WO2021189736A1 - Procédé et système de notation de séance d'exercices - Google Patents

Procédé et système de notation de séance d'exercices Download PDF

Info

Publication number
WO2021189736A1
WO2021189736A1 PCT/CN2020/104094 CN2020104094W WO2021189736A1 WO 2021189736 A1 WO2021189736 A1 WO 2021189736A1 CN 2020104094 W CN2020104094 W CN 2020104094W WO 2021189736 A1 WO2021189736 A1 WO 2021189736A1
Authority
WO
WIPO (PCT)
Prior art keywords
student
coach
pane
exercise
stability
Prior art date
Application number
PCT/CN2020/104094
Other languages
English (en)
Chinese (zh)
Inventor
庄龙飞
Original Assignee
庄龙飞
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from CN202010228332.0A external-priority patent/CN113449945B/zh
Application filed by 庄龙飞 filed Critical 庄龙飞
Publication of WO2021189736A1 publication Critical patent/WO2021189736A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/20—Education
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/20—Education
    • G06Q50/205—Education administration or guidance

Definitions

  • the present invention relates to a method and system for calculating sports scores by dividing sports courses into multiple sports types.
  • Common exercise judgment systems currently include: using multi-lens camera equipment and computer vision processing technology to capture the user’s movement gestures for analysis, whether it is the cost of hardware or the production of digital content.
  • the cost is high, which makes it difficult to promote;
  • a common way is to use a worn inertial measurement unit, which includes at least an accelerometer and a magnetometer to record the acceleration and angular velocity changes of the sensor in the three-dimensional space during exercise , So as to calculate the movement trajectory in space, and then record and analyze the comparison.
  • a worn inertial measurement unit which includes at least an accelerometer and a magnetometer to record the acceleration and angular velocity changes of the sensor in the three-dimensional space during exercise , So as to calculate the movement trajectory in space, and then record and analyze the comparison.
  • different sports types have different characteristics, so how to properly analyze and compare different sports types is a topic of concern to those skilled in the art.
  • the purpose of the present invention is to provide a sports course scoring method, which adopts different scoring methods for different sports types, so that the sports scores can be calculated more appropriately.
  • the embodiment of the present invention proposes an exercise course scoring method, which is suitable for an exercise course grading system.
  • the exercise course grading system includes a wearable device, and the wearable device includes an inertial measurement unit.
  • the scoring method of this exercise course includes: playing the course video, obtaining the coach movement data corresponding to the course video, and obtaining the coach pane from the coach movement data; obtaining the student movement data through the inertial measurement unit, and obtaining the student pane from the student movement data, where The length of the student pane is greater than the length of the coach pane; find the student segment in the student pane that is most similar to the coach pane; for the first exercise type, calculate the exercise score based on the stability of the student segment and the coach pane; and For the second type of exercise, a deduction mechanism is used to calculate the exercise score based on the difference between the student segment and the coach pane.
  • the start of the student pane is before the start of the coach pane, and the end of the student pane is after the end of the coach pane.
  • the above step of finding the student segment in the student pane that is most similar to the coach pane includes: executing an open start and end dynamic time warping algorithm on the student pane and the coach pane to obtain the student segment.
  • the step of calculating the exercise score according to the stability of the student segment includes: calculating the average absolute error of the force of the student segment at multiple sampling points as the student stability; calculating the coach pane on the multiple sampling points The average absolute error of the strength is used as the stability of the coach; and the stability ratio between the stability of the student and the stability of the coach is calculated.
  • the step of calculating the sports score according to the stability of the student segment further includes: converting the stability ratio into a stable score based on the Logger function; and calculating the corresponding coaching stability for each pane in the coaching sports data The first number of times in the first range and the second number of times in the second range of the coaching stability of the statistics pane; divide the first number by the sum of the first number and the second number to calculate the stability weight ; And calculate the sports score based on the stability score and the stability weight.
  • the step of calculating the sports score based on the stability score and the stability weight is calculated according to the following equation (1), where Score is the sports score, S stability is the stability score, w is the stability weight, and S 1 is a real number .
  • the aforementioned step of using the deduction mechanism to calculate the sports score based on the difference between the student segment and the coach pane includes: calculating the force error and direction error between the student segment and the coach pane; setting the force weight And direction weight, and when the larger of the force error and the direction error is greater than the error threshold, the corresponding force weight or direction weight is increased; and the motion score is calculated according to the force error, direction error, force weight and direction weight.
  • the aforementioned force error is calculated based on the difference between the L2 norm of the student segment at the paired sampling point and the L2 norm of the coach pane at the corresponding sampling point.
  • the above-mentioned direction error is calculated based on the cosine similarity between the paired sampling points of the student segment and the corresponding sampling points of the coach pane.
