An intelligent method for obtaining the pressure between lower limb joints during sprinting by means of resistance torsion

Abstract

The present invention relates to the technical field of human motion analysis. Specifically, it relates to an intelligent method for obtaining the pressure between lower limb joints during sprinting through resistance torsion, including the following steps: Step 1, obtain a number of six-degree-of-freedom values, resistance torsion values, and video acquisition results; Step 2, perform preprocessing to obtain effective six-degree-of-freedom values; Step 3, preprocess to obtain three-dimensional coordinate values; Step 4, obtain kinematic and dynamic values representing the pressure between lower limb joints; Step 5, through a motion capture model, obtain the kinematic and dynamic values of the pressure between lower limb joints, and perform three-dimensional motion analysis to obtain the final tracking motion values. Compared with the prior art, when an athlete conducts drag resistance running training, the waveform change of the pressure between the lower limb joints of the athlete during the movement can be understood by tracking the change of the motion values, so that the coach can further improve the effect of guiding the athlete to carry out sports training.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and more specifically, to an intelligent method for obtaining the pressure between lower limb joints during sprinting by resistance torsion. Background Art

[0002] Sprinting is a sport that aims to complete a fixed distance in the shortest possible time and maximize speed. Its main energy supply method is anaerobic metabolism, which requires high physical fitness from athletes and belongs to a periodic speed-strength event. To some extent, the specific strength quality of athletes determines the level of their sports performance. It affects and promotes other qualities and is the basis for sprint athletes to master sports skills and improve their performance.

[0003] In sprinting, lower limb strength is one of the physical fitness elements that determine sprinting performance. The common means to improve lower limb strength mainly involve drag-resistance running, that is, applying drag resistance (i.e., resistance torsion) to athletes during their training process.

[0004] At the same time, sprinting is a periodic speed-strength event, and its performance largely depends on the optimization of specific techniques and the effective development of specific strength. Currently, the relevant research on drag-resistance running mainly focuses on aspects such as kinematic changes in techniques, load effects, and training effects. Its drawback is that traditional drag-resistance running training has not analyzed the pressure between lower limb joints during an athlete's sprinting process under the condition of applying resistance torsion. It cannot accurately analyze the dynamic changes in related muscle groups and joint load changes of athletes during drag-resistance running training, thereby increasing the risk of sports injuries to athletes and causing mistakes in training plans. Summary of the Invention

[0005] The present invention provides an intelligent method for obtaining the pressure between lower limb joints during sprinting by resistance torsion to solve the problem that the prior art cannot accurately analyze the dynamic changes in related muscle groups and joint load changes of athletes during drag-resistance running training, thereby increasing the risk of sports injuries to athletes and causing mistakes in training plans.

[0006] An intelligent method for obtaining the pressure between lower limb joints during sprinting by resistance torsion includes the following steps:

[0007] Step 1: For each athlete on the sports field, obtain corresponding several six-degree-of-freedom values and a resistance torsion value during the athlete's movement process; at the same time, perform video acquisition on several cameras pre-set on the sports field to obtain a video acquisition result;

[0008] Step 2: Preprocess each of the six-degree-of-freedom values to obtain corresponding effective six-degree-of-freedom values;

[0009] Step 3: Preprocess the video acquisition result to obtain a three-dimensional coordinate value representing the three-dimensional data of the athlete's movement.

[0010] Step 4: Input each of the effective degree-of-freedom values and the resistance torsion value into a pre-trained motion capture model to obtain a kinematic and dynamic value representing the pressure between the lower limb joints.

[0011] Step 5: According to the three-dimensional coordinate value and the kinematic and dynamic value, perform three-dimensional motion analysis through a pre-set image recognition and deep learning model to obtain a tracking motion value representing the athlete's tracking motion data.

