rate threshold based on the curve characteristics, such as 600° / s corresponding to the 95% quantile point of the curve, that is, if the value exceeds the value, it is determined to be a deformation action; the embodiment can process the muscle strength change rate set in the same way to obtain an initial muscle change rate threshold (such as 8000 μV / s); and the two initial thresholds are combined into an initial action change rate threshold.
[0036] In the embodiment, the target movement function parameter includes a target joint activity parameter and a target muscle strength characteristic parameter; the movement function parameter includes a joint activity parameter and a muscle strength characteristic parameter; the movement function difference data is determined based on the target movement function parameter of the target athlete and the movement function parameter of the benchmark athlete, and specifically includes: determining the joint activity parameter difference based on the target joint activity parameter of the target athlete and the joint activity parameter of the benchmark athlete; determining the muscle strength characteristic parameter difference based on the target muscle strength characteristic parameter of the target athlete and the muscle strength characteristic parameter of the benchmark athlete; and taking the joint activity parameter difference and the muscle strength characteristic parameter difference as the movement function difference data.
[0037] In the embodiment, the initial action change rate threshold is adjusted based on the movement function difference data, and specifically includes: determining a function difference coefficient based on the movement function difference data, and adjusting the initial action change rate threshold based on the function difference coefficient.
[0038] In the embodiment, the functional difference coefficient includes a joint activity difference coefficient and a muscle force difference coefficient; the functional difference coefficient is determined based on the motion function difference data, and the initial motion change rate threshold is adjusted based on the functional difference coefficient, specifically including: determining the joint activity difference coefficient based on the joint activity parameter difference, and adjusting the initial joint change rate threshold based on the joint activity difference coefficient; determining the muscle force difference coefficient based on the muscle force feature parameter difference, and adjusting the initial muscle change rate threshold based on the muscle force difference coefficient.
[0039] For example, taking the joint activity parameter as an example, the knee joint activity parameter of the benchmark athlete can be obtained in the embodiment: the lower limit of the range is 80°, the upper limit of the range is 120°, and the range width is 40°; the knee joint activity parameter of the target athlete is called: the lower limit of the range is 75°, the upper limit of the range is 110°, and the range width is 35°.
[0040] The embodiment can make a difference for the obtained parameters one by one, and calculate the lower limit difference of the range 5° (80°-75°), the upper limit difference of the range 10° (120°-110°), and the range width difference 5° (40°-35°), to form the joint activity parameter difference.
[0041] The embodiment can perform weighted summation according to the range width weight 0.5, the range upper and lower limit weight 0.25 to obtain the joint activity difference coefficient: the lower limit coefficient of the range = 1-5° / 80°=0.9375, the upper limit coefficient of the range = 1-10° / 120°≈0.9167, and the range width coefficient = 1-5° / 40°=0.875; the joint activity difference coefficient = 0.25×0.9375+0.25×0.9167+0.5×0.875≈0.901.
[0042] The embodiment can adjust the initial joint change rate threshold based on the joint activity difference coefficient. Specifically, the embodiment can call the initial joint change rate threshold 600° / s, multiply the joint activity difference coefficient 0.901, and obtain the target joint change rate threshold ≈541° / s, which is adapted to the knee joint function characteristics of the target athlete.
[0043] For example, taking the muscle force feature parameter as an example, the quadriceps muscle force feature parameter of the target athlete can be obtained in the embodiment, specifically including: muscle force peak value 270μV, force rising rate 140μV / s, and force duration 0.9s. The embodiment can call the quadriceps muscle force feature parameter of the benchmark athlete, specifically including: muscle force peak value 300μV, force rising rate 160μV / s, and force duration 0.8s. The embodiment can calculate the muscle force peak difference 30μV (300μV-270μV), the muscle force rising rate difference 20μV / s (160μV / s-140μV / s), and the muscle force duration difference -0.1s (0.8s-0.9s, the negative value indicates that the muscle force duration of the target athlete is longer) for the obtained parameters, to form the muscle force characteristic parameter difference.
[0044] The embodiment can perform weighted summation according to the muscle force peak weight 0.4, the muscle force rising rate weight 0.3, and the muscle force duration weight 0.3 to obtain the muscle force difference coefficient. The muscle force peak coefficient = 1-30μV / 300μV = 0.9, the muscle force rising rate coefficient = 1-20μV / s / 160μV / s = 0.875, and the muscle force duration coefficient = 1-|-0.1s| / 0.8s = 0.875. The muscle force difference coefficient = 0.4×0.9+0.3×0.875+0.3×0.875 = 0.885.
[0045] The embodiment can adjust the initial muscle change rate threshold according to the muscle force difference coefficient. Specifically, the embodiment can call the initial muscle change rate threshold 8000μV / s, multiply the muscle force difference coefficient 0.885, and obtain the target muscle change rate threshold ≈7080μV / s, which is suitable for the quadriceps femoris function characteristics of the target athlete.
[0046] The embodiment can combine the two adjusted thresholds to obtain the target action change rate threshold, which is used for subsequent judgment of the action deformation of the target athlete.
[0047] S103: Obtain target action data of the target athlete, and calculate an action change rate based on the target action data. The target action data includes action parameters of a plurality of continuous frames. A target key frame is determined from the plurality of continuous frames based on a size relationship between the action change rate and the target action change rate threshold.
