An intelligent training method and device based on a strength mirror

Abstract

The application discloses a kind of intelligent training method and device based on strength body-building mirror, comprising the following steps: 1) the motion data of user in current group training is collected;2) the motion data collected is analyzed and calculated, and the reference data of multiple dimensions of user in current group training is obtained;3) the subsequent motion data of current group training is monitored, and compared and analyzed with the reference data of multiple dimensions, whether the subsequent motion of current group training is correct is judged, if there is error, the user is fed back, and it is guided to find the correct force mode.The application can fully combine own condition, dynamically identify the accuracy of action, and timely guide adjustment action when confirming action is wrong.

Description

Technical Field

[0001] This invention relates to the field of fitness, and in particular to an intelligent training method and device based on a strength training mirror. Background Technology

[0002] With the increasing awareness of fitness and the development of technology, smart fitness products have begun to integrate into people's lives, and strength training mirrors are one of them. Existing strength training mirrors may include a main frame, a strength training mechanism, and a display component covering the main frame. The strength training mechanism is used to perform strength training movements, and the display component may include a glass mirror and a screen. The glass mirror provides feedback on the user's posture, allowing for easy feedback, while the screen can display video content, meaning the user's posture image can be overlaid with video content to guide the user in adopting correct posture during exercise.

[0003] Currently, these strength training mirrors have the following problems: When users use such products for independent training, the lack of coaching guidance and real-time interaction makes it difficult to ensure consistent accuracy of movements. Especially during strength training, as physical fitness and strength gradually deplete, movements become increasingly prone to distortion, reducing training effectiveness and potentially causing sports injuries. Furthermore, due to individual differences in physical capabilities, the same movement performed by different people, or even by the same person at different stages, can vary. Therefore, even for the same movement, it's difficult to use a single, fixed standard to judge whether it has been performed correctly for different people or the same person in different states. Therefore, the problem that these strength training mirrors need to solve is—how to dynamically identify the accuracy of movements based on the user's individual situation during strength training, and promptly guide and adjust movements when errors are detected. Summary of the Invention

[0004] The main objective of this invention is to overcome the shortcomings of existing strength training mirrors in terms of accuracy in recognizing training movements. It proposes an intelligent training method and device based on a strength training mirror, which can dynamically recognize the accuracy of training movements and guide the user to adjust the movements in a timely manner when errors are confirmed.

[0005] The present invention adopts the following technical solution:

[0006] A smart training method based on a strength training mirror, characterized by the following steps:

[0007] 1) Collect the user's motion data during the current training group;

[0008] 2) Analyze and calculate the collected motion data to obtain the user's baseline data in multiple dimensions of training in the current group;

[0009] 3) monitoring the subsequent movement data of the current group training, and comparing and analyzing the benchmark data in multiple dimensions to determine whether the subsequent movement of the current group training is correct, and if not, giving feedback to the user to guide them to find the correct force mode.

[0010] Preferably, the movement data in the current group training includes at least the real-time rope amplitude, real-time return rope amplitude and real-time speed of each action.

[0011] Preferably, the collected movement data is analyzed and calculated to obtain the benchmark data of the user in the current group training in multiple dimensions, specifically including:

[0012] 2.1) In the current group training, locate several actions of the reference object that can be used as benchmark data, each action including a centripetal action and a centrifugal action;

[0013] 2.2) Retrieve the original data of the reference object and perform benchmark data analysis and calculation in multiple dimensions.

[0014] Preferably, in step 2.1), the movement data of the first N actions of the user performing the current group training is used as the reference object; and the movement data of the first action of the current group training is filtered.

[0015] Preferably, in step 2.1), when the user adjusts the resistance, the movement data of the action before adjustment is filtered, and the movement data of the N actions after adjustment is locked as the reference object.

[0016] Preferably, the benchmark data includes a centripetal amplitude benchmark, and the centripetal amplitude benchmark is the absolute value of the difference between the terminal amplitude of the current centripetal action and the terminal amplitude of the previous centrifugal action.

