Method and device for estimating energy consumption of combat sports
By obtaining the athlete's biological data and motion images, using the key point detection model to calculate the joint angle and input the energy consumption estimation model, the fighting sports energy consumption measurement without wearing equipment is realized, and the portable gas analyzer affects sports performance and operation is solved. Complex problems provide simple and accurate energy consumption estimates.
Patent Information
- Application Number
- CN202311102518.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-29
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2043-08-29
AI Technical Summary
The existing portable gas analyzers will affect their performance when measuring fighting sports energy consumption, the operation process is complex and expensive, and the equipment has high requirements for the conditions of the experimental site.
By obtaining the athlete's biological data and moving images, using the key point detection model to extract joint position data, calculate joint angles, and input a preset energy consumption estimation model for detection, estimate the athlete's energy consumption and realize energy without wearing equipment Consumption measurement.
It simplifies the operation process, reduces costs, avoids the interference of equipment on sports performance, provides accurate energy consumption estimation results, and data collection is possible without professional training.
Smart Images

Figure CN117275087B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of energy consumption detection, and in particular to a method and device for estimating energy consumption of fighting sports. Background Art
[0002] Energy expenditure clearly reflects the basic consumption and the intensity and duration of physical activity over a period of time. The energy expenditure generated by physical activity is measured in metabolic equivalents (METs), which are based on energy expenditure at rest and measure the energy expenditure generated by engaging in various activities. The MET is defined as the exercise intensity of 1 MET (1 MET = 3.5 mL / min / kg) when the human body consumes 3.5 ml of oxygen for 1 minute of activity per kilogram of body weight at rest.
[0003] The current energy consumption measurement methods are mainly divided into direct test method and indirect test method. The direct test method is to measure the heat generated by the relevant body in a closed and heat-insulated space, and convert it into the metabolic rate per unit time. The direct test method is relatively cumbersome in equipment design and processing flow, and is restricted by the application site.
[0004] Indirect testing methods include: double-labeled water method, gas metabolism analysis method and other testing methods. Among them, the double-labeled water method is the most accurate method for evaluating energy consumption at present. The essence of the double-labeled water method is to let the experimental subject ingest a small amount of water labeled with 180 and 2H (0.3g / kg), and then collect urine samples every one to two days, and use an isotope mass spectrometer to measure the proportion of 180 and 2H in the urine sample, and then calculate the energy consumption value according to the loss ratio of 180 and 2H. However, the consumables required for this method are expensive, and cannot meet the calculation of energy consumption values of some sports in a very short period of time.
[0005] At present, the most commonly used method in sports science is gas metabolism analysis. Gas metabolism analysis is to let the test personnel wear respiratory protection equipment and use cardiopulmonary function telemeter to analyze the oxygen consumption (VO 2 ) and CO 2 The value is generated, and then the energy consumption of the human body is calculated. The oxidation reaction of the substrates sugar and fat, which are the most important for exercise consumption, must consume oxygen, thereby generating carbon dioxide. Under normal circumstances, the replacement rate of oxygen and carbon dioxide in the lungs is equivalent to the rate at which the body consumes oxygen and releases carbon dioxide. According to the theory, we can measure the respiratory gas and calculate the energy consumption.
[0006] For combat sports, such as Taekwondo, martial arts, and Aikido, portable gas analyzers are generally used to measure energy consumption. Compared with desktop gas analyzers, it has the significant advantages of small size and light weight, less interference with physical activity, and can accurately detect energy consumption during physical exercise. Therefore, in the process of exploring special sports and energy consumption, this type of method has long been widely used. Before using the instrument, it is necessary to warm up for half an hour, then calibrate the gas volume, then calibrate the gas composition of the indoor air, and then connect the standard air for calibration. Make sure the test is carried out at an indoor temperature of 20-25℃ and a relative humidity of 60%. After the instrument is calibrated, the tester needs to wear the breathing mask on the subject's breathing area, and then turn on the monitoring mode of the instrument. During the experiment, attention should also be paid to avoid problems such as leakage of the breathing mask and looseness of the instrument to ensure the normal operation of each test instrument.