[0048] In the embodiment, the target action data refers to continuous dynamic data generated when the target athlete performs an action, including action parameters of a plurality of continuous frames, for example, a sequence of joint angles, muscle force intensity, etc. The action parameter refers to a specific characteristic value constituting the target action data, for example, the knee joint angle, the biceps brachii muscle force intensity, etc. The action change rate refers to the change rate of the action parameter with time, which is used to represent the dynamic change speed of the action parameter, for example, the degree number of the knee joint angle changed per second, the numerical value of the muscle force intensity changed per second, etc. The target key frame refers to a frame selected from the plurality of continuous frames, whose action change rate exceeds the target action change rate threshold, which is used to represent the key moment of the action deformation.
[0049] The consideration behind the embodiment is that the motion parameters of continuous frames can record the motion process completely, and the motion change rate can reflect the dynamic fluctuation of the motion; the target motion change rate threshold is the critical value of the target athlete, which can distinguish normal motion fluctuation from abnormal deformation; the embodiment can filter the key frames by comparing the change rate with the threshold, accurately capture the moment of motion deformation, avoid the blindness of fixed time point frame selection, and ensure that the key frames can truly reflect the motion defects, providing accurate analysis objects for subsequent evaluation.
[0050] For example, the specific implementation process of the embodiment can include: The embodiment can obtain target motion data of a target athlete. Specifically, the embodiment can use an inertial motion capture system to obtain the target motion data of the target athlete. During the collection process, the inertial motion capture system can collect data at a sampling rate of 100 Hz, generate motion parameters of continuous frames, including hip flexion angle, gastrocnemius muscle force intensity, etc., store them as time series data, and form the target motion data.
[0051] The embodiment can calculate the motion change rate based on the target motion data. Specifically, the embodiment can extract the motion parameters of continuous frames in the target motion data, for example, select the knee joint angle parameters: the 10th frame angle is 30°, the 11th frame is 35°, the time interval is 0.01 s, and the joint angle change rate of the frame is calculated as (35°-30°) / 0.01s=500° / s; similarly, the muscle force intensity parameters are extracted, the 10th frame of the quadriceps muscle force is 200μV, and the 11th frame is 230μV, and the muscle force intensity change rate is calculated as (230μV-200μV) / 0.01s=3000μV / s; the motion change rates of all continuous frames are calculated in turn to form a change rate sequence.
[0052] The embodiment can determine the target key frame. Specifically, the embodiment can call the target motion change rate threshold obtained before, such as the joint change rate threshold 570° / s and the muscle change rate threshold 7200μV / s; the embodiment can compare the motion change rate with the corresponding threshold frame by frame, for example, the 25th frame of the knee joint angle change rate is 600° / s (exceeding 570° / s), and the 40th frame of the quadriceps muscle force change rate is 7500μV / s (exceeding 7200μV / s); the embodiment can mark all frames exceeding the threshold such as the 25th frame and the 40th frame as target key frames, record their time points and corresponding motion parameters, and use them for subsequent evaluation and analysis.
[0053] S104: determining a motion evaluation result based on the motion parameters corresponding to the target key frame and the motion data of the benchmark athlete.
[0054] In the embodiment, the action evaluation result includes a first action evaluation result and a second action evaluation result; the action evaluation result is determined based on the action parameter corresponding to the target key frame and the action data of the benchmark athlete, and specifically includes: calculating a first difference between the action parameter corresponding to the target key frame and the action data of the benchmark athlete, and determining the first action evaluation result based on the first difference; adjusting the action data of the benchmark athlete based on the motor function difference data, calculating a second difference between the action parameter corresponding to the target key frame and the adjusted action data of the benchmark athlete, and determining the second action evaluation result based on the second difference.
[0055] In the embodiment, the first action evaluation result refers to an evaluation result obtained based on the difference between the target key frame action parameter and the unadjusted benchmark athlete action data, and is used to reflect the deviation degree of the action from the original benchmark standard, for example, can include joint angle deviation score, muscle force synchronization score, etc. The second action evaluation result refers to an evaluation result obtained based on the difference between the target key frame action parameter and the benchmark athlete action data adjusted by the motor function difference data, and is used to reflect the actual standardization of the action after excluding individual function difference, for example, can include corrected joint angle matching degree, force intensity adaptability, etc.
[0056] The first difference refers to the difference between the action parameter corresponding to the target key frame and the unadjusted benchmark athlete action data, for example, can include the difference between the knee joint angle of the target key frame and the angle of the same frame of the benchmark athlete, the absolute difference of muscle force intensity, etc. The second difference refers to the difference between the action parameter corresponding to the target key frame and the adjusted action data of the benchmark athlete, for example, can include the difference between the elbow joint angle of the target key frame and the benchmark angle adjusted by the function difference, the difference after the force intensity correction, etc.
[0057] The consideration behind the embodiment is that the unadjusted benchmark data represents the standard of the same type, but there is a motor function difference (such as different joint range of motion and muscle strength) between the target athlete and the benchmark athlete, and direct comparison will exaggerate or reduce the action defects; the first evaluation result can intuitively show the gap between the action and the original standard, and provide a basis for overall reference; and the benchmark data is adjusted by the motor function difference data, which can adapt the comparison benchmark to the function level of the target athlete, and the second evaluation result can more accurately reflect the technical defects of the action itself, avoiding misjudgment due to individual function limitation; the combination of the two can reflect the gap between the action and the absolute standard, and highlight the real technical problems, and improve the objectivity and pertinence of the evaluation.