[0017] Preferably, the benchmark data includes a centrifugal amplitude benchmark, and the centrifugal amplitude benchmark is the absolute value of the difference between the terminal amplitude of the current centrifugal action and the terminal amplitude of the current centripetal action.

[0018] Preferably, the benchmark data includes a centrifugal speed benchmark, and the centrifugal speed benchmark is (the absolute value of the difference between the terminal amplitude of the current centrifugal action and the terminal amplitude of the current centripetal action) / (the time length between the terminal amplitude of the current centrifugal action and the terminal amplitude of the current centripetal action).

[0019] Preferably, the benchmark data includes an average benchmark value, and when the current group training performs multiple actions, multiple benchmark values are obtained in each dimension, then the multiple benchmark values in each dimension are averaged to obtain the average benchmark value of each dimension of the current group training.

[0020] Preferably, the reference data comprises a reference value range, and a fluctuation amplitude is given to the average reference value of each dimension of the group training, so as to obtain the reference value range of each dimension of the group training.

[0021] Preferably, in step 3), the motion data of each subsequent action of the current group training is obtained and compared with the average reference value or the reference value range of the plurality of dimensions of the current group training, if the motion data is greater than or less than the average reference value or the reference value range, the action is wrong, and the user is given feedback through the fitness mirror to guide him to find the correct force mode; the feedback includes playing video, voice or graphics, and displaying real-time motion data and the average reference value or the reference value range of the plurality of dimensions.

[0022] An intelligent training device based on a strength fitness mirror, characterized in that it comprises

[0023] A collection device for collecting motion data of a user in a current group training;

[0024] An analysis and calculation device for analyzing and calculating the collected motion data to obtain reference data of a plurality of dimensions of the user in the current group training;

[0025] A comparison and judgment device for monitoring motion data of a subsequent action of the current group training and comparing and analyzing the motion data with the reference data of the plurality of dimensions to judge whether the subsequent action of the current group training is correct;

[0026] A feedback device for giving feedback to the user to guide him to find the correct force mode when the subsequent action of the current group training is wrong.

[0027] From the above description of the present application, compared with the prior art, the present application has the following beneficial effects:

[0028] 1. In the present application, the collected motion data is analyzed and calculated to obtain reference data of a plurality of dimensions of the user in the current group training; the subsequent motion data of the current group training is monitored and compared with the reference data of the plurality of dimensions to judge whether the subsequent motion of the current group training is correct, if it is wrong, the user is given feedback, the accuracy of the action is dynamically identified in combination with his own situation, and the action is adjusted in time when it is confirmed to be wrong.

[0029] 2. In the present application, the several times of actions of positioning the reference object of the reference data are taken as the reference object of the user's personal data at present, which avoids the natural individual differences caused by factors such as gender, height and sports ability, and also avoids the case that the same user has large difference in sports performance in different states (for example, healthy and sick, active and tired, poor sports ability and strengthened sports ability).

[0030] 3. In this invention, the motion data of the first movement in the current training group is filtered, and the motion data of the movement before the adjustment is filtered when the user adjusts the resistance. The motion data of the N movements after the adjustment is locked as the reference object. This filters out important interference factors, including invalid data generated during the movement preparation stage, invalid data caused by shaking or other abnormalities during the movement, and differences in movement data caused by adjusting the resistance, thereby improving the accuracy of the benchmark value.

[0031] 4. In this invention, the reference data includes centripetal amplitude reference, centrifugal amplitude reference, and centrifugal velocity reference. The calculation method is simple, and the average reference value and reference value range can also be set. That is, after obtaining the reference data, the reference range is further defined to provide tolerance space.

[0032] 5. In this invention, after acquiring the reference data, it is directly displayed on the interface, and the user's current action execution status is fed back through real-time animation, thereby guiding the user to apply force correctly and reducing the error rate. Simultaneously, the real-time feedback animation system monitors the left and right lever arms, allowing the user to monitor the execution status of both sides at the same time.