[0012] Advantages and beneficial effects of the method of the present invention: Through the motion capture model, obtain the kinematic and dynamic values of the pressure between the lower limb joints, and input the values and the obtained three-dimensional coordinate value into the image recognition and deep learning model for three-dimensional motion analysis to obtain the final tracking motion value; compared with the prior art, when the athlete conducts drag resistance running training, the waveform change of the pressure between the lower limb joints of the athlete during the movement can be understood through the change of the tracking motion value, so that the coach can further improve the effect of guiding the athlete's sports training.

[0013] Preferably, in step 2, low-pass filtering is performed on each of the six-degree-of-freedom values to obtain the corresponding effective six-degree-of-freedom value.

[0014] Preferably, step 3 includes:

[0015] Step 31: Analyze the video acquisition result, and obtain several pictures in the same time-sequence space according to the analysis result.

[0016] Step 32: Analyze each of the pictures to obtain an analysis result.

[0017] Step 33: According to the analysis result and the effective six-degree-of-freedom value, obtain the three-dimensional coordinate value representing the three-dimensional data of the athlete's movement.

[0018] Preferably, step 4 includes:

[0019] Step 41: Input each of the effective degree-of-freedom values and the resistance torsion value into the pre-trained motion capture model to obtain a gravitational moment, a muscle moment, a joint angular velocity, a joint angular acceleration, and an external moment.

[0020] Step 42: Calculate a joint net moment according to the gravitational moment, the muscle moment, the joint angular velocity, the joint angular acceleration, and the external moment; the joint net moment is the kinematic and dynamic value.

[0021] Preferably, after performing the step 41, it further includes calculating an inertial torque according to the link angular velocity and the link angular acceleration.

[0022] Preferably, in the step 42, a joint net torque is calculated according to the gravity torque, the muscle torque, the inertial torque, and the external torque; the joint net torque is the kinematic and dynamic value.

[0023] Preferably, the lower limb joints include a hip joint, a knee joint, and an ankle joint. In the step 4, the effective degree-of-freedom values and the resistance torsion values are input into a pre-trained motion capture model to obtain a kinematic and dynamic value representing the pressure between the hip joint, the knee joint, and the ankle joint.

[0024] Preferably, the camera is a monocular camera or a binocular camera. In the step 1, for each athlete in the sports field, a corresponding number of six-degree-of-freedom values and a resistance torsion value are obtained during the athlete's movement; at the same time, video acquisition is performed on the monocular camera or the binocular camera pre-set in the sports field to obtain a video acquisition result. Description of the Drawings

[0025] Figure 1 It is a schematic flowchart of the intelligent method of the present invention;

[0026] Figure 2 It is a detailed flowchart of step 3 of the intelligent method of the present invention;

[0027] Figure 3 It is a detailed flowchart of step 4 of the intelligent method of the present invention. Detailed Embodiments

[0028] The following is a description of the preferred embodiments of the present invention with reference to the drawings. It should be understood that the preferred embodiments described herein are only for the purpose of illustrating and explaining the present invention, and are not used to limit the present invention.

[0029] The present invention provides an intelligent method for obtaining the pressure between lower limb joints during sprinting with resistance torsion, so as to solve the problem that the prior art cannot accurately analyze the dynamic changes of related muscle groups during the drag resistance running training of athletes, resulting in the inability to further improve the guiding effect of sports training.

[0030] Combined with Figure 1 As shown, an intelligent method for obtaining the pressure between lower limb joints during sprinting with resistance torsion includes the following steps:

[0031] Step 1. For each athlete in the sports field, obtain a corresponding number of six-degree-of-freedom values and a resistance torsion value during the athlete's movement; at the same time, collect video through a number of cameras pre-set in the sports field to obtain a video collection result;

[0032] Step 2. Preprocess each of the six-degree-of-freedom values to obtain corresponding effective six-degree-of-freedom values;

[0033] Step 3. Preprocess the video collection result to obtain a three-dimensional coordinate value representing the three-dimensional data of the athlete's movement;

[0034] Step 4. Input each of the effective degree-of-freedom values and the resistance torsion value into a pre-trained motion capture model to obtain a kinematic and dynamic value representing the pressure between the lower limb joints;

[0035] Step 5. According to the three-dimensional coordinate value and the kinematic and dynamic value, perform three-dimensional motion analysis through a pre-set image recognition and deep learning model to obtain a tracking motion value representing the tracking motion data of the athlete.