  • the step of calculating the motion score according to the force error, the direction error, the force weight, and the direction weight is calculated according to the following equation (2), where Score is the motion score, D a is the direction error, and w a is the direction Weight, D m is the force error, w m is the force weight.
  • the embodiment of the present invention provides a sports course scoring system, which includes a wearable device and an intelligent device.
  • the wearable device includes an inertial measurement unit to obtain student motion data.
  • the smart device is used to play the course video through the display and obtain the coach's exercise data corresponding to the course video.
  • the intelligent device is used to perform the following multiple steps or transmit the student's exercise data and coach's exercise data to the calculation module to perform these steps: obtain the student pane from the student's exercise data, and obtain the coach pane from the coach's exercise data, where the student The length of the pane is greater than the length of the coach pane; find the student segment in the student pane that is most similar to the coach pane; for the first exercise type, calculate the exercise score based on the stability of the student segment and the coach pane; and For the second type of exercise, a deduction mechanism is used to calculate the exercise score based on the difference between the student segment and the coach pane.
  • Fig. 1A is a schematic diagram illustrating a sports course scoring system according to an embodiment.
  • FIG. 1B is a partial flowchart of the sports course scoring system 100 according to an embodiment.
  • Fig. 2 is a schematic diagram showing coach exercise data and student exercise data according to an embodiment.
  • Fig. 3 is a flowchart illustrating a method for scoring sports courses according to an embodiment.
  • 110-wearable device 112-student, 120-display, 121-score, 130-intelligent device, 131-processor, 132-memory, 133-wireless communication module, 140-cloud database, 141-computing module, 151 ⁇ 156-step, 210-trainer exercise data, 220-student exercise data, 211,212-coach pane, 221-student pane, 230-student segment, 231,232-sampling point, 301-306-step.
  • the present invention provides a sports course scoring method combining smart devices and cloud application services, which can capture the student's body part movement data while the student exercises according to the sports course video, and compare the student's movement data with the coach The exercise data of the users is compared, and the exercise results of the users are evaluated accordingly. In this way, the students can clearly know whether they perform the required actions correctly, thereby improving the exercise effect and increasing the willingness to exercise.
  • different methods are used to calculate scores, and for some types of sports that are not very important to posture or fast movements, a relatively large tolerance will be given.
  • Fig. 1A is a schematic diagram illustrating a sports course scoring system according to an embodiment.
  • the sports course scoring system 100 includes at least one wearable device 110, a display 120 and a smart device 130.
  • the wearable device 110 may, for example, be implemented as a sports bracelet and worn by the student 112, but in other embodiments may also be implemented as a watch, a strap, or other device that can be worn on the body.
  • the wearable device 110 includes an inertia measurement unit, which includes at least an acceleration sensor to measure acceleration values on three axes, such as X, Y, and Z.
  • the inertia measurement unit may also include angular velocity. Sensors and/or magnetometers.
  • the data measured by the inertial measurement unit is called student motion data.
  • the wearable device 110 also includes a wireless communication module, such as a Bluetooth communication module, a wireless fidelity (WiFi) module, or other suitable low-power wireless transmission modules.
  • the wearable device 110 may further include a display panel or other arbitrary elements, and the invention is not limited thereto.
  • the display 120 is used to play a lesson video, in which a coach is demonstrating an action.
  • the smart device 130 includes a processor 131, a memory 132, and a wireless communication module 133.
  • the memory 132 stores program codes and is executed by the processor 131.
  • the processor 131 may be a central processing unit, a microprocessor, a microcontroller, a digital signal processor, an image processing chip, a special application integrated circuit, etc., and the invention is not limited thereto.
  • the wireless communication module 133 is, for example, a Bluetooth communication module, a wireless fidelity (WiFi) module or other suitable low-power wireless transmission module, and is used to receive the student's exercise data from the wearable device 110.
  • FIG. 1B is a partial flowchart of the sports course scoring system 100 according to an embodiment. Please refer to FIG. 1A and FIG. 1B.
  • a course video must be produced.
  • the course video is shot and the coach's exercise data is recorded at the same time.
  • multiple cameras may be used to shoot the coach's actions first, and these cameras are set up at different positions and angles, and these shot videos can be edited to produce course videos.
  • the coach also brings one or more wearable devices 110 during the shooting, and the signals sensed by the inertial measurement unit are also recorded as coach exercise data.
  • coach sports data includes multiple sampling points, and each sampling point includes three acceleration values such as X, Y, and Z.
  • step 152 post-production of the course film is performed, and the coach exercise data and the course film are synchronized according to the time code.
  • the course videos and the coach's exercise data are drawn and stored in the cloud database 140.
  • the smart device 130 can obtain the course video and coach exercise data from the cloud database 140 and play the course video through the display 120.
  • the smart device 130 obtains the student's exercise data through the wearable device 110 (via the wireless communication module).
