CN114558313A

CN114558313A - Motion recognition and scoring method for strength training instrument

Info

Publication number: CN114558313A
Application number: CN202210235052.1A
Authority: CN
Inventors: 陈利民
Original assignee: Beijing Taxing Tianji Science And Technology Development Co ltd
Current assignee: Beijing Taxing Tianji Science And Technology Development Co ltd
Priority date: 2022-03-11
Filing date: 2022-03-11
Publication date: 2022-05-31
Anticipated expiration: 2042-03-11
Also published as: CN114558313B

Abstract

The invention discloses a motion recognition and scoring method for a strength training instrument, which relates to the technical field of training auxiliary devices and comprises the following steps: testing and determining a limit weight RM and determining a training target reference weight M; generating a complete training set according to indexes of the limit weight RM and the reference weight M; generating a minimum starting dead zone L1 of the training instrument according to the attribute of the training instrument; by comparing the standard curve graph with the user training curve graph, scoring each action stage and each action completion quality of the mechanical training of the strength; carrying out accurate data quantization on the total amount of motion and storing; by repeating the steps, the training action quality of the user can be detected and scored by the strength instrument, and the user can find problems existing in the training in time and correct the problems in time. The data analysis is realized only by using the algorithm which can be realized by the conventional chip, the operation and the running are simple, the cost is reduced, and the method can be widely popularized and used.

Description

Action recognition and scoring method for strength training instrument

Technical Field

The invention relates to the technical field of training auxiliary devices, in particular to a motion recognition and scoring method for a strength training instrument.

Background

The existing strength training instrument can only simply collect exercise load and times, cannot detect whether the requirement of action quality is met or not in practical use, namely cannot analyze the problems in the action and cannot give reasonable scores.

The motion analysis method, the conventional camera vision scheme and the conventional wearable sensor scheme have the disadvantages of complex structure, high cost, fussy use and difficulty in large-scale popularization and use.

The invention patent with the existing publication number of CN114011045A discloses a wearable device-based training action counting method and a wearable device, wherein the wearable device is provided with an inertia measurement unit, and the method comprises the following steps: step 1, acquiring six-axis data through an inertia measurement unit and calculating three attitude angle data; step 2, selecting an attitude angle, and fitting a sine curve according to data of the selected attitude angle; step 3, after the fitting of a single action period is completed, judging whether the characteristic parameters of the fitted sinusoidal curve in the single action period meet a preset standard range, if so, adding 1 to the training action count; if not, the training action count is unchanged, and the characteristic parameters comprise one or more of the angle difference range of the single action period, the time length of the single action period, the symmetry degree of the curve of the single action period and the difference amplitude of two end points of the curve of the single action period.

The technical scheme provides a training action counting method based on wearable equipment and the wearable equipment, wherein the characteristic parameters of the sine curve comprise one or more of a single action period angle difference range, a single action period time length, a single action period curve symmetry degree and a single action period curve two-endpoint difference amplitude.

In this embodiment, the action counting module is further configured to: when the characteristic parameters of the fitted sinusoidal curve in n continuous action periods appear for the first time all accord with a preset standard range, and the duration of the n action periods all accord with a preset time length, recording the first action in the n actions as an initial action, counting from the initial action, and n is more than or equal to 2.

In n continuous action periods appearing for the first time, after fitting of each action period is completed, a preset standard range is adjusted according to characteristic parameters of a fitted sine curve in the current single action period, namely whether counting is judged by taking the initial preset standard range as a standard in the initial action period, whether counting is judged by taking the adjusted preset standard range in the previous action period as a standard in the second action period to the nth action period, and after the n continuous action periods appearing for the first time are ended, the finally adjusted preset standard range is taken as a judgment standard of subsequent training action counting. The system realizes the approximate monitoring of training personnel, but does not have the scoring and data analysis capabilities, and still has the defects of complex structure, high cost, fussy use and difficulty in large-scale popularization and use.

Disclosure of Invention

In view of the above technical problems, an object of the present invention is to provide a motion recognition and scoring method for a strength training apparatus, wherein a central processing unit and a mass detection mechanism are disposed on the training apparatus, so as to ensure that the strength training apparatus can detect and score the quality of the user's training motion, and facilitate the user to find and correct the problems existing in the training in time. The data analysis is realized only by using the algorithm which can be realized by the conventional chip, the operation and the running are simple, the cost is reduced, and the method can be widely popularized and used.

