Football flight trajectory prediction method based on motion sensor and related equipment

Abstract

The invention discloses a football flight trajectory prediction method based on a motion sensor and related equipment. The football flight trajectory prediction method based on the motion sensor comprises the following steps: acquiring football motion data when a football is kicked out within a preset time through the motion sensor; feature extraction is conducted on the football motion data through a trained football flight trajectory prediction model, the football kicking feature and the initial flight feature of the football motion data are obtained, and the trained football flight trajectory prediction model is obtained by training a pre-trained football flight trajectory prediction model through a training data set; and outputting a football motion track prediction result after a preset time based on a football kicking feature of the football motion data and the initial flight feature. The method can solve the problem that the existing method cannot accurately predict the flight path of the football based on artificial experience judgment.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and particularly to a football flight trajectory prediction method and related devices based on a motion sensor. Background Art

[0002] Football is a sport highly sought after by teenagers and is also recognized as the world's number one sport, which can play a good role in strengthening the body. In football training, accurately predicting the flight trajectory of a football is crucial for training players' defensive and offensive strategies, and can train and cultivate players' ball sense and reaction ability. However, traditional prediction methods are based on manual experience judgment and cannot accurately predict the flight trajectory of a football. Therefore, there is an urgent need for a football flight trajectory prediction method based on a motion sensor to solve the problem that existing methods are based on manual experience judgment and cannot accurately predict the flight trajectory of a football. Summary of the Invention

[0003] The present invention provides a football flight trajectory prediction method based on a motion sensor, aiming to solve the problem that existing methods are based on manual experience judgment and cannot accurately predict the flight trajectory of a football, and to achieve accurate prediction of the football flight trajectory.

[0004] To achieve the above object, the football flight trajectory prediction method based on a motion sensor proposed by the present invention, wherein the motion sensor is arranged in a hand wearable device, and the motion sensor is used to collect hand motion data corresponding to the hand wearable device. The method includes:

[0005] Obtaining football running data when the football is kicked within a preset time through the motion sensor;

[0006] Performing feature extraction on the football motion data through a trained football flight trajectory prediction model to obtain the kicking feature and the initial flight feature of the football motion data, and the trained football flight trajectory prediction model is obtained by training a pre-trained football flight trajectory prediction model with a training data set;

[0007] Based on the kicking feature and the initial flight feature of the football motion data, outputting a prediction result of the football motion trajectory after a preset time.

[0008] Optionally, before performing feature extraction on the football motion data through the trained football flight trajectory prediction model to obtain the kicking feature and the initial flight feature of the football motion data, the method further includes:

[0009] Obtaining a training data set and a pre-trained football flight trajectory prediction model, where the training data set includes sample data in a historical kicking process and corresponding football flight trajectory annotation data;

[0010] Train the pre-trained football flight trajectory prediction model with the sample data in the historical kicking process, and adjust the parameters of the pre-trained football flight trajectory prediction model with the corresponding football flight trajectory annotation data. After training is completed, a trained football flight trajectory prediction model is obtained.

[0011] Optionally, the sample data in the historical kicking process includes the kicking data when the football is kicked, the flight data within a preset time, and the flight data after the preset time. The football flight trajectory prediction model includes a first feature extraction network, a second feature extraction network, a third feature extraction network, and a feature fusion network. Training the pre-trained football flight trajectory prediction model with the sample data in the historical extraction process includes:

[0012] Perform first feature extraction on the kicking data when the football is kicked through the first feature extraction network to obtain the kicking feature of the sample data;

[0013] Perform second feature extraction on the flight data within the preset time through the second feature extraction network to obtain the first flight feature of the sample data;

[0014] Perform third feature extraction on the flight data after the preset time through the third feature extraction network to obtain the second flight feature of the sample data;

[0015] Perform fusion processing on the kicking feature, the first flight feature, and the second flight feature through the feature fusion network to obtain the flight trajectory feature of the sample data.

[0016] Optionally, the kicking data when the football is kicked includes the kicking force, the kicking angle, and the rotation angle of the ball. Performing first feature extraction on the kicking data when the football is kicked through the first physical sign extraction network to obtain the kicking feature of the sample data includes:

[0017] Perform force feature extraction on the kicking force through the first feature extraction network to obtain the kicking force feature;

[0018] Perform angle feature extraction on the kicking angle through the first feature extraction network to obtain the kicking angle feature;

[0019] Perform rotation angle feature extraction on the rotation angle of the ball through the first feature extraction network to obtain the rotation angle feature;

[0020] Based on the kicking force feature, the kicking angle feature, and the rotation angle feature, obtain the kicking feature of the sample data.