  • the processor 131 will start to receive the student's exercise data from the wearable device 110 before the course video starts to be played, but the student's exercise data will be discarded, and the current data will not be stored until the course video starts to be played. Student movement data.
  • the obtained student motion data will correspond to the course video. For example, it can be known which picture in the course video each sampling point in the student motion data corresponds to.
  • step 154 when the exercise is suspended or ended (at the end of the course movie), the acquired exercise data of the student is collected. Since the corresponding relationship between the student's exercise data and the course video has been known in step 153, and the corresponding relationship between the course video and the coach's exercise data can be obtained according to the SMPTE time code, the corresponding relationship can be obtained according to the screen number of the course video The exercise data of the trainees and the exercise data of the coaches.
  • step 155 the coach sports data is divided into a plurality of coach panes, and the sports score of each coach pane is calculated.
  • step 156 the exercise score 121 is displayed, so that the student 112 can clearly know whether he or she performs the required action correctly, thereby improving the exercise effect and increasing the willingness to exercise.
  • the sports score is calculated by the processor 131, but in other embodiments, the sports score may also be calculated by a server on the cloud or other electronic devices.
  • the smart device 130 can be connected to a computing module 141 on the cloud.
  • the computing module 141 can be a server, a virtual machine, or a network application that provides computing services, and the present invention is not limited thereto.
  • the processor 131 may transmit the coach exercise data and the student exercise data to the calculation module 141, and the calculation module 141 calculates the exercise score and then sends it back to the processor 131.
  • the exercise score is calculated after the exercise is stopped or ended.
  • the exercise score can also be calculated while obtaining the student's exercise data, that is, the exercise score can also be displayed in real time.
  • the score calculation in the smart device 130 can be started immediately, or after the exercise, the exercise data of the student can be transmitted to the calculation module 141 to perform the exercise score. calculate.
  • the calculation of the sports score will be described in detail below.
  • FIG. 2 is a schematic diagram illustrating the coach exercise data and the student exercise data according to an embodiment.
  • the coach exercise data 210 and the student exercise data 220 are shown as one-dimensional signals (that is, there is only one acceleration value on each sampling point), but this is only a schematic diagram, in fact, each sampling point should have multiple There are two acceleration values (that is, a vector is formed), and the coach exercise data 210 and the student exercise data 220 are a set of vectors.
  • the coach sports data 210 can be expressed as Where i is a positive integer, representing the i-th sampling point, They respectively represent the acceleration values of the X-axis, Y-axis, and Z-axis of the coach exercise data 210 at the i-th sampling point.
  • the student exercise data 220 is expressed as in They respectively represent the acceleration values of the X-axis, Y-axis, and Z-axis of the student's motion data 220 at the i-th sampling point.
  • the coach exercise data 210 and the student exercise data 220 correspond to the same course video. Therefore, if the coach exercise data 210 and the student exercise data 220 are more similar, the calculated score should be higher.
  • the length of the coach exercise data 210 may be different from the length of the student exercise data 220.
  • the sampling frequency of the student exercise data 220 is different from the sampling frequency of the coach exercise data 210, the one with the larger sampling frequency may be resampling, so that the sampling frequency of the two is the same.
  • the inertial measurement unit also includes an angular accelerometer
  • a six-axis sensor data fusion algorithm can be used to obtain three-dimensional acceleration vectors F c [i] and F s [i], which can properly distinguish The posture of the action and the direction of the action can be used to judge more types of exercises and provide a better discrimination rate.
  • the inertial measurement unit also includes a three-axis magnetometer
  • the nine-axis sensor data fusion algorithm can be used to obtain the three-dimensional acceleration vectors F c [i] and F s [i] more accurately for faster comparison. Calculation of similarity rate with precision.
  • the aforementioned fusion algorithm can use any direction and heading reference system (AHRS) algorithm, and the present invention is not limited thereto.
  • AHRS direction and heading reference system
  • the length of the coach pane 211 can be determined according to different exercise types, and the length can be 2 seconds to 10 seconds, and the present invention is not limited thereto.
  • the student pane 221 is also obtained from the student motion data 220, wherein the length of the student pane 221 is greater than the length of the coach pane 211.
  • the starting point of the student pane 221 may be before the starting point of the coaching pane 211, and the ending point of the student pane 221 may be after the ending point of the coaching pane 211. This is because the student may experience delays when imitating the actions of the coach.
  • a longer length of the student pane 221 can have a greater tolerance for these phenomena.
  • the length of the coach pane 211 is 4 seconds and the sampling frequency is 25 Hz
  • an open start and end dynamic time warping algorithm is performed on the student pane 221 and the coach pane 211.
  • -begin-end dynamic time warping, OBE-DTW -begin-end dynamic time warping
  • OBE-DTW releases the limitation of the same starting point and the same end point, so as to find partially matched fragments from the student pane 221. This algorithm can be expressed as the following equations (1) to (3).