The technical purpose of the invention is realized by the following technical scheme:

a method for motion recognition and scoring for a strength training apparatus, the method comprising the steps of:

the method comprises the following steps: the central processing unit determines the limit weight RM through testing and determines the reference weight M of the training target (the value can be obtained through two modes of manual input or testing and inputting through the central processing unit);

step two: the central processing unit generates a complete training set according to indexes of the limit weight RM and the reference weight M, and calculates the load W of the action of the training set according to the reference weight;

step three: the central processing unit generates a minimum starting dead zone L1 and a stability coefficient K of the training instrument according to the self attribute of the training instrument (the values can be obtained by manual input or test input of the central processing unit);

step four: before training, introducing a complete curve chart of three stages of centripetal contraction, shrinkage maintaining and centrifugal contraction of standard action into a middle letting processor;

step five: when the training instrument is used for training, the central processing unit identifies three stages of centripetal contraction, contraction keeping and centrifugal contraction of real-time training actions of a user by detecting the training instrument and generates a complete curve graph;

step six: the central processing unit scores the completion quality of each action stage and each action of the mechanical training of the strength by comparing the standard curve graph with the user training curve graph;

step seven: the central processing unit carries out accurate data quantization and storage on the total amount of movement, and provides a data basis for big movement data.

By adopting the technical scheme, the training intensity suitable for the training personnel can be determined by collecting data before the training of the training personnel, real-time monitoring is carried out when the training personnel train, the training start time and the training end time are automatically detected, real-time feedback is carried out according to the action standard degree, the training data is subjected to data analysis and stored after the training of the training personnel, the training data is used as available data during the next training, and the motion amount is obtained by analyzing the integral of effective work. In addition, the system can judge whether the user is exhausted in real time, and reduce the training risk. The problem of current strength training apparatus can only be simple to gather motion heavy burden and number of times, can not detect whether reach the requirement of action quality in the in-service use, also can not analyze in the action promptly is solved, can not give reasonable grade technical problem. Compared with the conventional camera vision scheme and the conventional wearable sensor scheme, the structure is simpler and more concentrated, the use is more convenient and faster, and the wide popularization and use are easy.

The invention is further configured to: the number of times N of the training set is default to 12, but the number of times N of the training set is dynamically adjusted according to the user limit weight RM, the body basic state and different training actions.

By adopting the technical scheme, after the weight W is calculated, the training instrument automatically adjusts the weight of the equipment to the weight W, or informs a user to automatically adjust the weight to the weight W. And performing action recognition and scoring within the weight W range in subsequent strength training of the user.

The invention is further configured to: the stability factor K can be set to 0.85 in general

By adopting the technical scheme, the stability of training and data acquisition can be optimized to a great extent by setting the stability coefficient K (K is the optimal undetermined coefficient obtained after big data analysis).

The invention is further configured to: the complete graph of the three phases of centripetal contraction, holding contraction and centrifugal contraction can be automatically generated by the central processor according to the property of the instrument (the value can be obtained by manual input or test input of the central processor).

By adopting the technical scheme, the training effect during training can be judged by taking the icon obtained by the existing data statistics as a standard.

The invention is further configured to: the training apparatus provided with the central processing unit is provided with a counterweight measuring device and a counterweight displacement detecting device.

By adopting the technical scheme, the movement condition of the training instrument can be judged by detecting the counterweight measuring device and the counterweight displacement detecting device on the training instrument, and the data acquisition of the training instrument is optimized to a great extent.

The invention is further configured to: and a display screen is arranged on the instrument provided with the central processing unit and can feed back the action scores in real time.

By adopting the technical scheme, the action standard can be fed back in real time when the training personnel train.

The invention is further configured to: the instrument provided with the central processing unit is provided with a protection device, and the real-time protection device is controlled by the central processing unit.

By adopting the technical scheme, whether the user is exhausted can be judged in real time, the protection device is rapidly started, and the training risk is reduced.

In conclusion, the beneficial technical effects of the invention are as follows:

(1) the training intensity suitable for the training personnel can be determined by collecting data before the training of the training personnel, real-time monitoring is carried out when the training personnel train, the training start time and the training end time are automatically detected, real-time feedback is carried out according to the action standard degree, data analysis is carried out on the training data after the training of the training personnel, the training data are stored and used as available data during next training, and the motion amount is obtained through integral calculation of effective work of analysis.

(2) The training system has the function of judging whether the user is exhausted in real time, and can reduce the training risk.

(3) The real-time feedback can be carried out on the standard of action achieved when the training personnel train.

Drawings

FIG. 1 is a schematic overall flow diagram of one embodiment of the present invention;

FIG. 2 is a schematic flow chart of a detection and protection apparatus according to an embodiment of the present invention;

FIG. 3 is a graph illustrating a standard motion force curve according to an embodiment of the present invention;

FIG. 4 is a graph illustrating an actual force curve according to an embodiment of the present invention;

FIG. 5 is a force curve fit diagram in accordance with an embodiment of the present invention;

FIG. 6 is a schematic diagram of the slope product of the action score curve according to an embodiment of the present invention;

FIG. 7 is a graphical representation of a force-out curve in one embodiment of the invention.