[0033] 6. In this invention, more vivid and timely virtual coach voice and text prompts are used to remind users of errors through hearing and vision as soon as an error is detected, helping users to perceive the error and the correction method in a timely manner. Attached Figure Description

[0034] Figure 1 This is a flowchart of the method of the present invention;

[0035] Figure 2 A diagram illustrating the preparation phase for a kneeling lat pulldown;

[0036] Figure 3 This is a schematic diagram illustrating the actual execution phase of the action.

[0037] Figure 4 This is a schematic diagram illustrating the change in the amplitude of the rope pull;

[0038] Figure 5 This is a diagram illustrating error correction.

[0039] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Detailed Implementation

[0040] The present invention will be further described below through specific embodiments.

[0041] The power fitness mirror in the application can include a main body support, a power training mechanism and a display assembly covering the main body support, the power training mechanism can be used to perform power training actions, and the display assembly can include a glass mirror and a display screen, etc., the glass mirror is used to feedback the motion posture image of the user, so as to facilitate the user to obtain motion feedback, and the display screen can be used to display video content, etc., that is, the motion posture image of the user can be superimposed and displayed with the video content, so as to guide the user to adopt a correct posture for fitness.

[0042] The power training mechanism can include a universal guide assembly and a driving assembly, the universal guide assembly is provided with a pulling element, i.e., a pulling rope, and the driving assembly is provided with a resistance element, which can be a motor, a coil spring or a tension spring, etc. The user can perform power training by pulling the pulling element and resisting the resistance driven by the resistance element.

[0043] Referring to Figure 1 The application provides an intelligent training method based on a power fitness mirror, which includes the following steps:

[0044] 1) Collecting motion data of the user in the current group training.

[0045] In the application, the driving assembly of the power fitness mirror, such as a motor, can collect the motion data of the user in real time during the training, including but not limited to left / right motor, rope amplitude, real-time rope amplitude, real-time speed, resistance value, motor temperature, etc.

[0046] At present, in power training, the main judgment standard for judging whether an action is correct is:

[0047] (1) Whether the action amplitude is in place when pulling out and restoring, and whether the stretching amplitude of the muscle can be ensured;

[0048] (2) Whether the action speed is reasonable in different execution stages, and whether the effect of fully stimulating the muscle is achieved.

[0049] Therefore, the main data target of the application is the rope amplitude and real-time speed at each time node for each execution of each action.

[0050] When the above target data is collected, it is read at a frequency of about 20 times per second. For example, when the user performs each action A, the N real-time motion data read is as follows:

[0051]

[0052] 2) Analyzing and calculating the collected motion data to obtain the reference data of the user in the current group training in multiple dimensions.

[0053] In the present application, the reference data is the main basis for judging whether the user has accurately completed the action. For example, when the relevant data obtained after the action is performed has a large difference with the reference data, it can be judged that the accuracy of the action is wrong and needs to be corrected.

[0054] In the present application, the target reference data type can include: (1) centripetal amplitude reference: the amplitude of the pull rope when exerting force; (2) centrifugal amplitude reference: the amplitude of the return rope when restoring; (3) centrifugal speed reference: the time required to complete the restoration stroke, the shorter the time, the faster the speed, and vice versa.

[0055] In this step, the way to obtain the reference data can be divided into two steps:

[0056] 2.1) In the current group training, locate several actions that can be used as reference objects for reference data, each action including centripetal action and centrifugal action.

[0057] The reference data is the core reference value for judging whether the action is standard, therefore, the object that can be used as reference data itself needs to be as standard as possible. Then, how to ensure the reliability of the reference object is the first problem to be solved.

[0058] Since the physical functions of each individual are not the same, the reference object is first locked to the user himself, that is, only his own condition is used as a reference.

[0059] Secondly, in the fitness field, a single training contains multiple actions, each action needs to be trained for multiple groups, and each group needs to be executed for multiple times. The user's physical strength and power are obviously different in different stages of single training and single group training, for example, in the first N times of each group, since the user is in a stage where physical strength and power are relatively full, the action accuracy is the highest. However, in the last few actions of each group, due to the gradual consumption of physical strength and power, most of the action deformation and error scenarios are concentrated in the second half of each group.