[0036] Specifically, through the motion capture model, obtain the kinematic and dynamic values of the pressure between the lower limb joints, and input this value and the obtained three-dimensional coordinate value into the image recognition and deep learning model for three-dimensional motion analysis to obtain the final tracking motion value; compared with the prior art, when an athlete is performing drag resistance running training, the change situation of the pressure between the lower limb joints of the athlete during the movement can be understood through the change situation of the tracking motion value, so that the coach can further improve the effect of guiding the athlete's sports training.

[0037] During the sprint, multiple cameras and two three-dimensional force platforms are used. The three-dimensional motion capture technology of multiple cameras is used to collect the kinematic data of the athlete (i.e., the video collection result). Among them, the three-dimensional force platform is buried in the center of the plastic runway, and the surface is covered with the same PU material as the runway. At a certain distance from the starting point, at the same time, a number of six-degree-of-freedom activation signals are obtained through the ground pressure sensor device, and this six-degree-of-freedom activation signal is the six-degree-of-freedom value of this application. After the regular warm-up, the subject starts running in a standing position and sequentially completes the resistance torsion sprint test from no load to the preset specified pulling force with the maximum ability.

[0038] In a preferred embodiment of the present invention, in the step 2, low-pass filtering is performed on each of the six-degree-of-freedom values to obtain the corresponding effective six-degree-of-freedom values.

[0039] Specifically, several six-degree-of-freedom activation signals obtained by the ground pressure sensor device, preferably two in number in the present invention, and the six-degree-of-freedom activation signals are the six-degree-of-freedom values of this application. After low-pass filtering, effective six-degree-of-freedom values can be obtained.

[0040] Combined with Figure 2 As shown, in a preferred embodiment of the present invention, step 3 includes:

[0041] Step 31: Analyze the video acquisition result, and obtain several pictures in the same time-sequence space according to the analysis result;

[0042] Step 32: Analyze each of the pictures to obtain an analysis result;

[0043] Step 33: According to the analysis result and the effective six-degree-of-freedom value, obtain the three-dimensional coordinate value representing the three-dimensional data of the athlete's movement.

[0044] Combined with Figure 3 As shown, in a preferred embodiment of the present invention, step 4 includes:

[0045] Step 41: Input each of the effective degree-of-freedom values and the resistance torsion value into the pre-trained motion capture model to obtain a gravitational moment, a muscle moment, a joint angular velocity, a joint angular acceleration, and an external moment;

[0046] Step 42: Calculate a joint net moment according to the gravitational moment, the muscle moment, the joint angular velocity, the joint angular acceleration, and the external moment; the joint net moment is the kinematic and dynamic value.

[0047] To further optimize the above solution, after executing step 41, it further includes calculating an inertial moment according to the joint angular velocity and the joint angular acceleration.

[0048] To further optimize the above solution, in step 42, a joint net moment is calculated according to the gravitational moment, the muscle moment, the inertial moment, and the external moment; the joint net moment is the kinematic and dynamic value.

[0049] Specifically, through the motion capture model, the muscle moment between the lower limb joints and the dynamic interaction between the links are calculated. In this model, the calculation formula is;

[0050] N ET = M US + G RA + M DT + E XF ;

[0051] Among them, NET is the joint net moment;

[0052] G RA is the gravity moment;

[0053] M US is the muscle moment;

[0054] M DT is the inertia moment;

[0055] E XF is the external moment.

[0056] In this formula, G RA is the moment generated by gravity at the joint, M DT is the moment generated by movement, that is, the sum of the moments generated by the segment angular velocity and the segment angular acceleration at the joint, E XF is the moment of the ground reaction force during the support phase of the sprint action at the joint, M US is the moment generated by the muscle at the joint.