[0007] However, the main disadvantage of using a portable gas analyzer to measure the energy expenditure of a specific sport is that the device will affect the subject's athletic performance. The portable gas analyzer is mainly composed of two parts: the analyzer body and the breathing mask. The structure of the breathing mask increases the subject's breathing resistance, which forces the subject to breathe with more force, causing the measured energy consumption to be greater than the actual energy consumption of the exercise. The analyzer body is fixed at the lower back. Although the current portable gas analyzers are generally lightweight, they will still be affected to varying degrees when the subject performs large movements or movements that are difficult to balance, resulting in errors between the test value and the actual value. Secondly, portable gas analyzers also have the disadvantages of being expensive, having a complicated operating process, and having high requirements for experimental site conditions. Summary of the invention
[0008] The present invention provides a method and device for estimating energy consumption of fighting sports, which are used to solve the problems that the existing method needs to wear a gas analyzer, affects sports performance, and has a complicated operation process.
[0009] The present invention provides a method for estimating energy consumption of fighting sports, comprising: acquiring biological data of a person to be estimated, and acquiring spatial position data of all joint points of the person to be estimated by acquiring motion images of the person to be estimated based on an acquisition device;
[0010] Based on the spatial position data of all the related nodes, the angle between each skeleton joint point and two adjacent skeleton joint points is calculated to obtain joint angle data;
[0011] Inputting the joint angle data and biological data into a preset energy consumption estimation model for detection to obtain an estimation result of the energy consumption of the person to be estimated;
[0012] The preset energy consumption estimation model is trained using joint angle sample data and biological sample data as samples, and energy consumption data corresponding to the samples as labels.
[0013] According to a method for estimating energy consumption of fighting sports provided by the present invention, the method collects motion images of a person to be estimated based on an acquisition device to obtain spatial position data of all joint points of the person to be estimated, including:
[0014] Acquire a collection device to collect motion video of the person to be estimated;
[0015] Performing image processing on the motion video to obtain a motion image of the person to be estimated in the motion video;
[0016] The motion image is processed using a key point detection model to obtain spatial position data of all relevant nodes of the person to be estimated.
[0017] According to a method for estimating the energy consumption of fighting sports provided by the present invention, the motion image is processed using a key point detection model to obtain spatial position data of all relevant nodes of the person to be estimated, including:
[0018] Using a key point detection model, extracting joint points from the motion image to obtain the position coordinates of each joint point;
[0019] Each of the position coordinates is normalized to obtain the spatial position data of all related nodes.
[0020] According to a method for estimating energy consumption of fighting sports provided by the present invention, based on the spatial position data of all the related nodes, the angle between each skeleton joint point and its two adjacent skeleton joint points is calculated to obtain joint angle data, including:
[0021] Based on the spatial position data of all the joint points, the angle between each skeleton joint point and its two adjacent skeleton joint points is calculated using an angle calculation formula to obtain the joint angle data of the person to be estimated;
[0022] The angle calculation formula is:
[0023]
[0024] Among them, J 2 is the spatial position of a skeleton joint point to be calculated, J 1 and J 3 and J 2 The spatial positions of two adjacent joint points.
[0025] According to a method for estimating energy consumption of fighting sports provided by the present invention, the training step of the preset energy consumption estimation model includes:
[0026] The joint angle sample data and the biological sample data are used as samples, and the energy consumption data corresponding to the samples are used as labels to be divided into a training set and a test set;
[0027] Inputting the training set into the energy consumption basic model for training, thereby obtaining a trained energy consumption basic model;
[0028] After the trained basic energy consumption model is tested using the test set, when the trained basic energy consumption model meets the accuracy of classification and recognition, a preset energy consumption estimation model is obtained.