[0060] Therefore, the reference object of the reference value has the following characteristics: (1) locking the user himself as the reference object (2) obtaining reference data in groups: that is, each group has its own reference data (3) using the first N movement data of each group training as the reference object of the reference data. That is, the movement data of the first N actions of the user himself in the current group training is used as the reference object, where N is an integer, and its specific value can be configured.

[0061] At the same time, in order to further ensure the reliability of the acquisition object, the application also adds the following mechanism: when the user adjusts the resistance size, filter the motion data of the action before adjustment, and lock the motion data of the N times of action after adjustment as the reference object. The specific description is as follows: (1) filter the first action of each group of training: the purpose is to avoid the preparation distance and preparation time generated in the action preparation stage from affecting the calculation of the reference value; (2) when the user adjusts the resistance size, the motion data of the action before adjustment needs to be filtered, and the motion data of the N times of action after adjustment is locked as the reference object. The function of this mechanism is explained as follows.

[0062] 1. What is the action preparation stage?

[0063] Reference Figure 2 It is the preparation stage of kneeling high pull-down, when the user puts on the preparation posture, it can be seen that the rope has been pulled out a distance at this time, when the user completes the pull-out stage and restores, it will also restore to this position. This distance is the preparation stage, and this distance actually does not participate in the actual action execution stage, reference Figure 3 The actual action execution stage. Since the first action of each group of actions needs to be adjusted by the user, a preparation distance will be generated at this time, therefore, the first action of each group is filtered.

[0064] 2. Why filter the action before resistance adjustment and take the action after resistance adjustment?

[0065] As mentioned earlier, the reference data finally acquired by the application mainly includes action amplitude and speed. The resistance size is the key factor affecting amplitude and speed. For example: the same user uses the same action to lift the dumbbell, when the weight of the dumbbell is 5KG and 20KG, the action amplitude with 20KG is relatively small and the speed is relatively slow, at this time it will have a significant impact on the reference data determination, which is also the same in the scene of strength fitness mirror.

[0066] When the user adjusts the resistance size, it means that the resistance before adjustment does not reach the ideal state, only when the user determines the resistance, the training after that is the ideal state, and the data at this time has reference value. Therefore, when the user adjusts the resistance, the motion data of the N times of action after adjustment is locked as the reference object.

[0067] In summary, the application can be specifically illustrated by the following examples how the reference object of the reference data is positioned.

[0068] Example 1: Assuming that the training data of a user training a certain action is as follows:

[0069] Resistance (KG) 1st 10 2nd 10 3rd 10

[0070] First, filter the first action,

[0071] Secondly, since the user did not adjust the resistance in the first 3 times, it can be determined that the resistance is suitable for the user, and therefore the reference object of the baseline data of this set of training is positioned as the movement data of the second and third times.

[0072] Example 2: Assuming that a user trains a movement with the following training data:

[0073] Resistance (KG) 1st 10 2nd 11 3rd 13 4th 13

[0074] First, filter the first time,

[0075] Secondly, since the user adjusted the resistance in the second time, and then adjusted the resistance again in the third time, the fourth and third times have the same resistance, it can be determined that the third and fourth times are closer to the ideal state of the user, and therefore the reference object of the baseline data of this set of training is positioned as the movement data of the third and fourth times.

[0076] 2.2) Retrieve the original data of the reference object and perform baseline data analysis and calculation in multiple dimensions.

[0077] In this step, the baseline data analyzed and calculated includes the centripetal amplitude baseline, the centrifugal amplitude baseline, and the centrifugal speed baseline. The specific calculation is as follows:

[0078] Taking each set of training, including several movements, and each movement including a centripetal rope and a centrifugal rope as an example, referring to the rope amplitude change schematic diagram of Figure 4 , the points are explained as follows:

[0079]

[0080] According to the above positioning points, the following baseline data is further calculated:

[0081]

[0082] Specifically, the centripetal amplitude baseline is the absolute value of the difference between the end amplitude G of the current centripetal movement and the end amplitude E of the last centrifugal movement, i.e. |G-E|.

[0083] The centrifugal amplitude baseline is the absolute value of the difference between the end amplitude E of the current centrifugal movement and the end amplitude C of the current centripetal movement, i.e. |E-C|.