[0058] For example, the specific implementation process of the embodiment can include: (1) The embodiment can obtain the action parameters of the target key frame. Specifically, the embodiment can extract the action parameters from the labeled target key frame, for example, the knee joint angle of the 25th frame is 42°, and the quadriceps muscle strength is 260μV, the hip joint angle of the 40th frame is 58°, and the gluteus muscle strength is 310μV. The embodiment can extract the benchmark data of the same action stage as the target key frame from the benchmark action sample corresponding to the target athlete category, such as the benchmark knee joint angle of the 25th frame is 40°, and the quadriceps muscle strength is 250μV, the benchmark hip joint angle of the 40th frame is 55°, and the gluteus muscle strength is 300μV. These data are the median value of the standard action parameter range of the benchmark athlete of the same body type.
[0059] The embodiment can calculate the difference between the target parameters and the benchmark data: the knee joint angle difference of the 25th frame is 2°, and the strength difference is 10μV; the hip joint angle difference of the 40th frame is 3°, and the strength difference is 10μV. A preset scoring rule is used, such as 2 points deducted for every 1° of angle difference, and 1 point deducted for every 10μV of strength difference. Finally, 5 points are deducted for the 25th frame, and 7 points are deducted for the 40th frame, with a comprehensive score of 88 points. At the same time, the difference details can be labeled, such as “the knee joint angle of the 25th frame is slightly larger, and the quadriceps muscle strength is slightly stronger”, to form the action evaluation result and provide specific direction for training correction.
[0060] For example, the specific implementation process of the embodiment can further include: The embodiment can calculate the first difference and determine the first action evaluation result. Specifically, the embodiment can call the action parameters corresponding to the target key frame (such as the 25th frame and the 40th frame), including the knee joint angle of the 25th frame 45°, and the quadriceps muscle strength 280μV, the hip joint angle of the 40th frame 60°, and the gluteus muscle strength 320μV; the embodiment can extract the parameters of the same key frame in the action data of the benchmark athlete corresponding to the target athlete category, such as the benchmark knee joint angle of the 25th frame 40°, and the quadriceps muscle strength 250μV, the benchmark hip joint angle of the 40th frame 55°, and the gluteus muscle strength 300μV; the embodiment can calculate the first difference, such as the knee joint angle difference of the 25th frame 5°, and the strength difference 30μV, the hip joint angle difference of the 40th frame 5°, and the strength difference 20μV; the embodiment can calculate the first action evaluation result (such as 13 points for the 25th frame, 12 points for the 40th frame, and a comprehensive score of 75 points) based on a preset scoring rule (such as 2 points deducted for every 1° of angle difference, and 1 point deducted for every 10μV of strength difference).
[0061] The embodiment can adjust the action data of the benchmark athlete. Specifically, the embodiment can call the previously obtained motion function difference data, such as that the knee joint range of motion of the target athlete is 5° less than the benchmark (the maximum of the benchmark is 40°, and the maximum of the target is 35°), and the corresponding adjustment coefficient is 0.875 (35° / 40°); the peak value of the quadriceps femoris of the target athlete is 10% lower than the benchmark (the peak value of the benchmark is 250 μV, and the peak value of the target is 225 μV), and the adjustment coefficient is 0.9 (225 μV / 250 μV); the embodiment can adjust the action data of the benchmark athlete, such as the knee joint angle of the benchmark in the 25th frame is 40° x 0.875 = 35° (adapted to the maximum range of motion of the target), and the quadriceps femoris force is 250 μV x 0.9 = 225 μV (adapted to the strength of the target); the hip joint angle of the benchmark in the 40th frame is 55°, it is assumed that there is no difference in hip joint function, and the corresponding adjustment coefficient is 1.0, so the hip joint angle of the benchmark is still 55°, and the gluteus maximus force is 300 μV (it is assumed that there is no difference) and is still 300 μV; the adjusted action data of the benchmark athlete is obtained.
[0062] The embodiment can calculate the second difference and determine the second action evaluation result. Specifically, the embodiment can calculate the second difference between the target key frame parameters and the adjusted benchmark data, such as the knee joint angle of the target athlete in the 25th frame is 34° (close to the maximum range of motion of 35°), and the quadriceps femoris force is 220 μV (close to the peak value of 225 μV), and the difference between the adjusted benchmark data is 34°-35°=-1°, and the force difference is 220 μV-225 μV=-5 μV; the hip joint angle of the target athlete in the 40th frame is 54°, and the gluteus maximus force is 290 μV, and the difference between the adjusted benchmark data is 54°-55°=-1°, and the force difference is 290 μV-300 μV=-10 μV; the embodiment can calculate using the same scoring rules, such as 2 points for each 1° of angle difference, and 1 point for each 10 μV of force difference; then the 25th frame is deducted 1 x 2 + 5 / 10 x 1 = 2 + 0.5 = 2.5 points, and the 40th frame is deducted 1 x 2 + 10 / 10 x 1 = 2 + 1 = 3 points, and the comprehensive score is 94.5 points; the embodiment can integrate the first action evaluation result (85 points) and the second action evaluation result (94.5 points) to generate a final action evaluation report, wherein the first result reflects the difference from the original benchmark (the score is lower due to the function limitation), and the second result reflects the actual technical level excluding the individual function difference (the score is higher, and the target athlete's action specification within the self ability range is better), which provides a more practical reference for training improvement.