[0029] According to a method for estimating energy consumption of fighting sports provided by the present invention, the energy consumption basic model includes: an input layer, an intermediate layer and an output layer;
[0030] Among them, the input layer includes an attention layer, the intermediate layer includes a long short-term memory network layer, a dense layer and a linear regression layer, and the long short-term memory network layer, the dense layer and the linear regression layer are arranged in sequence from the input layer to the output layer.
[0031] The present invention also provides a device for estimating energy consumption of fighting sports, comprising:
[0032] An acquisition module is used to acquire biological data of the person to be estimated, and to acquire the motion image of the person to be estimated based on the acquisition device, and to acquire the spatial position data of all joint points of the person to be estimated;
[0033] A calculation module, used to calculate the angle between each skeleton joint point and two adjacent skeleton joint points based on the spatial position data of all the related nodes, to obtain joint angle data;
[0034] A detection module is used to input joint angle data and biological data into a preset energy consumption estimation model for detection to obtain an estimated result of the energy consumption of the person to be estimated;
[0035] The preset energy consumption estimation model is trained using joint angle sample data and biological sample data as samples, and energy consumption data corresponding to the samples as labels.
[0036] According to a combat sports energy consumption estimation device provided by the present invention, the acquisition module is used for acquiring the motion image of the person to be estimated based on the acquisition device to acquire the spatial position data of all joint points of the person to be estimated, including:
[0037] Acquire a collection device to collect motion video of the person to be estimated;
[0038] Performing image processing on the motion video to obtain a motion image of the person to be estimated in the motion video;
[0039] The motion image is processed using a key point detection model to obtain spatial position data of all relevant nodes of the person to be estimated.
[0040] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for estimating the energy consumption of fighting sports as described in any one of the above-mentioned methods is implemented.
[0041] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for estimating the energy consumption of combat sports as described in any one of the above is implemented.
[0042] The present invention provides a method and device for estimating energy consumption in combat sports. The method and device obtain biological data of a person to be estimated and spatial position data of all joints of the person to be estimated, thereby obtaining joint angle data. The angle and biological data of each skeleton joint point are input into a preset energy consumption estimation model for detection, thereby obtaining an estimated result of the energy consumption of the person to be estimated. As a result, the person to be estimated does not need to wear any detection equipment during exercise, thereby realizing energy consumption estimation, and the operation process is simpler and faster. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0044] Figure 1 It is a flow chart of the method for estimating the energy consumption of fighting sports provided by the present invention.
[0045] Figure 2 It is a schematic diagram of the angle between each skeleton joint point and its two adjacent skeleton joint points in the present invention.
[0046] Figure 3 A schematic diagram of testing the preset energy consumption estimation model in the present invention.
[0047] Figure 4 It is a structural schematic diagram of the device for estimating the energy consumption of fighting sports provided by the present invention.
[0048] Figure 5It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION
[0049] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0050] Combine the following Figure 1 , Figure 2 and Figure 3 A method for estimating energy consumption of fighting sports of the present invention is described, comprising:
[0051] S1. Obtain the biological data of the person to be estimated, and based on the acquisition device, collect the motion image of the person to be estimated, and obtain the spatial position data of all joints of the person to be estimated. In this embodiment, the acquisition device can be a camera. The joints include nose, left eye, right eye, left ear, right ear, right shoulder, left shoulder, right span, left span, right hand joint, left hand joint, right knee, left knee, right wrist, left wrist, right ankle and left ankle. At the same time, the biological data includes gender (in this embodiment, 0 represents male and 1 represents female), height (in centimeters, rounded to an integer), weight (in kilograms, rounded to an integer), age (calculated in years) and body fat percentage (multiplied by one hundred and rounded to an integer).
[0052] S2. Based on the spatial position data of all the related nodes, the angle between each skeleton joint point and two adjacent skeleton joint points is calculated to obtain joint angle data. Specifically, the joint angle data includes data of the left shoulder angle, the right shoulder angle, the left elbow angle, the right elbow angle, the left hip angle, the right hip angle, the left knee angle, and the right knee angle.