Detailed Description

The present invention will now be described more fully hereinafter with reference to the accompanying examples.

Referring to fig. 1 and 2, a motion recognition and scoring method for a strength training apparatus comprises the following steps:

step four: before training, introducing a complete curve chart of three stages of centripetal contraction, shrinkage maintaining and centrifugal contraction of standard actions into a central processing unit;

This device is when using, as the transfer through central processing unit, the connection data, the icon, the apparatus, show and protection device, can confirm the training intensity who is fit for training personnel at the data acquisition before training personnel trains, carry out real time monitoring when training personnel trains, automated inspection training begins and finish time, and make real-time feedback according to action standard degree, and carry out data analysis and save with training data after training personnel trains, as the available data when training next time, the integral calculation through the effective work of analysis obtains the amount of exercise. In addition, the system can judge whether the user is exhausted in real time, and reduce the training risk. The problem of current strength training apparatus can only be simple to gather motion heavy burden and number of times, can not detect whether reach the requirement of action quality in the in-service use, also can not analyze in the action promptly is solved, can not give reasonable grade technical problem. Compared with the conventional camera vision scheme and the conventional wearable sensor scheme, the structure is simpler and more concentrated, the use is more convenient and faster, and the wide popularization and use are easy.

Referring to fig. 1, the number of exercise motions N of the training set is default to 12, but is dynamically adjusted according to the user's ultimate weight RM, the basic body state and the different exercise motions.

The number of times N of the training set is default to 12, but the dynamic adjustment is performed according to the user limit weight RM, the body basic state and the training action.

The weight W of the training set action is derived by the following algorithm:

W＝M·(1-0.0306·N+0.00045·N2)

after the weight W is calculated, the training apparatus automatically adjusts the weight of the device to W, or informs the user to automatically adjust the weight to W. And the subsequent strength training of the user is to perform action recognition and scoring within the weight W range.

With reference to fig. 1, the stability factor K can be set to 0.85 in general.

The parameters are usually preset before the training apparatus leaves the factory. The motion recognition and scoring method proposes to set the instrument parameters as follows:

the sampling rate defaults to 100 times per second,

in order to avoid the accidental touch of the instrument, an instrument minimum starting dead zone L1 is set

To ensure that the training motion samples are stable, a stability factor K should be set, which may typically be set to 0.85.

With reference to fig. 3, the complete graph of the three phases of centripetal contraction, holding contraction and centrifugal contraction can be automatically generated by the cpu according to the properties of the instrument itself (this value can be obtained by manual input or by test input via the cpu). The training apparatus provided with the central processing unit is provided with a counterweight measuring device and a counterweight displacement detecting device.

Introducing an action reference curve before the training of the device, wherein different action reference curves are used for different training actions; the action curve can be preset in the instrument or can be acquired by other data acquisition modes.

Comprises a centripetal contraction phase SS2, a holding contraction phase SS3 and a centrifugal contraction phase SS 4.

The reference curve contains the standard trip LL, the action timeout TT,

and calculating out the time used TTS3 ═ TT 3-TT 2 in the standard retention shrinkage period SS 3;

the slope mSS2 of the standard centripetal contraction phase SS2 is LL/(TT 2-TT 1);

the slope mSS4 of the standard centrifugal contraction phase SS4 is LL/(TT 4-TT 3).

With reference to fig. 4, a single data curve can be obtained for each action record during training with the apparatus.

In the single data plot, the minimum startup dead zone L1 is the following segment S1, which is negligible,

the single motion is initiated when the instrument stroke exceeds the minimum actuation deadband L1, the start time T1 is recorded,

the single action ends when the instrument travel falls below the minimum actuation dead band L1 again, recording an end time T4,

the maximum stroke of a single action is L3, a stable sampling stroke L2 is obtained by calculation, namely the maximum stroke L3 multiplied by a stable coefficient K,

the point in time when the stroke exceeds the stable sampling stroke L2 is a hold segment start time T2,

the point of time when the stroke is lower than the stable sampling stroke L2 is the hold section end time T3,

the time slicing can obtain a single centripetal contraction stage S2, a single retention contraction stage S3 and a single centrifugal contraction stage S4.

And further calculating the duration TS3 of the single keeping segment to be T3-T2;

single centripetal contraction phase S2 slope mS2 ═ L2-L1)/(T2-T1;

the slope mS4 of the single centrifugation contraction phase S4 is (L2-L1)/(T4-T3).

With reference to fig. 5, a display screen is arranged on the instrument provided with the central processing unit, and the display screen can feed back the action scores in real time.