[0063] From the above, the embodiment determines the athlete category and the corresponding benchmark action sample through the body data of the athlete, and calculates the movement function difference data of the athlete and the benchmark, so that the evaluation benchmark matches the individual characteristics of the athlete, solves the evaluation deviation problem caused by the unified standard, and improves the individual adaptability of the evaluation.
[0064] The embodiment adjusts the initial action change rate threshold value based on the movement function difference data to obtain a target threshold value, so that the critical value for judging the action deformation adapts to the individual movement function characteristics of the athlete, avoids the limitation of the fixed threshold value, and provides a reasonable basis for accurate selection of the key frame. The action data of the athlete in the embodiment is calculated to determine the target key frame based on the action change rate and the target threshold value. Compared with fixed or random time node selection, this method can dynamically capture the key moment related to individual action deformation, so that the key frame selection is more targeted and accurate.
[0065] The embodiment determines the evaluation result based on the target key frame parameter and the benchmark data. Since the previous steps provide an individual-adapted benchmark, a reasonable judgment threshold value and an accurate key frame for evaluation, the final evaluation result is more scientific and accurate, effectively solving the problem of insufficient evaluation accuracy in the prior art.
[0066] In an embodiment of the present application, the initial action change rate threshold value includes an initial joint change rate threshold value and an initial muscle change rate threshold value, and the target action change rate threshold value includes a target joint change rate threshold value and a target muscle change rate threshold value. Adjusting the initial action change rate threshold value based on the movement function difference data to obtain a target action change rate threshold value, comprising: Adjusting the initial joint change rate threshold value based on the movement function difference data to obtain a target joint change rate threshold value. Adjusting the initial muscle change rate threshold value based on the movement function difference data to obtain a target muscle change rate threshold value.
[0067] In the embodiment, the target joint change rate threshold value refers to the critical value adapted to the joint function characteristics of the target athlete obtained by adjusting the initial joint change rate threshold value based on the movement function difference data, which can include a change rate threshold value adapted to the knee joint of the target athlete, etc. The target muscle change rate threshold value refers to the critical value adapted to the muscle function characteristics of the target athlete obtained by adjusting the initial muscle change rate threshold value based on the movement function difference data, which can include a change rate threshold value adapted to the quadriceps muscle of the target athlete, etc.
[0068] The consideration behind this embodiment is that joint movement and muscle effort are two core dimensions of action execution, and their functional characteristics differ, such as individual differences in joint range of motion and muscle strength intensity come from different sources; although the initial joint change rate threshold and the initial muscle change rate threshold are determined based on the same type of benchmark athletes, the target athletes may have specific differences in joint function and muscle function; if a unified adjustment method is used for both, it is difficult to accurately adapt to individual characteristics, which may lead to threshold deviation in one dimension; this embodiment can make the critical values of joints and muscles respectively fit the actual functional level of the target athletes by adjusting based on the motion function difference data, ensuring that the action deformation judgment is accurate and reliable in both dimensions, and providing a more scientific basis for subsequent key frame determination.
[0069] For example, the specific implementation process of this embodiment can include: This embodiment can obtain the motion function difference data of the target athlete. Specifically, this embodiment can call the previously calculated motion function difference data, including joint function differences (such as the target athlete's knee joint range of motion being 8° smaller than the benchmark athlete, and the elbow joint range of motion being the same as the benchmark) and muscle function differences (such as the target athlete's quadriceps muscle force peak being 15% lower than the benchmark athlete, and the biceps muscle force peak being the same as the benchmark).
[0070] This embodiment can adjust the initial joint change rate threshold to obtain the target joint change rate threshold. Specifically, this embodiment can call the initial joint change rate threshold, such as the knee joint initial threshold of 600° / s and the elbow joint initial threshold of 500° / s; for the knee joint, an adjustment coefficient is calculated according to its range of motion difference, such as (benchmark range of motion - 8°) / benchmark range of motion = 0.92, and this embodiment can adjust the initial threshold to 600° / s x 0.92 = 552° / s; the elbow joint has no functional difference, the adjustment coefficient is 1.0, and the initial threshold remains 500° / s; finally, this embodiment can obtain the target joint change rate threshold: knee joint 552° / s, elbow joint 500° / s.
[0071] The embodiment can adjust the initial muscle change rate threshold to obtain a target muscle change rate threshold. Specifically, the embodiment can call the initial muscle change rate threshold, such as the initial threshold of quadriceps femoris 8000 μV / s and the initial threshold of biceps brachii 6000 μV / s. For the quadriceps femoris, the embodiment can calculate an adjustment coefficient according to the difference in peak force, such as (benchmark peak force-15% x benchmark peak force) / benchmark peak force=0.85. The embodiment can adjust the initial threshold to 8000 μV / s x 0.85=6800 μV / s. The biceps brachii has no functional difference, the adjustment coefficient is 1.0, and the initial threshold remains 6000 μV / s. Finally, the embodiment can obtain the target muscle change rate threshold: 6800 μV / s for the quadriceps femoris and 6000 μV / s for the biceps brachii. The embodiment can integrate the target joint change rate threshold and the target muscle change rate threshold to form a target action change rate threshold, which is used for subsequent comparison of action change rate and determination of key frame.