[0053] S3. Input the joint angle data and biological data into a preset energy consumption estimation model for detection to obtain an estimated result of the energy consumption of the person to be estimated.
[0054] The preset energy consumption estimation model is trained using joint angle sample data and biological sample data as samples, and energy consumption data corresponding to the samples as labels.
[0055] The present invention collects the motion images of the person to be estimated based on the acquisition device, obtains the spatial position data of all joints of the person to be estimated and the biological data of the person to be estimated, and then obtains the joint angle data, so as to realize the data collection more simply without the need for professional training. Then the angle and biological data of each skeleton joint point are input into the preset energy consumption estimation model for detection, and the estimated result of the energy consumption of the person to be estimated is obtained. The person to be estimated does not need to wear any detection equipment when exercising, so that the energy consumption can be estimated, and the operation process is simpler and faster. At the same time, the present invention only needs a video acquisition device and a computer with a certain computing power to complete the estimation of energy consumption, which is cheaper than the existing detection through a portable gas analyzer.
[0056] On the basis of the above embodiment, the acquisition device is used to acquire the motion image of the person to be estimated, and obtain the spatial position data of all joint points of the person to be estimated, including:
[0057] Acquire the acquisition device to collect the motion video of the person to be estimated. The acquisition device collects the motion video of the person to be estimated to achieve contactless data collection, so that the person to be estimated, such as an athlete, can collect data during daily training, avoiding the wearing of existing detection equipment that affects the sports performance of the person to be estimated.
[0058] The motion video is subjected to image processing to obtain a motion image of the person to be estimated in the motion video.
[0059] The motion image is processed using a key point detection model to obtain spatial position data of all relevant nodes of the person to be estimated. Specifically, the key point detection model includes an OpenPose method.
[0060] On the basis of the above embodiment, the motion image is processed using a key point detection model to obtain spatial position data of all nodes related to the person to be estimated, including:
[0061] The key point detection model is used to extract the joint points of the motion image to obtain the position coordinates of each joint point. In this embodiment, the coordinates of the joint points can be expressed as L with the upper left corner as the origin. n =(a n ,b n ), the value range of n is from 0 to 16,
[0062] Each of the position coordinates is normalized to obtain the spatial position data of all the related nodes. Specifically, the position coordinates of each joint point are normalized, and the normalized position coordinates of the 17 joint points constitute a coordinate matrix P.
[0063] P=[(x 0 ,y0 ),(x 2 ,y 2 ),...,(x k ,y k ),...,(x 16 ,y 16 )];
[0064] Among them, (x k ,y k ) represents the coordinate position of the kth joint point after normalization, which is composed of the horizontal coordinate x and the vertical coordinate y.
[0065] Based on the above embodiments, please refer to Figure 2 , based on the spatial position data of all the related nodes, the angle between each skeleton joint point and its two adjacent skeleton joint points is calculated to obtain joint angle data, including:
[0066] Based on the spatial position data of all the joint points, the angle between each skeleton joint point and its two adjacent skeleton joint points is calculated using the angle calculation formula to obtain the joint angle data of the person to be estimated.
[0067] The angle calculation formula is:
[0068]
[0069] Among them, J 2 is the spatial position of a skeleton joint point to be calculated, J 1 and J 3 and J 2 The spatial position of two adjacent joint points. In this embodiment, the angle at the right knee joint is taken as an example, where j 1 、j 2 and j 3 They respectively represent the right hip joint (the upper joint of the right knee joint), the right knee joint (the target joint) and the right ankle (the lower joint of the right knee joint), and the angle at the right knee joint is calculated by the angle calculation formula.
[0070] Based on the above embodiment, the training step of the preset energy consumption estimation model includes:
[0071] The joint angle sample data and the biological sample data are used as samples, and the energy consumption data corresponding to the samples are used as labels to be divided into a training set and a test set. The training set and the test set can be divided using a python script file to increase the division speed.