[0021] Optionally, the pre-trained football flight trajectory model further includes a trajectory feature extraction network. Adjusting the parameters of the pre-trained football flight trajectory prediction model using the corresponding football flight trajectory annotation data, after training is completed, obtaining the trained football flight trajectory prediction model includes:

[0022] Performing trajectory prediction on the flight trajectory features of the sample data through the trajectory feature extraction network to obtain the predicted trajectory data of the sample data;

[0023] Calculating the loss value between the predicted trajectory data of the sample data and the corresponding football flight trajectory annotation data through a preset loss function, and taking minimizing the loss value as the optimization objective to adjust the model parameters of the pre-trained football flight trajectory prediction model. After training is completed, obtaining the trained football flight trajectory prediction model.

[0024] Optionally, extracting the features of the football motion data through the trained football flight trajectory prediction model to obtain the kicking features and the initial flight features of the football motion data includes:

[0025] Performing first feature extraction on the football motion data through the first feature extraction network to obtain the kicking features of the football motion data;

[0026] Performing second feature extraction on the football motion data through the second feature extraction network to obtain the initial flight features of the football motion data.

[0027] Optionally, based on the kicking features and the initial flight features of the football motion data, outputting the football motion trajectory prediction result after a preset time includes:

[0028] Performing fusion processing on the kicking features and the initial flight features of the football motion data through the feature fusion network to obtain the flight trajectory features of the football motion data;

[0029] Performing trajectory prediction on the flight trajectory features of the football motion data through the trajectory feature extraction network and outputting the football motion trajectory prediction result after a preset time.

[0030] Optionally, after outputting the football motion trajectory prediction result after a preset time based on the kicking features and the initial flight features of the football motion data, the method further includes:

[0031] Constructing a two-dimensional plane of the ground;

[0032] Predicting the plane football landing point in the two-dimensional plane based on the football motion trajectory prediction result;

[0033] Based on the mapping relationship between the ground and the two-dimensional plane, determine the corresponding ground football landing point of the plane football landing point, and the ground football landing point is used to guide the player to move forward to the ground football landing point in advance.

[0034] The present invention also provides a football flight trajectory prediction system, which includes a server and a motion sensor disposed in the football, and the processor is wirelessly connected to the motion sensor;

[0035] The motion sensor is disposed in the football, and the motion sensor is configured to collect motion data corresponding to the football and transmit the collected motion data to the server;

[0036] When the server executes, it implements the steps of the football flight trajectory prediction method provided by the embodiment of the present invention based on the motion sensor.

[0037] The present invention also provides a football flight trajectory prediction device based on a motion sensor. The football flight trajectory prediction device based on a motion sensor includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of the above-mentioned football flight trajectory prediction method based on a motion sensor.

[0038] The technical solution of the football flight trajectory prediction method based on a motion sensor of the present invention obtains football motion data when the football is kicked within a preset time through the motion sensor; extracts features from the football motion data through a trained football flight trajectory prediction model to obtain the kicking features and initial flight features of the football motion data, and the trained football flight trajectory prediction model is obtained by training a pre-trained football flight trajectory prediction model with a training data set; based on the kicking features and the initial flight features of the football motion data, output the football motion trajectory prediction result after a preset time. The present invention extracts features from the football motion data when the football is kicked within a preset time through a trained football flight trajectory prediction model to obtain the kicking features and initial flight features of the football motion data, and utilizes the kicking features and initial flight features of the football motion data, thereby outputting the football motion trajectory prediction result after a preset time, solving the problem that the existing method is based on manual experience judgment and cannot accurately predict the flight trajectory of the football. Description of the Drawings

[0039] Figure 1 It is a schematic flowchart of the football flight trajectory prediction method based on a motion sensor provided by an embodiment of the present invention;

[0040] Figure 2Schematic structural diagram of a football flight trajectory prediction device based on a motion sensor provided by an embodiment of the present invention. Detailed implementation manners

[0041] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0042] It should be noted that all directional indications (such as up, down, left, right, front, back...) in the embodiments of the present invention are only used to explain the relative position relationship and movement conditions between components in a specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly.