[0084] The centrifugal speed baseline is: (the absolute value of the difference between the end amplitude E of the current centrifugal movement and the end amplitude C of the current centripetal movement) / (the time length between the end amplitude E of the current centrifugal movement and the end amplitude C of the current centripetal movement), i.e. |E-C| / (the time length of E-C).

[0085] Further, the reference data includes average reference values. When the current group of training performs multiple actions, multiple reference values are obtained for each dimension, and then the average of each dimension is calculated to obtain the average reference value of each dimension of the current group of training. For example, it is known that the second and third times of user A training a certain group of actions are locked as the reference value of the current group, and then the reference value of the group is calculated as follows:

[0086]

[0087] The average of 2 times is taken, and the centripetal amplitude reference of the current group of training is (10+12) / 2=11 cm.

[0088] The average of 2 times is taken, and the centripetal amplitude reference of the current group of training is (10+12) / 2=11 cm.

[0089] The average of 2 times is taken, and the centripetal amplitude reference of the current group of training is (10+12) / 2=11 cm.

[0090] The average of 2 times is taken, and the centripetal amplitude reference of the current group of training is (10+12) / 2=11 cm.

[0091] For example, assuming that the fluctuation of the centripetal amplitude is 90%-110%, then the final centripetal amplitude reference range is 11*90%-11*110%, i.e. 9.9 cm-12.1 cm. The reference range of other dimensions is the same.

[0092] 3) Monitor the subsequent motion data of the current group of training, and compare and analyze the reference data of multiple dimensions to determine whether each action of the subsequent motion of the current group of training is correct. If not, feedback is given to the user to guide him to find the correct force mode.

[0093] After successfully obtaining the reference range of each dimension of a certain group of actions, the mirror interface will combine graphics and action videos to guide the user to follow the video to correctly exert force, and on the other hand, the reference data is marked in the dynamic graphics, and the graphics will react to the user's current action amplitude and speed in real time, so as to guide the user to correctly complete the action.

[0094] For example, see Figure 5 , the upper part of the figure is the action monitoring panel. The real-time amplitude and speed of the left and right force arms can be monitored. The user can refer to the reference value marked on the monitoring panel to control the force of the action in real time. The lower left is the action video demonstrated by the coach; the lower right is the muscle hot zone map; the subtitle in the middle is the virtual coach error correction subtitle, which is triggered when the action is determined to be substandard.

[0095] Specifically, the motion data of each subsequent action of the current group training is acquired and compared with the average reference value or reference value range of multiple dimensions of the current group training, if the motion data is greater than or less than the average reference value or reference value range, the action is wrong, that is, it does not meet the standard, and the user is fed back through the fitness mirror to guide him to find the correct power mode; the feedback can include playing video, voice or graphics, and displaying real-time motion data and average reference value or reference value range of multiple dimensions, etc.

[0096] See the error correction scenarios in the following table:

[0097] Error correction scenario Criteria Excessive centripetal amplitude Centripetal amplitude > centripetal reference range Insufficient centripetal amplitude Centripetal amplitude < centripetal reference range Excessive centrifugal amplitude Centrifugal amplitude > centrifugal reference range Insufficient centrifugal amplitude Centrifugal amplitude < centrifugal reference range Excessive centrifugal speed Centrifugal speed < centrifugal reference range

[0098] At the same time, in order to further achieve the purpose of error correction, various voice and script libraries under various error correction scenarios can be configured in advance for each action, when it is determined that the action does not meet the standard, the system will play the corresponding voice and subtitle prompt to guide the user to adjust the state in time. On the other hand, when the user correctly completes the action, the virtual trainer will encourage the user to maintain the status quo through voice and barrage, continue to cheer, until the user completes the training.

[0099] The application also provides an intelligent training device based on the strength fitness mirror, comprising

[0100] The acquisition device acquires the motion data of the user in the current group training.

[0101] The analysis and calculation device analyzes and calculates the acquired motion data to obtain the reference data of the user in the current group training in multiple dimensions.