[0057] Specifically, by simulating effective six - degree - of - freedom values, that is, the synchronization of six - degree - of - freedom activation signals, reconstructing the labeled three - dimensional trajectory and ground reaction force data, after low - pass filtering, through inverse dynamic modeling, and using a human body full - body dynamics model, fitting the segment angular velocity and the segment angular velocity, the inertia moment is calculated.

[0058] In a preferred embodiment of the present invention, the lower limb joints include the hip joint, the knee joint, and the ankle joint. In step 4, the respective effective degree - of - freedom values and the resistance torsion values are input into a pre - trained motion capture model to obtain a kinematic and dynamic value representing the pressure between the hip joint, the knee joint, and the ankle joint.

[0059] In a preferred embodiment of the present invention, the camera is a monocular camera or a binocular camera. In step 1, for each athlete in the sports field, a corresponding number of the six - degree - of - freedom values and a resistance torsion value are obtained during the movement of the athlete; at the same time, by performing video acquisition on the monocular camera or the binocular camera pre - set in the sports field, a video acquisition result is obtained.

[0060] In sprinting, based on the data feedback from the motion capture model, image recognition, and deep learning model, the kinematic and kinetic values between the lower limb joints of athletes are further updated and iterated. By repeatedly measuring the kinematic, kinetic, and three-dimensional coordinate values during the running process of athletes, and through the pre-set image recognition and deep learning model for three-dimensional motion analysis, a tracking motion value representing the athlete's tracking motion data is obtained. The optimized technical application and sports performance status are evaluated, the differences before and after the tracking motion value are compared, and the effectiveness and benefit ratio are evaluated.

[0061] Based on multiple cameras, the three-dimensional coordinate values of the human body, i.e., the three-dimensional coordinate system, are constructed. By analyzing multiple frames captured in the same temporal space (i.e., the video acquisition results), combining the spatial calibration data and the recognition of the motion environment, the three-dimensional data of the motion is synthesized. Through the pre-set image recognition and deep learning model for three-dimensional motion analysis, the automatic identification of human body key points without markers is realized, helping athletes and coaches comprehensively analyze sports problems and track training data.

[0062] Combined with the mechanical characteristics of specific sports, a special technical plan for motion action recognition, evaluation, and feedback is formulated. By re-learning the tracking motion values of different athletes, the recognition errors and inaccuracies of special actions are supplemented, and the integrity of the model recognition is improved.

[0063] The specific implementation of this application is as follows: 1. Based on a markerless three-dimensional motion capture system (i.e., the aforementioned two three-dimensional force plates and several cameras, and the cameras in this application are preferably infrared cameras), several six-degree-of-freedom values and resistance torsion values of the athlete are obtained. At the beginning of each data acquisition, it is induced by the three-dimensional motion capture system, and two six-degree-of-freedom activation signals captured by the ground pressure sensor device (after low-pass filtering the six-degree-of-freedom activation signals, the corresponding effective six-degree-of-freedom values are obtained) are combined to obtain the kinematic and dynamic values between the lower limb joints; 2. The net knee joint force is determined through inverse dynamic modeling by calculating the joint angular velocity and joint angular acceleration, and the inertial torque is calculated using the human body whole-body dynamics model combined with the resistance torsion value. Finally, based on the inertial torque, gravitational torque, muscle torque, and external torque, the joint net torque is calculated; 3. A signal matrix is created by vertically connecting all the signals and time series of all the tests respectively. The neural network is used to map the signals of all the movements to the time series of all the movements, and in the MATLAB neural network toolbox, the signal matrix is used as the input and the time series matrix is used as the target output. The backpropagation error correction of the dedicated algorithm is used, combined with the spatial calibration data and the recognition of the motion environment, to synthesize the three-dimensional data of the motion. Through the pre-set image recognition and deep learning model (preferably using the OpenSim full body musculoskeletal model), three-dimensional motion analysis is carried out, and finally the three-dimensional coordinate values of the three-dimensional data representing the athlete's motion are obtained, realizing the markerless automatic recognition of human key points, and helping athletes and coaches comprehensively analyze motion problems and track training data.