  • the length of can be the same as the length of the student pane 221 or different.
  • This algorithm is to find the positive integers d and e from the student pane 221 so that the error Will be the smallest, where Cost OBE-DTW represents the average of the error between the sampling points paired according to the OBE-DTW, and the paired sampling points in the student pane 221 will form the student segment 230.
  • the sampling point 231 in the student segment 230 is matched to the sampling point 232 in the coach pane 211.
  • the sampling points in the student segment 230 may be duplicated (paired to multiple sampling points in the coach pane 211). Points) and not necessarily continuous (some sampling points are not matched to the coach pane 211). In some embodiments, different errors can be used according to different types. If the evaluation is based on the direction of action, Cost OBE-DTW refers to the average of cosine similarity, if it is based on the force of action. For evaluation, Cost OBE-DTW refers to the average of the L2 norm (also known as the Euler distance). In addition, N matcted refers to the length of the student segment found and a is a real number, which can be determined by experiments (for example, 2).
  • the above equation (3) is to adjust the error according to the length of the student segment
  • the length of the found student segment 230 may be less than or greater than the coach pane 211, so when a segment is closer to the length of the coach pane 211, it will be selected first, which is different from the conventional OBE-DTW.
  • the student’s actions are slightly behind the coach’s actions. Therefore, according to the above algorithm, the first part of the student pane 221 will be discarded, and only the latter half of the student pane 221 will be matched. This way, the backward movement of the student can be given. Greater tolerance.
  • the exercise type of the student segment 230 will be analyzed, and different types of exercises will adopt different score calculation methods.
  • At least two types of movement are set here, the first type of movement refers to a stable static movement, and the second type of movement refers to a general or fast dynamic movement.
  • the absolute value of the acceleration values of the X-axis, Y-axis, and Z-axis at each sampling point can be added to obtain a force, which is expressed as the following equation (4).
  • the acceleration value in the equation (4) may belong to coach movement data or student movement data.
  • the strength of the coach’s exercise data can be used, and then the average, standard deviation, or mean absolute deviation of all strengths x[i] is used to classify the exercise, such as when the average
  • the absolute difference is less than a threshold, it can be determined as the first exercise type, and when the absolute difference is greater than or equal to the threshold, it can be determined as the second exercise type.
  • it can also be classified according to the sports event (such as boxing aerobic, Latin dance, intermittent intensity exercise, yoga, etc.) to which the coach's exercise data belongs, or according to the length of the coach pane 211 (or the student pane 221) To classify.
  • the weight of the first exercise type (between 0 and 1) and the weight of the second exercise type (between 0 and 1) may also be calculated for the student pane 221, and the first exercise type may be calculated separately.
  • the score of one sport type and the score of the second sport type, and then the two scores are added together according to the weight, and different student panes 221 may belong to different sport types or have different sport type weights.
  • the present invention does not limit how to perform sports classification. The following describes the calculation of sports scores in two types of sports.
  • the above average absolute deviation can be expressed as the following equation (5), where the average absolute deviation is referred to as the stability MAD.
  • Equation (5) is used to calculate the stability MAD of the student's motion data, hereinafter referred to as the student's stability MAD user .
  • Equation (5) can also be used to calculate the stability of the coach's sports data. It is only necessary to replace the positive integer N in the equation (5) with M, which is hereinafter referred to as the coach stability MAD ref .
  • the sports score can be calculated according to the student stability MAD user and the coach stability MAD ref . The more similar the two are, the higher the calculated sports score.
  • the stability ratio r between the student stability MAD user and the coach stability MAD ref can be calculated, as shown in the following equation (6).
  • the stability ratio is shifted, scaled, and flipped to convert it into a stable score, as shown in the following equations (7) to (9).
  • S stability (r) represents the stability score.
  • q is a real number used to control the changing trend of equation (8) and can be obtained through experiments.
  • the logit function described above can be implemented in a table lookup manner.
  • this embodiment uses double The threshold method subdivides static motion into true static and fuzzy static. Specifically, two thresholds are defined here, namely the upper bound of the static threshold t h and the true static threshold t, where the upper bound of the static threshold t h is greater than the true static threshold t. These two thresholds determine two ranges, which are The first range [0, t] and the second range (t, t h ). Next, the coach sports data is divided into multiple panes.
  • the corresponding coach stability MAD ref is calculated, and the first time the coach stability MAD ref is in the first range [0, t] is counted. number and a second range (t, t h) a second number of times, and then divided by the sum of the first frequency and the first and second times to calculate a stable weight w, expressed as the following equation (10).
  • the motion score is calculated according to the stability score S stability (r) and the stability weight w, as shown in the following equation (11).
  • Score is the sports score.
  • S 1 is a real number, which can be set through experiments, for example, 90.