[0102] The comparison and judgment device monitors the motion data of the subsequent action of the current group training and compares and analyzes the reference data in multiple dimensions to determine whether the subsequent action of the current group training is correct.

[0103] The feedback device gives feedback to the user when the subsequent action of the current group training is wrong, guiding him to find the correct power mode.

[0104] The application provides an intelligent training device based on the strength fitness mirror, which is used to execute the intelligent training method based on the strength fitness mirror.

[0105] The above is only a specific embodiment of the application, but the design concept of the application is not limited thereto, and any non-substantial modification of the application using this concept shall be deemed to infringe the protection scope of the application.

Claims

1. A method for intelligent training based on a strength mirror, characterized in that, It comprises the following steps: 1) Collecting the motion data of the user in the current group training, including the rope amplitude, the return rope amplitude and the real-time speed; 2) Analyzing and calculating the collected motion data to obtain the reference data of the user in the current group training in multiple dimensions; Specifically, it comprises: 2.1) In the current group training, positioning several actions of the reference object which can be used as reference data, each action including centripetal action and centrifugal action; the motion data of the first N actions of the user performing the current group training is used as the reference object; and filtering the motion data of the first action of the current group training; in step 2.1), when the user adjusts the resistance size, filter the motion data of the action before adjustment, and lock the motion data of the N actions after adjustment as the reference object 2.2) Retrieving the original data of the reference object and performing reference data analysis and calculation in multiple dimensions, the reference data including centripetal amplitude reference, centrifugal amplitude reference and centrifugal speed reference; the centripetal amplitude reference is the absolute value of the difference between the terminal amplitude of the current centripetal action and the terminal amplitude of the last centrifugal action; the centrifugal amplitude reference is the absolute value of the difference between the terminal amplitude of the current centrifugal action and the terminal amplitude of the current centripetal action; the centrifugal speed reference is (the absolute value of the difference between the terminal amplitude of the current centrifugal action and the terminal amplitude of the current centripetal action) / (the time length between the terminal amplitude of the current centrifugal action and the terminal amplitude of the current centripetal action); 3) Monitoring the subsequent motion data of the current group training and comparing and analyzing it with the reference data in multiple dimensions to determine whether the subsequent motion of the current group training is correct, and if not, giving feedback to the user to guide him to find the correct force way.

2. The intelligent training method based on the power fitness mirror according to claim 1, characterized in that: The motion data in the current group training includes at least the real-time rope amplitude, the real-time return rope amplitude and the real-time speed of each action. 3.The intelligent training method based on the power fitness mirror according to claim 1, characterized in that: The reference data includes average reference value, when the current group training performs multiple actions, multiple reference values are obtained in each dimension, then the average value of multiple reference values in each dimension is calculated to obtain the average reference value of each dimension of the current group training.

4. The intelligent training method based on the power fitness mirror according to claim 3, characterized in that: The reference data includes reference value range, an amplitude of fluctuation is given to the average reference value of each dimension of the current group training, thereby obtaining the reference value range of each dimension of the current group training.

5. The intelligent training method based on the power fitness mirror according to claim 4, characterized in that: In step 3), the motion data of each subsequent action of the current group training is obtained and compared and analyzed with the average reference value or reference value range of the multiple dimensions of the current group training obtained, if the motion data is greater than or less than the average reference value or reference value range, the action is wrong, then feedback is given to the user through the fitness mirror to guide him to find the correct force way; the feedback includes playing video, voice or graphics, and displaying real-time motion data and the average reference value or reference value range of the multiple dimensions.

6. A smart training device based on a strength mirror, characterized by: The intelligent training method based on the strength fitness mirror of any one of claims 1 to 5 comprises a collecting device for collecting the motion data of the user in the current group training; The analysis and calculation device analyzes and calculates the collected movement data to obtain reference data of the user in multiple dimensions of the current set of training; The comparison and judgment device monitors movement data of the subsequent action of the current set of training, and compares and analyzes the reference data in the multiple dimensions to determine whether the subsequent action of the current set of training is correct; The feedback device provides feedback to the user to guide him / her to find the correct force mode when the subsequent action of the current set of training is incorrect.

Images