[0072] The embodiment adjusts the initial joint and muscle change rate thresholds based on the motor function difference data respectively, so that the two types of target thresholds accurately adapt to the joint and muscle function characteristics of the target athlete, avoids dimensional deviation caused by single adjustment, ensures that the action deformation judgment is reliable in two dimensions, provides a scientific basis for key frame determination, and improves the evaluation accuracy.
[0073] In an embodiment of the present application, the target action data includes a joint angle sequence and a muscle force intensity sequence, and the action change rate includes a joint angle change rate and a muscle force intensity change rate. The action change rate is calculated based on the target action data, including: The joint angle change rate is calculated based on the joint angle sequence. The joint angle sequence includes joint angle parameters of a plurality of consecutive frames. The muscle force intensity change rate is calculated based on the muscle force intensity sequence. The muscle force intensity sequence includes muscle force intensity parameters of a plurality of consecutive frames.
[0074] In this embodiment, the joint angle sequence refers to a sequence composed of joint angle parameters of consecutive multiple frames in chronological order, used to record the dynamic changes of joint angles completely, which may include, for example, the sequence of continuous angle values of the knee joint, elbow joint, hip joint, etc. during the action. The muscle strength sequence refers to a sequence composed of muscle strength parameters of consecutive multiple frames in chronological order, used to record the dynamic changes of muscle strength completely, which may include, for example, the sequence of continuous strength values of the quadriceps femoris, biceps brachii, gluteus maximus, etc. during the action. The joint angle parameter refers to the specific angle value corresponding to a single frame in the joint angle sequence, used to represent the position state of the joint at a certain time, which may include, for example, the flexion angle of the knee joint, the extension angle of the elbow joint, etc. in a certain frame. The muscle strength parameter refers to the specific strength value corresponding to a single frame in the muscle strength sequence, used to represent the activation state of the muscle at a certain time, which may include, for example, the strength of the quadriceps femoris, the strength of the biceps brachii, etc. in a certain frame.
[0075] The consideration behind this embodiment is that joint angle and muscle strength are two core elements of action, and their dynamic changes reflect the morphological characteristics and dynamic characteristics of action respectively, with different physiological and mechanical significance; the joint angle sequence can reflect the continuity of the limb trajectory, and the muscle strength sequence can reflect the power source and regulation process of the action; this embodiment calculates the change rate based on the two types of sequences respectively, which can capture the fast and slow changes of joint movement and the dynamic adjustment of muscle strength, avoiding information loss or deviation caused by mixing the two types of characteristics; this classification calculation method can provide accurate data basis for subsequent comparison with the corresponding target change rate threshold (joint to joint, muscle to muscle), ensuring the pertinence and reliability of action deformation judgment.
[0076] For example, the specific implementation process of this embodiment can include: This embodiment can calculate the joint angle change rate based on the joint angle sequence. Specifically, this embodiment can extract the joint angle sequence in the target action data, such as the knee joint angle sequence, containing angle values of the 1st to 100th frames, with a frame interval of 0.01s; this embodiment can select the joint angle parameters of two consecutive frames, such as the 20th frame of the knee joint angle 30° and the 21st frame of 36°, calculate the angle difference of the two frames as 6°, and divide by the time interval 0.01s to obtain the joint angle change rate of this frame as 600° / s; this embodiment can calculate the joint angle change rate of all consecutive frames in the sequence in turn to form the joint angle change rate sequence.
[0077] The embodiment can calculate the muscle force strength change rate based on the muscle force strength sequence. Specifically, the embodiment can extract the muscle force strength sequence in the target action data, such as the quadriceps force strength sequence, which contains force values of the first to 100th frames, and each frame interval is 0.01 s. The embodiment can select the muscle force strength parameters of two consecutive frames, such as the 20th frame of the quadriceps force 200 μV and the 21st frame of 260 μV, calculate the force difference of the two frames as 60 μV, and divide by the time interval 0.01 s to obtain the muscle force strength change rate of the frame as 6000 μV / s. The embodiment can sequentially calculate the muscle force strength change rates of all consecutive frames in the sequence to form a muscle force strength change rate sequence.
[0078] The embodiment can accurately capture the morphological dynamics and dynamic dynamics of the action by calculating the corresponding change rates based on the joint angle sequence and the muscle force strength sequence respectively, provide targeted data for subsequent comparison with the target change rate threshold, avoid deviation caused by feature confusion, improve the accuracy and pertinence of the action change rate calculation, and lay a foundation for accurate determination of the key frame.
[0079] In an embodiment of the present application, the action change rate includes a joint angle change rate and a muscle force strength change rate, and the target action change rate threshold includes a target joint change rate threshold and a target muscle change rate threshold. The target key frame is determined from the plurality of consecutive frames based on the size relationship between the action change rate and the target action change rate threshold, comprising: For each frame in the plurality of consecutive frames, if the joint angle change rate corresponding to the frame exceeds the target joint change rate threshold, and / or the muscle force strength change rate corresponding to the frame exceeds the target muscle change rate threshold, the frame is marked as a target key frame.
[0080] In the embodiment, the frame is a time point in the action data sequence, which is used to detect whether the action change rate exceeds the threshold in real time. The consideration behind the embodiment is that the joint angle change rate and the muscle force strength change rate respectively reflect the morphological dynamics and dynamic dynamics of the action, and the action deformation types indicated by the two abnormalities are different, such as joint angle abnormality may be trajectory deviation, and muscle force abnormality may be power deficiency. If only a single dimension is used to determine the key frame, the deformation moment of another dimension will be missed. Using the logic of marking as long as any one exceeds, the abnormal moment of the joint or muscle can be captured comprehensively, and the key frame of all potential action defects is ensured to be identified, providing a complete analysis object for subsequent evaluation.