  • This type of exercise is based on the deduction mechanism and the exercise score is calculated based on the difference between the student segment and the coach pane.
  • the force error will be shifted and scaled through the sigmoid function.
  • the S function is as shown in the following equation (12).
  • m and q are parameters.
  • the S function can be implemented in a look-up table.
  • the strength of each sampling point in the coach pane can be expressed as the following equation (13).
  • mag upper [i] std(mag c ,i,w)+mean(mag c ,i,w),
  • std(mag c ,i,w) represents the standard deviation of the force mag c [] from the i-th sampling point to the i+w-th sampling point
  • mean(mag c ,i,w) represents The force mag c [] is the average value from the i-th sampling point to the i+w-th sampling point.
  • P represents a set, which includes the sampling points paired between the student segment and the coach pane.
  • This set is generated by the OBE-DTW algorithm.
  • the sampling point 231 in Figure 2 is paired to the sampling point 232, ( i c , i s ) represents each pair in the set P.
  • m m and q m are real numbers, used to control the tolerance of the force difference, which can be set arbitrarily through experiments.
  • D m represents the points to be deducted, which is the aforementioned force error.
  • the dynamic factor mag upper [i c ] is used to adaptively adjust the force error D m according to the coach's exercise data.
  • the force error D m is based on the L2 norm ⁇ F s [i s ] ⁇ of the student segment at each sampling point and the L2 norm of the coach pane at the corresponding sampling point ⁇ F c [ i] ⁇ is calculated.
  • m a, q a is a real number, for controlling the angular difference of tolerance, can be set by any experiment.
  • D a represents the points to be deducted, which is the above-mentioned directional error.
  • the direction error is D a [i s] corresponding to the sample points pane coach in accordance with each sampling point F c fragment student F s [i c] cosine similarity between the computed.
  • the force error D m is used to determine whether the student the correct operation of the force
  • D a direction error is determined to the direction of the correct operation of the student.
  • the force weight w m and the direction weight w a are also set, when the larger one of the force error and the direction error is greater than the error threshold (indicating that the movements of the student and the coach are very different) , Increase the corresponding force weight or direction weight, and decrease the other weight, so that you won’t be given a high score because it happens to have a great similarity in direction or force.
  • the above step of increasing the weight is adjusted according to the S function, as shown in the following equation (17).
  • the sum of the weight w m and right direction strength weights w a is 1, the force of the weight w m w a weight direction is determined according to the following pseudo code.
  • q is a real number, used to determine the magnitude of the weight increase.
  • m is a real number, used to determine the displacement of the S function on the X axis, which represents the above critical value.
  • the real numbers q and m can be set arbitrarily through experiments. In other words, when the direction D a relatively large error, error in a direction D a function S to be input, when D a direction error rises rapidly above the threshold amplitude low rest condition is changed.
  • the sports score of the next coach pane 212 can be calculated.
  • the coach pane 212 and the coach pane 211 partially overlap. For example, you can slide the coach pane 211 to the right by a distance to obtain the coach pane 212, and then calculate the sports score of the coach pane 212. Too small a sliding distance will increase a lot of unnecessary calculations and cannot improve the resolution of motion analysis. A too large sliding distance will reduce a lot of calculations but also reduce the resolution of motion analysis. In some implementation forces, the above-mentioned sliding distance is 0.15 seconds.
  • Fig. 3 is a flowchart illustrating a method for scoring sports courses according to an embodiment.
  • step 301 the course video is played, the coach movement data corresponding to the course video is obtained, and the coach pane is obtained from the coach movement data.
  • step 302 the student movement data is obtained through the inertial measurement unit, and the student pane is obtained from the student movement data, wherein the length of the student pane is greater than the length of the coach pane.
  • the student segment in the student pane that is most similar to the coach pane is searched for.
  • step 304 determine the type of exercise. If it is the first type of exercise, in step 305, calculate the exercise score according to the stability of the student segment and the coach pane.
  • step 306 deduct The score mechanism is calculated based on the difference between the student segment and the coach pane.
  • step 304 can be replaced by calculating the weight of the first type of exercise and the weight of the second type of exercise.
  • the weight of the exercise type is added to the weight of the second exercise type.
  • the weight of the first exercise type and the weight of the second exercise type can be determined by looking up the table according to the sports items (such as boxing aerobic, Latin dance, intermittent intensity exercise, yoga, etc.), or according to the coach pane
  • the length of 211 (or student pane 221) is determined.
  • each step in FIG. 3 has been described in detail as above, and will not be repeated here. It should be noted that each step in FIG.
  • FIG. 3 can be implemented as multiple program codes or circuits, and the present invention is not limited thereto.
  • the method in FIG. 3 can be used in conjunction with the above embodiments, or can be used alone. In other words, other steps can also be added between the steps in FIG. 3.
  • the aforementioned L2 norm can also be replaced with L1 norm, maximum norm, histogram distance, or other suitable distance calculation methods.
  • the above-mentioned OBE-DTW can also be replaced with other types of DTW, or replaced by similarity algorithms such as the longest common sequence (LCS).