When the device works, whether the training action is effective or not can be determined by comparing and analyzing the single action curve and the reference curve.

When the ending time T4 is greater than the action timeout TT, the action is overtime, the score can be reduced or the action is judged to be invalid.

When TS3 is greater than the standard TTS3 for the duration of the systolic phase, then the action is held valid, otherwise the action remains unsuccessful and the score can be decreased or the action determined to be invalid.

The action completion degree can be used as a scoring coefficient of the current action, and if the action completion degree is low, the action can be judged to be invalid.

And (3) calculating the completion degree P2 of the centripetal contraction phase action according to an algorithm by combining the figure 6:

similarly, the completion degree of the centrifugal contraction stage action P4:

keeping the motion finish degree P3 of the contraction stage as L3/standard stroke LL of the maximum stroke of single motion.

The total completion (score) P can be obtained by summing or integrating the three stage completions P2, P3 and P4.

The total completion is the score, which is ultimately a percentage value, with the closer the score to 100%, the more standard the training action, and the higher the score. From the graph, the higher the degree of coincidence of ms2 with mss2, the higher the score.

With reference to fig. 4, data statistics and analysis can be performed by calculating the amount of exercise.

The motion amount PL is calculated from the load W, the stroke L, the start time T1 to the end time T4 for each operation record.

PL can scientifically and efficiently express the amount of exercise at a time for statistics and analysis of data.

With reference to fig. 7, a protection device is arranged on the apparatus on which the central processing unit is installed, and the real-time protection device is controlled by the central processing unit.

In order to prevent the user from taking off the force and prevent danger, when the condition that the user is exhausted is obtained through the analysis of the training action curve, the user is immediately informed or the protection function of the instrument is started. In a typical training scenario, each action takes 3-6 seconds to complete. And in one second, if the change of the positive direction and the negative direction of the stroke is detected for three times, the user is judged to be exhausted, and the protection device is controlled and started through the central processing unit.

The working principle of the embodiment is as follows: when the device is used, the device is used as a transfer through a central processing unit, data are connected, an icon, an instrument, a display and protection device are connected, real-time monitoring is carried out according to seven steps controlled by the central processing unit, data are collected and output in real time, the training intensity suitable for training personnel can be determined by collecting the data before the training of the training personnel, real-time monitoring is carried out when the training personnel train, the training start time and the training end time are automatically detected, real-time feedback is carried out according to action standard degree, data analysis and storage are carried out on the training data after the training of the training personnel, the training data are used as available data during next training, and the amount of exercise is obtained through analyzing integral calculation of effective work. In addition, the system can judge whether the user is exhausted in real time, and reduce the training risk. The problem of current strength training apparatus can only be the simple collection motion heavy burden and number of times, can not detect in the in-service use whether to reach the requirement of action quality, also can not analyze in the action promptly, can not give reasonable grade technical problem is solved. Compared with the conventional camera vision scheme and the conventional wearable sensor scheme, the structure is simpler and more concentrated, the use is more convenient and faster, and the wide popularization and use are easy.

The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited thereby, so: equivalent changes made according to the structure, shape and principle of the invention shall be covered by the protection scope of the invention.

Claims

1. A method for motion recognition and scoring for a strength training apparatus, the method comprising the steps of:

step three: the central processing unit generates a minimum starting dead zone L1 and a stability coefficient K of the training instrument according to the self attribute of the training instrument (the values can be obtained by two modes of manual input or test input through the central processing unit);

step seven: the central processing unit carries out accurate data quantization and storage on the total movement amount, and provides a data base for big movement data.

2. The motion recognition and scoring method for a strength training instrument as recited in claim 1, wherein: the number of training set actions N is default to 12, but can be dynamically adjusted according to the user limit weight RM, the body basic state and different training actions.

3. The motion recognition and scoring method for a strength training instrument according to claim 1, wherein: the stability factor K may be set to 0.85 in general.

4. The motion recognition and scoring method for a strength training instrument as recited in claim 1, wherein: the complete graph of the three phases of centripetal contraction, holding contraction and centrifugal contraction can be automatically generated by the central processor according to the property of the instrument (the value can be obtained by manual input or test input of the central processor).

5. The motion recognition and scoring method for a strength training instrument according to claim 1, wherein: the training apparatus provided with the central processing unit is provided with a counterweight measuring device and a counterweight displacement detecting device.

6. The motion recognition and scoring method for a strength training instrument according to claim 1, wherein: and a display screen is arranged on the instrument provided with the central processing unit and can feed back the action scores in real time.

7. The motion recognition and scoring method for a strength training instrument according to claim 1, wherein: the apparatus provided with the central processing unit is provided with a protection device, and the real-time protection device is controlled by the central processing unit.