[0081] For example, the specific implementation process of the embodiment can include that the embodiment can call a joint angle change rate sequence (such as the knee joint change rate of the first to the 100th frame, which is 500° / s, 550° / s, etc. in turn), a muscle strength change rate sequence (such as the quadriceps femoris change rate of the first to the 100th frame, which is 6500 μV / s, 7000 μV / s, etc. in turn), and a target motion change rate threshold (knee joint 552° / s, quadriceps femoris 6800 μV / s).
[0082] The embodiment can start checking from the first frame: the knee joint change rate of the 20th frame is 560° / s (exceeding 552° / s), which is marked as a target key frame; the quadriceps femoris change rate of the 30th frame is 6900 μV / s (exceeding 6800 μV / s), which is marked as a target key frame; neither of the 40th frame exceeds the threshold, which is not marked; the judgment of all frames is completed in turn, and finally the target key frame set (such as the 20th frame, the 30th frame, etc.) is obtained.
[0083] The embodiment marks the key frame by judging whether the joint and muscle change rates exceed the corresponding threshold, can comprehensively capture the key moment of the two types of motion deformation, avoids the omission of single dimension judgment, ensures the integrity of the key frame, and provides a reliable basis for accurate evaluation.
[0084] A method for evaluating an athlete's motion corresponding to the above embodiment, Figure 2 A structural block diagram of an athlete motion evaluation system provided by an embodiment of the present application is shown. For ease of illustration, only parts related to the embodiments of the present application are shown. For reference Figure 2 The athlete motion evaluation system 20 includes an athlete category division module 21, a deformation threshold determination module 22, a key frame extraction module 23, and a motion evaluation module 24.
[0085] The athlete category division module 21 is configured to determine a target athlete category and a target athlete category corresponding benchmark motion sample based on target body type data of a target athlete, the target athlete category corresponding benchmark motion sample including motion function parameters of a benchmark athlete and motion data of the benchmark athlete, determine motion function difference data based on target motion function parameters of the target athlete and the motion function parameters of the benchmark athlete, and the target athlete is an athlete to be evaluated for motion. The deformation threshold determination module 22 is configured to adjust an initial motion change rate threshold based on the motion function difference data to obtain a target motion change rate threshold, and the target motion change rate threshold is a critical value for judging motion deformation. The key frame extraction module 23 is configured to obtain target action data of the target athlete, calculate an action change rate based on the target action data, and determine a target key frame from the continuous multiple frames based on a size relationship between the action change rate and a target action change rate threshold, wherein the target action data comprises action parameters of the continuous multiple frames. The action evaluation module 24 is configured to determine an action evaluation result based on the action parameters corresponding to the target key frame and the action data of the benchmark athlete.
[0086] In an embodiment of the present application, the athlete action evaluation system 20 further comprises: The data preparation module is configured to: obtain body shape data of multiple athletes, cluster the body shape data of the multiple athletes to obtain multiple athlete category subsets, and each athlete category subset corresponds to a cluster center body shape data and an athlete category; obtain body shape data, movement function parameters and action data of multiple benchmark athletes; For each benchmark athlete: calculate a matching degree between the body shape data of the benchmark athlete and the cluster center body shape data of each athlete category subset, take an athlete category corresponding to an athlete category subset to which the cluster center body shape data with the highest matching degree with the body shape data of the benchmark athlete belongs as an athlete category corresponding to the benchmark athlete, and take the movement function parameters and the action data of the benchmark athlete as a benchmark action sample, wherein the benchmark action sample and the athlete category corresponding to the benchmark athlete have a corresponding relationship.
[0087] In an embodiment of the present application, the athlete category division module 21 is specifically configured to: calculate a matching degree between the target body shape data of the target athlete and the cluster center body shape data of each athlete category subset respectively, and take an athlete category corresponding to an athlete category subset to which the cluster center body shape data with the highest matching degree with the target body shape data of the target athlete belongs as the target athlete category.
[0088] In an embodiment of the present application, the initial action change rate threshold comprises an initial joint change rate threshold and an initial muscle change rate threshold, and the deformation threshold determination module 22 is specifically configured to: extract a joint angle change rate and a muscle strength change rate based on the action data of the benchmark athlete corresponding to the target athlete category; perform probability density estimation on the joint angle change rate to obtain a joint change rate density curve, and determine the initial joint change rate threshold based on the joint change rate density curve; perform probability density estimation on the muscle strength change rate to obtain a muscle change rate density curve, and determine the initial muscle change rate threshold based on the muscle change rate density curve.
[0089] In one embodiment of this application, the initial motion change rate threshold includes an initial joint change rate threshold and an initial muscle change rate threshold, and the target motion change rate threshold includes a target joint change rate threshold and a target muscle change rate threshold; the deformation threshold determination module 22 is further configured to: The initial joint rate of change threshold is adjusted based on the motor function difference data to obtain the target joint rate of change threshold; The initial muscle change rate threshold is adjusted based on motor function difference data to obtain the target muscle change rate threshold.