Abstract

L'invention concerne un procédé et un système de notation de séance d'exercices. Le procédé de notation de séance d'exercices consiste à : lire la vidéo d'une séance, acquérir des données d'exercice d'entraîneur correspondant à la vidéo de la séance, puis acquérir une fenêtre d'entraîneur à partir des données d'exercice d'entraîneur (301) ; acquérir des données d'exercices d'étudiant au moyen d'une unité de mesure inertielle, puis acquérir une fenêtre d'étudiant à partir des données d'exercices d'étudiant, la longueur de la fenêtre d'étudiant étant supérieure à la longueur de la fenêtre d'entraîneur (302) ; rechercher, dans la fenêtre d'étudiant, un segment d'étudiant qui est le plus similaire à la fenêtre d'entraîneur (303) ; pour un premier type d'exercice, calculer un score d'exercice en fonction de la stabilité du segment d'étudiant et de la stabilité de la fenêtre d'entraîneur (305) ; et pour un second type d'exercice, utiliser un mécanisme de points d'inaptitude pour calculer un score d'exercice en fonction d'une différence entre le segment d'étudiant et la fenêtre d'entraîneur (306). De cette manière, différents procédés de notation sont utilisés pour différents types d'exercice, et le score d'exercice peut être calculé de manière plus appropriée.
PCT/CN2020/104094 2020-03-27 2020-07-24 Procédé et système de notation de séance d'exercices WO2021189736A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010228332.0A CN113449945B (zh) 2020-03-27 运动课程评分方法与系统
CN202010228332.0 2020-03-27

Publications (1)