[0064] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent method for obtaining the pressure between lower limb joints during sprinting by means of resistance torsion, characterized in that Including the following steps: Step 1: For each athlete in the sports field, obtain a corresponding number of six-degree-of-freedom values and a resistance torsion value during the athlete's movement; at the same time, perform video acquisition on several cameras preset in the sports field to obtain a video acquisition result; Step 2: Preprocess each of the six-degree-of-freedom values to obtain corresponding effective six-degree-of-freedom values; Step 3: Preprocess the video acquisition result to obtain a three-dimensional coordinate value representing the three-dimensional data of the athlete's movement; Step 4: Input each effective degree-of-freedom value and the resistance torsion value into a pre-trained motion capture model to obtain a kinematic and dynamic value representing the pressure between the lower limb joints; Step 5: According to the three-dimensional coordinate value and the kinematic and dynamic value, perform three-dimensional motion analysis through a preset image recognition and deep learning model to obtain a tracking motion value representing the athlete's tracking motion data.

2. The intelligent method for obtaining the pressure between lower limb joints during sprinting by resistance torsion according to claim 1, characterized in that, In Step 2, perform low-pass filtering on each of the six-degree-of-freedom values to obtain the corresponding effective six-degree-of-freedom values.

3. The intelligent method for obtaining the pressure between lower limb joints during sprinting by resistance torsion according to claim 1, characterized in that, Step 3 includes: Step 31: Analyze the video acquisition result, and obtain several frames in the same time-sequence space according to the analysis result; Step 32: Analyze each of the frames to obtain an analysis result; Step 33: According to the analysis result and the effective six-degree-of-freedom values, obtain the three-dimensional coordinate value representing the three-dimensional data of the athlete's movement.

4. The intelligent method for obtaining the pressure between lower limb joints during sprinting by resistance torsion according to claim 1, characterized in that, Step 4 includes: Step 41: Input each of the effective degree-of-freedom values and the resistance torsion value into the pre-trained motion capture model to obtain a gravitational moment, a muscle moment, a segment angular velocity, a segment angular velocity, and an external moment; Step 42: Calculate a joint net moment according to the gravitational moment, the muscle moment, the segment angular velocity, the segment angular velocity, and the external moment; the joint net moment is the kinematic and dynamic value.

5. The intelligent method for obtaining the pressure between lower limb joints during sprinting by resistance torsion according to claim 4, wherein, After executing Step 41, it further includes calculating an inertial moment according to the segment angular velocity and the segment angular acceleration.

6. The intelligent method for obtaining the inter-joint pressure of the lower limbs during sprinting by resistance torsion according to claim 5, characterized in that, In Step 42, calculate a joint net moment according to the gravitational moment, the muscle moment, the inertial moment, and the external moment; the joint net moment is the kinematic and dynamic value.

7. The intelligent method for obtaining the pressure between lower limb joints during sprinting by resistance torsion according to claim 1, characterized in that, The lower limb joints include the hip joint, the knee joint, and the ankle joint. In Step 4, input each of the effective degree-of-freedom values and the resistance torsion value into the pre-trained motion capture model to obtain a kinematic and dynamic value representing the pressure between the hip joint, the knee joint, and the ankle joint.

8. The intelligent method for obtaining the pressure between lower limb joints during sprinting by resistance torsion according to claim 1, wherein, The camera is a monocular camera or a binocular camera. In Step 1, for each athlete in the sports field, obtain a corresponding number of six-degree-of-freedom values and a resistance torsion value during the athlete's movement; at the same time, perform video acquisition on the monocular camera or the binocular camera preset in the sports field to obtain a video acquisition result.

Images