[0090] In one embodiment of this application, the target motion data includes a joint angle sequence and a muscle exertion intensity sequence, and the motion change rate includes the joint angle change rate and the muscle exertion intensity change rate; the keyframe extraction module 23 is specifically used for: calculating the joint angle change rate based on the joint angle sequence; the joint angle sequence includes joint angle parameters of multiple consecutive frames; calculating the muscle exertion intensity change rate based on the muscle exertion intensity sequence; the muscle exertion intensity sequence includes muscle exertion intensity parameters of multiple consecutive frames.
[0091] In one embodiment of this application, the rate of change of motion includes the rate of change of joint angle and the rate of change of muscle exertion intensity, and the target rate of change of motion threshold includes the target joint rate of change threshold and the target muscle rate of change threshold; the keyframe extraction module 23 is specifically used to: for each frame in a series of consecutive frames, if the rate of change of joint angle corresponding to the frame exceeds the target joint rate of change threshold, and / or, the rate of change of muscle exertion intensity corresponding to the frame exceeds the target muscle rate of change threshold, then the frame is marked as a target keyframe.
[0092] See Figure 3 , Figure 3 This is a schematic block diagram of an electronic device provided according to an embodiment of this application. Figure 3 The electronic device 300 in this embodiment may include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memories 304 store computer programs, including program instructions. The processors 301 execute the program instructions stored in the memories 304. Specifically, the processors 301 are configured to invoke the program instructions to perform the functions of the modules in the aforementioned device embodiments, for example... Figure 2 The functions of the athlete category classification module 21, deformation threshold determination module 22, keyframe extraction module 23, and motion evaluation module 24 are shown.
[0093] It should be appreciated that in the embodiments of the present application, the processor 301 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.
[0094] The input device 302 can include a touchpad, a fingerprint collection sensor (for collecting fingerprint information and direction information of a fingerprint of a user), a microphone, etc., and the output device 303 can include a display (LCD, etc.), a speaker, etc.
[0095] The memory 304 can include a read-only memory and a random access memory, and provide instructions and data for the processor 301. A portion of the memory 304 can also include a nonvolatile random access memory. For example, the memory 304 can also store information of a player category.
[0096] In specific implementations, the processor 301, the input device 302 and the output device 303 described in the embodiments of the present application can execute the implementation manners described in the embodiments of the player action evaluation method provided by the embodiments of the present application, and can also execute the implementation manners of the electronic device 300 described in the embodiments of the present application, which will not be described herein again.
[0097] In another embodiment of the present application, a computer readable storage medium is provided, which stores a computer program. The computer program includes program instructions, which, when executed by a processor, implement all or part of the processes of the above-mentioned embodiment methods. The computer program can also instruct related hardware to complete the implementation. The computer program can be stored in a computer readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various method embodiments can be implemented. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate form. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.
[0098] The computer readable storage medium can be an internal storage unit of the electronic device, such as a hard disk or a memory of the electronic device. The computer readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the computer readable storage medium can include both the internal storage unit and the external storage device of the electronic device. The computer readable storage medium is used to store the computer program and other programs and data required by the electronic device. The computer readable storage medium can also be used to temporarily store data that has been output or will be output.
[0099] Those skilled in the art can appreciate that the modules / units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. In order to clearly illustrate the interchangeability of hardware and software, the components and steps of the examples have been described in general terms in the above description. Whether the functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. A person 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 the present application.
[0100] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the electronic device and the units described above can refer to the corresponding processes in the above-mentioned method embodiments, which will not be described here.
[0101] In several embodiments provided in the present application, it should be understood that the disclosed electronic device and method can be implemented in other manners. For example, the division of the above-described apparatus embodiments is merely an example, and there can be other division manners. For example, multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be indirect couplings or communication connections through some interfaces, or can be in electrical, mechanical or other forms.
[0102] The modules or units illustrated as separate parts can or can not be physically separate, and the parts illustrated as modules or units can or can not be physical modules or units, i.e., can be located in one place, or can be distributed on multiple network modules or units. Some or all of the modules or units can be selected according to actual needs to achieve the purposes of the embodiments of the present application.
[0103] In addition, each functional module / unit in each embodiment of the present application can be integrated into one processing module / unit, or each module / unit can exist physically separately, or two or more modules / units can be integrated into one module / unit. The integrated module / unit can be implemented in the form of hardware or in the form of a software functional module / unit.
[0104] The above is merely specific embodiments of the present application, and the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.
Claims
1. A method of assessing the performance of an athlete, characterized in that, The method comprises the following steps: determining a target athlete category and a benchmark action sample corresponding to the target athlete category based on target body type data of a target athlete, wherein the benchmark action sample corresponding to the target athlete category comprises motion function parameters of a benchmark athlete and action data of the benchmark athlete; determining motion function difference data based on target motion function parameters of the target athlete and the motion function parameters of the benchmark athlete; the target athlete is an athlete to be evaluated for action; adjusting an initial action change rate threshold based on the motion function difference data to obtain a target action change rate threshold; the target action change rate threshold is a critical value for judging action deformation; obtaining target action data of the target athlete, and calculating an action change rate based on the target action data; the target action data comprises action parameters of a plurality of continuous frames, and a target key frame is determined from the plurality of continuous frames based on a size relationship between the action change rate and the target action change rate threshold; determining an action evaluation result based on action parameters corresponding to the target key frame and the action data of the benchmark athlete.