Publication Number Publication Date
WO2021189736A1 true WO2021189736A1 (fr) 2021-09-30

Family

ID=77807903

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/104094 WO2021189736A1 (fr) 2020-03-27 2020-07-24 Procédé et système de notation de séance d'exercices

Country Status (1)

Country Link
WO (1) WO2021189736A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116611969A (zh) * 2023-07-17 2023-08-18 北京彩蛋科技有限公司 传统武术智能学习与评分系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103007514A (zh) * 2011-09-21 2013-04-03 财团法人工业技术研究院 运动学习系统与辅助使用者学习运动的方法
EP3090684A1 (fr) * 2015-05-08 2016-11-09 The Swatch Group Research and Development Ltd. Podomètre et procédé d'analyse de données de mouvement
CN107530030A (zh) * 2015-03-13 2018-01-02 赫尔比公司 确定人类运动活动类型的方法及其实施装置
CN109461016A (zh) * 2018-09-10 2019-03-12 平安科技（深圳）有限公司 数据评分方法、装置、计算机设备及存储介质
CN110148072A (zh) * 2018-02-12 2019-08-20 庄龙飞 运动课程评分方法与系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103007514A (zh) * 2011-09-21 2013-04-03 财团法人工业技术研究院 运动学习系统与辅助使用者学习运动的方法
CN107530030A (zh) * 2015-03-13 2018-01-02 赫尔比公司 确定人类运动活动类型的方法及其实施装置
EP3090684A1 (fr) * 2015-05-08 2016-11-09 The Swatch Group Research and Development Ltd. Podomètre et procédé d'analyse de données de mouvement
CN110148072A (zh) * 2018-02-12 2019-08-20 庄龙飞 运动课程评分方法与系统
CN109461016A (zh) * 2018-09-10 2019-03-12 平安科技（深圳）有限公司 数据评分方法、装置、计算机设备及存储介质

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116611969A (zh) * 2023-07-17 2023-08-18 北京彩蛋科技有限公司 传统武术智能学习与评分系统
CN116611969B (zh) * 2023-07-17 2023-10-03 北京彩蛋科技有限公司 传统武术智能学习与评分系统

Also Published As

Publication number Publication date
CN113449945A (zh) 2021-09-28

Similar Documents

Publication Publication Date Title
Ghasemzadeh et al. Sport training using body sensor networks: A statistical approach to measure wrist rotation for golf swing
US10803762B2 (en) Body-motion assessment device, dance assessment device, karaoke device, and game device
Nakai et al. Prediction of basketball free throw shooting by openpose
US20130018494A1 (en) System and method for motion analysis and feedback with ongoing dynamic training orientation determination
US20140018181A1 (en) Golf swing analysis method and apparatus
JP6943294B2 (ja) 技認識プログラム、技認識方法および技認識システム
JP2014521139A (ja) スポーツ動作のビデオを解析する方法
JPWO2019130527A1 (ja) 抽出プログラム、抽出方法および情報処理装置
TWI681798B (zh) 運動課程評分方法與系統、電腦程式產品
Chen et al. Using real-time acceleration data for exercise movement training with a decision tree approach
US10726738B1 (en) Automated fitness coaching based on motion data
US20230169710A1 (en) Techniques for establishing biomechanical model through motion capture
Chun et al. A sensor-aided self coaching model for uncocking improvement in golf swing
US11682157B2 (en) Motion-based online interactive platform
Malawski Depth versus inertial sensors in real-time sports analysis: A case study on fencing
Kim et al. Golf swing analysis system with a dual band and motion analysis algorithm
CN110751100A (zh) 一种体育场馆辅助训练方法与系统
Wei et al. Performance monitoring and evaluation in dance teaching with mobile sensing technology
CN110148072B (zh) 运动课程评分方法与系统
WO2021189736A1 (fr) Procédé et système de notation de séance d'exercices
Torigoe et al. Strike Activity Detection and Recognition Using Inertial Measurement Unit towards Kendo Skill Improvement Support System.
TWI766259B (zh) 運動課程評分方法與系統、電腦程式產品
CN113449945B (zh) 运动课程评分方法与系统
Zefinetti et al. Goalkeeper’s Performances Assessed with Action Cameras Based Mocap System
CN111803074A (zh) 肢体运动指导方法、装置、系统、电子设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20926872

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20926872

Country of ref document: EP

Kind code of ref document: A1