[0072] The training set is input into the energy consumption basic model for training to obtain the trained energy consumption basic model.
[0073] In this embodiment, the energy consumption basic model includes: an input layer, an intermediate layer and an output layer.
[0074] The input layer includes an attention layer, the intermediate layer includes a long short-term memory network layer (Long Short-Term Memory, LSTM), a dense layer and a linear regression layer, and the long short-term memory network layer, dense layer and linear regression layer are arranged in sequence from the input layer to the output layer, so as to first input the sample and the label to the attention layer (Attention Layer), and then use the output of the attention layer as the input of the LSTM layer.
[0075] The accuracy of the output model of the output layer of the energy consumption estimation model is calculated using the mean square error (MSE). The calculation formula is:
[0076]
[0077] Where: n is the amount of data in the test set, i is the sequence number of the data, f(x i ) is the predicted value given by the model, y i is the energy consumption data (energy consumption value) measured by the gas analyzer. During retraining, the lower the error value, the better the prediction effect.
[0078] The device for estimating the energy consumption of fighting sports provided by the present invention is described below. The device for estimating the energy consumption of fighting sports described below and the method for estimating the energy consumption of fighting sports described above can be referred to each other.
[0079] Please also read Figure 4 , a device for estimating energy consumption of fighting sports, including: an acquisition module 410, a calculation module 420 and a detection module 430.
[0080] The acquisition module 410 is used to acquire biological data of the person to be estimated, and to acquire spatial position data of all joint points of the person to be estimated by acquiring motion images of the person to be estimated based on acquisition equipment.
[0081] The calculation module 420 is used to calculate the angle between each skeleton joint point and its two adjacent skeleton joint points based on the spatial position data of all the related nodes to obtain joint angle data.
[0082] The detection module 430 is used to input the joint angle data and the biological data into a preset energy consumption estimation model for detection, so as to obtain an estimation result of the energy consumption of the person to be estimated.
[0083] The preset energy consumption estimation model is trained using joint angle sample data and biological sample data as samples, and energy consumption data corresponding to the samples as labels.
[0084] The present invention acquires biological data of the person to be estimated through the acquisition module 410, and acquires the motion image of the person to be estimated based on the acquisition device, acquires the spatial position data of all joints of the person to be estimated, the calculation module 420 obtains the estimated result of the energy consumption of the person to be estimated, and the detection module 430 obtains the estimated result of the energy consumption of the person to be estimated. As a result, the person to be estimated does not need to wear any detection equipment when exercising, and the energy consumption can be estimated, and the operation process is simpler and faster.
[0085] The acquisition module 410 is used to acquire the motion image of the person to be estimated based on the acquisition device, and acquire the spatial position data of all joint points of the person to be estimated, including:
[0086] Acquire the acquisition device to collect the motion video of the person to be estimated.
[0087] The motion video is subjected to image processing to obtain a motion image of the person to be estimated in the motion video.
[0088] The motion image is processed using a key point detection model to obtain spatial position data of all nodes related to the person to be estimated. Specifically, it includes:
[0089] The key point detection model is used to extract the joint points of the motion image to obtain the position coordinates of each joint point.
[0090] Each of the position coordinates is normalized to obtain the spatial position data of all related nodes.
[0091] The calculation module 420 is specifically used to: calculate the angle between each skeleton joint point and its two adjacent skeleton joint points based on the spatial position data of all the joint points, using the angle calculation formula, to obtain the joint angle data of the person to be estimated;
[0092] The angle calculation formula is:
[0093]
[0094] Among them, J 2 is the spatial position of a skeleton joint point to be calculated, J 1 and J 3 and J 2 The spatial positions of two adjacent joint points.
[0095] Figure 5 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 5 As shown, the electronic device may include: a processor 510, a communication interface 520, a memory 530 and a communication bus 540, wherein the processor 510, the communication interface 520 and the memory 530 communicate with each other via the communication bus 540. The processor 510 may call the logic instructions in the memory 530 to execute the method for estimating the energy consumption of fighting sports, which method includes:
[0096] S1. Acquire biological data of the person to be estimated, and acquire motion images of the person to be estimated based on an acquisition device to acquire spatial position data of all joint points of the person to be estimated.