2. A method of assessing the performance of an athlete as claimed in claim 1 wherein, Before the step of determining a target athlete category and a benchmark action sample corresponding to the target athlete category based on target body type data of a target athlete, the method further comprises the following steps: obtaining body type data of a plurality of athletes, and clustering the body type data of the plurality of athletes to obtain a plurality of athlete category subsets, wherein each athlete category subset corresponds to a clustering center body type data and an athlete category; obtaining body type data, motion function parameters and action data of a plurality of benchmark athletes; for each benchmark athlete, calculating a matching degree of the body type data of the benchmark athlete and the clustering center body type data of each athlete category subset, and taking an athlete category corresponding to an athlete category subset to which the clustering center body type data having the highest matching degree with the body type data of the benchmark athlete belongs as an athlete category corresponding to the benchmark athlete; taking the motion function parameters and the action data of the benchmark athlete as a benchmark action sample, and there is a corresponding relationship between the benchmark action sample and the athlete category corresponding to the benchmark athlete.
3. A method of assessing the performance of an athlete as claimed in claim 2, wherein, The step of determining a target athlete category based on target body type data of a target athlete comprises the following steps: calculating a matching degree of the target body type data of the target athlete and the clustering center body type data of each athlete category subset, respectively; taking an athlete category corresponding to an athlete category subset to which the clustering center body type data having the highest matching degree with the target body type data of the target athlete belongs as the target athlete category.
4. A method of assessing the performance of an athlete as claimed in claim 1, wherein, The initial action change rate threshold comprises an initial joint change rate threshold and an initial muscle change rate threshold; The initial joint change rate threshold and the initial muscle change rate threshold are obtained by the following method: extracting a joint angle change rate and a muscle strength change rate based on the action data of the benchmark athlete corresponding to the target athlete category; performing probability density estimation on the joint angle change rate to obtain a joint change rate density curve, and determining the initial joint change rate threshold based on the joint change rate density curve; The probability density estimation is performed on the muscle strength change rate to obtain a muscle change rate density curve, and the initial muscle change rate threshold is determined based on the muscle change rate density curve.
5. A method of assessing the performance of an athlete as claimed in claim 1, wherein, The initial motion change rate threshold includes an initial joint change rate threshold and an initial muscle change rate threshold, and the target motion change rate threshold includes a target joint change rate threshold and a target muscle change rate threshold. The adjustment of the initial motion change rate threshold based on the motion function difference data to obtain the target motion change rate threshold includes: The initial joint change rate threshold is adjusted based on the motion function difference data to obtain the target joint change rate threshold. The initial muscle change rate threshold is adjusted based on the motion function difference data to obtain the target muscle change rate threshold.
6. A method of assessing the performance of an athlete as claimed in claim 1, wherein, The target motion data includes a joint angle sequence and a muscle strength sequence, and the motion change rate includes a joint angle change rate and a muscle strength change rate. The calculation of the motion change rate based on the target motion data includes: The joint angle change rate is calculated based on the joint angle sequence; the joint angle sequence includes joint angle parameters of a plurality of continuous frames. The muscle strength change rate is calculated based on the muscle strength sequence; the muscle strength sequence includes muscle strength parameters of a plurality of continuous frames.
7. A method of assessing the performance of an athlete as claimed in claim 1, wherein, The motion change rate includes a joint angle change rate and a muscle strength change rate, and the target motion change rate threshold includes a target joint change rate threshold and a target muscle change rate threshold. The determination of the target key frame from the plurality of continuous frames based on the size relationship between the motion change rate and the target motion change rate threshold includes: For each frame in the plurality of continuous frames, if the joint angle change rate corresponding to the frame exceeds the target joint change rate threshold, and / or, the muscle strength change rate corresponding to the frame exceeds the target muscle change rate threshold, the frame is marked as a target key frame.
8. An athlete performance assessment system characterised in that, It includes: The athlete category division module is configured to determine a target athlete category and a benchmark motion sample corresponding to the target athlete category based on target body type data of a target athlete, wherein the benchmark motion sample corresponding to the target athlete category includes motion function parameters of a benchmark athlete and motion data of the benchmark athlete. The motion function difference data is determined based on target motion function parameters of the target athlete and the motion function parameters of the benchmark athlete; the target athlete is an athlete to be evaluated for motion; The deformation threshold determination module is configured to adjust an initial motion change rate threshold based on the motion function difference data to obtain a target motion change rate threshold; the target motion change rate threshold is a critical value for judging motion deformation. The key frame extraction module is configured to obtain target motion data of the target athlete, calculate a motion change rate based on the target motion data, and determine a target key frame from a plurality of continuous frames based on a size relationship between the motion change rate and the target motion change rate threshold; the target motion data includes motion parameters of the plurality of continuous frames. An action evaluation module is configured to determine an action evaluation result based on the action parameter corresponding to the target key frame and the action data of the benchmark athlete.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, The computer program is executed by the processor to implement the steps of the method according to any one of claims 1 to 7.
10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. The computer program is executed by the processor to implement the steps of the method according to any one of claims 1 to 7.
Citation Information
Patent Citations
Standard evaluation method and system for sports actions of sports personnel
CN118015315A
Athlete action recognition method based on posture evaluation
CN118470806A
Exercise evaluation system and method based on multi-dimensional data analysis
CN118986339A
Pull-up process action decomposition and key frame extraction method and computer equipment
CN119068555A
Cleaning personnel work supervision method based on machine vision analysis
CN119672384A
Cited By
Sports interdisciplinary interaction system and method based on multi-modal alignment
CN122290221A