[0097] S2. Based on the spatial position data of all the related nodes, calculate the angle between each skeleton joint point and its two adjacent skeleton joint points to obtain joint angle data.
[0098] S3. Input the joint angle data and biological data into a preset energy consumption estimation model for detection to obtain an estimated result of the energy consumption of the person to be estimated.
[0099] The preset energy consumption estimation model is trained using joint angle sample data and biological sample data as samples, and energy consumption data corresponding to the samples as labels.
[0100] In addition, the logic instructions in the above-mentioned memory 530 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.
[0101] On the other hand, the present invention further provides a computer program product, the computer program product includes a computer program, the computer program can be stored in a non-transitory computer-readable storage medium, when the computer program is executed by a processor, the computer can execute the method for estimating the energy consumption of fighting sports provided by the above methods, the method comprising:
[0102] S1. Acquire biological data of the person to be estimated, and acquire motion images of the person to be estimated based on an acquisition device to acquire spatial position data of all joint points of the person to be estimated.
[0103] S2. Based on the spatial position data of all the related nodes, calculate the angle between each skeleton joint point and its two adjacent skeleton joint points to obtain joint angle data.
[0104] S3. Input the joint angle data and biological data into a preset energy consumption estimation model for detection to obtain an estimated result of the energy consumption of the person to be estimated.
[0105] The preset energy consumption estimation model is trained using joint angle sample data and biological sample data as samples, and energy consumption data corresponding to the samples as labels.
[0106] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the computer program is executed by a processor to perform the method for estimating the energy consumption of fighting sports provided by the above methods, the method comprising:
[0107] S1. Acquire biological data of the person to be estimated, and acquire motion images of the person to be estimated based on an acquisition device to acquire spatial position data of all joint points of the person to be estimated.
[0108] S2. Based on the spatial position data of all the related nodes, calculate the angle between each skeleton joint point and its two adjacent skeleton joint points to obtain joint angle data.
[0109] S3. Input the joint angle data and biological data into a preset energy consumption estimation model for detection to obtain an estimated result of the energy consumption of the person to be estimated.
[0110] The preset energy consumption estimation model is trained using joint angle sample data and biological sample data as samples, and energy consumption data corresponding to the samples as labels.
[0111] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.
[0112] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0113] 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 it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for estimating energy consumption in combat sports. It is characterized in that include: Acquire biological data of the person to be estimated, and acquire motion images of the person to be estimated based on an acquisition device to acquire spatial position data of all joint points of the person to be estimated; Based on the spatial position data of all the related nodes, the angle between each skeleton joint point and two adjacent skeleton joint points is calculated to obtain joint angle data; Inputting the joint angle data and biological data into a preset energy consumption estimation model for detection to obtain an estimation result of the energy consumption of the person to be estimated; The preset energy consumption estimation model is trained using joint angle sample data and biological sample data as samples, and energy consumption data corresponding to the samples as labels; The step of calculating the angle between each skeleton joint point and two adjacent skeleton joint points based on the spatial position data of all the related nodes to obtain joint angle data includes: Based on the spatial position data of all the joint points, the angle between each skeleton joint point and its two adjacent skeleton joint points is calculated using an angle calculation formula to obtain the joint angle data of the person to be estimated; The angle calculation formula is: Among them, J 2 is the spatial position of a skeleton joint point to be calculated, J 1 and J 3 and J 2 The spatial positions of two adjacent joints; The training step of the preset energy consumption estimation model includes: The joint angle sample data and the biological sample data are used as samples, and the energy consumption data corresponding to the samples are used as labels to be divided into a training set and a test set; Inputting the training set into the energy consumption basic model for training, thereby obtaining a trained energy consumption basic model; After testing the trained basic energy consumption model using the test set, when the trained basic energy consumption model meets the accuracy of classification recognition, a preset energy consumption estimation model is obtained; The energy consumption basic model includes: an input layer, an intermediate layer and an output layer; Among them, the input layer includes an attention layer, the intermediate layer includes a long short-term memory network layer, a dense layer and a linear regression layer, and the long short-term memory network layer, the dense layer and the linear regression layer are arranged in sequence from the input layer to the output layer.
2. The method for estimating energy consumption of combat sports according to claim 1, It is characterized in that The method of collecting the motion image of the person to be estimated based on the acquisition device to obtain the spatial position data of all joint points of the person to be estimated includes: Acquire a collection device to collect motion video of the person to be estimated; Performing image processing on the motion video to obtain a motion image of the person to be estimated in the motion video; The motion image is processed using a key point detection model to obtain spatial position data of all relevant nodes of the person to be estimated.
3. The method for estimating energy consumption of combat sports according to claim 2, It is characterized in that The motion image is processed using a key point detection model to obtain spatial position data of all relevant nodes of the person to be estimated, including: Using a key point detection model, extracting joint points from the motion image to obtain the position coordinates of each joint point; Each of the position coordinates is normalized to obtain the spatial position data of all related nodes.
4. A device for estimating energy consumption in combat sports, It is characterized in that include: An acquisition module is used to acquire biological data of the person to be estimated, and to acquire the motion image of the person to be estimated based on the acquisition device, and to acquire the spatial position data of all joint points of the person to be estimated; A calculation module, used to calculate the angle between each skeleton joint point and two adjacent skeleton joint points based on the spatial position data of all the related nodes, to obtain joint angle data; A detection module is used to input joint angle data and biological data into a preset energy consumption estimation model for detection to obtain an estimated result of the energy consumption of the person to be estimated; The preset energy consumption estimation model is trained using joint angle sample data and biological sample data as samples, and energy consumption data corresponding to the samples as labels; The step of calculating the angle between each skeleton joint point and two adjacent skeleton joint points based on the spatial position data of all the related nodes to obtain joint angle data includes: Based on the spatial position data of all the joint points, the angle between each skeleton joint point and its two adjacent skeleton joint points is calculated using an angle calculation formula to obtain the joint angle data of the person to be estimated; The angle calculation formula is: Among them, J 2 is the spatial position of a skeleton joint point to be calculated, J 1 and J 3 and J 2 The spatial positions of two adjacent joints; The training step of the preset energy consumption estimation model includes: The joint angle sample data and the biological sample data are used as samples, and the energy consumption data corresponding to the samples are used as labels to be divided into a training set and a test set; Inputting the training set into the energy consumption basic model for training, thereby obtaining a trained energy consumption basic model; After testing the trained basic energy consumption model using the test set, when the trained basic energy consumption model meets the accuracy of classification recognition, a preset energy consumption estimation model is obtained; The energy consumption basic model includes: an input layer, an intermediate layer and an output layer; Among them, the input layer includes an attention layer, the intermediate layer includes a long short-term memory network layer, a dense layer and a linear regression layer, and the long short-term memory network layer, the dense layer and the linear regression layer are arranged in sequence from the input layer to the output layer.
5. The device for estimating energy consumption of fighting sports according to claim 4, It is characterized in that The acquisition module is used to acquire the motion image of the person to be estimated based on the acquisition device, and acquire the spatial position data of all joint points of the person to be estimated, including: Acquire a collection device to collect motion video of the person to be estimated; Performing image processing on the motion video to obtain a motion image of the person to be estimated in the motion video; The motion image is processed using a key point detection model to obtain spatial position data of all relevant nodes of the person to be estimated.
6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the program, the method for estimating the energy consumption of fighting sports as described in any one of claims 1 to 3 is implemented.
7. A non-transitory computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the method for estimating the energy consumption of fighting sports as described in any one of claims 1 to 3 is implemented.