[0043] It should also be noted that when an element is referred to as being "fixed to" or "disposed on" another element, it can be directly on the other element or there may be an intermediate element at the same time. When an element is referred to as being "connected" to another element, it can be directly connected to the other element or there may be an intermediate element at the same time.

[0044] In addition, the descriptions involving "first", "second", etc. in the present invention are only for descriptive purposes, and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the fact that those of ordinary skill in the art can implement them. When the combination of technical solutions appears to be contradictory or unable to be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present invention.

[0045] Refer to Figure 1 , Figure 1 which is a schematic flowchart of a football flight trajectory prediction method based on a motion sensor provided by an embodiment of the present invention.

[0046] In this embodiment, the above-mentioned motion sensor is disposed in the football, and the motion sensor is used to collect motion data corresponding to the football. The football flight trajectory prediction method based on the motion sensor includes the following steps:

[0047] Step S10: Obtain the football motion data when the football is kicked within a preset time through the motion sensor.

[0048] In this embodiment, the above-mentioned motion sensor can capture and record the football motion data during the process of dribbling the ball, such as speed, acceleration, direction, vibration frequency, etc., and can monitor the football motion state in real time and continuously monitor the change of the football motion state.

[0049] The above-mentioned preset time is the time period when the sensor starts to record data, which can specifically be within a few seconds before and after the football is kicked, such as within 1 second, within 2 seconds, within 3 seconds, etc. For example, if the preset time is 3 seconds, then the sensor will collect data in real time within 3 seconds before and after the football is kicked; if the preset time is 2 seconds, then the sensor will collect data in real time within 2 seconds before and after the football is kicked; if the preset time is 1 second, then the sensor will collect data in real time within 1 second before and after the football is kicked.

[0050] The above-mentioned football motion data can be motion data such as the speed, acceleration, rotation angle, and direction change of the football.

[0051] It should be noted that the motion sensor collects the motion data of the football when it is kicked in real time. Through these motion data, the motion state of the football at the moment of being kicked can be understood.

[0052] Step S20: Extract features from the football motion data through the trained football flight trajectory prediction model to obtain the kicking features and initial flight features of the football motion data.

[0053] In this embodiment, the above-mentioned trained football flight trajectory prediction model is obtained by training the pre-trained football flight trajectory prediction model with a training data set.

[0054] The above-mentioned football flight trajectory prediction model can be a football flight trajectory prediction model constructed based on machine learning or deep learning, such as a convolutional neural network (CNN), a recurrent neural network (RNN), etc.

[0055] The above-mentioned training data set includes sample data and corresponding football flight trajectory annotation data during historical kicking processes. The above-mentioned sample data includes kicking data when the football is kicked, flight data within the preset time, and flight data after the preset time.

[0056] The above-mentioned feature extraction can be understood as a processing process of identifying and extracting the features of the football flight trajectory from the football motion data.

[0057] The above-mentioned kicking features can be understood as various parameters and characteristics when kicking the ball, such as the force, angle, speed, direction, etc. of kicking the ball.

[0058] The above-mentioned initial flight features can be understood as the flight features that the football has at the moment of being kicked, such as the initial speed, initial angle, initial direction, etc.

[0059] It should be noted that the trained feature extraction model can identify football motion data features, such as sudden changes in acceleration, vibration frequency and amplitude, etc.

[0060] Step S30: Based on the kicking characteristics and initial flight characteristics of the football motion data, output the prediction result of the football motion trajectory after a preset time.

[0061] In this embodiment, the prediction result of the football motion trajectory after a preset time can be output according to the force, angle, speed, direction during kicking, and the initial speed, initial angle, and initial direction of the football when it is just kicked.

[0062] The above football motion trajectory prediction result includes information such as the flight trajectory, predicted landing point, flight time, etc. of the football after a preset time.

[0063] The above "after a preset time" is after a preset time point. For example, if the preset time is 1 second, then after a preset time can be 1 second later; if the preset time is 2 seconds, after a preset time can be 2 seconds later; if the preset time is 3 seconds, then after a preset time can be 3 seconds later, etc.

[0064] In this embodiment, the present invention captures the initial motion state of the football in real time through a sensor within a preset time, and uses the trained football flight trajectory prediction model to predict the football flight trajectory according to the initial motion state, and predicts the football motion trajectory after a preset time. According to the predicted football motion trajectory, the defensive player can adjust the position in advance to intercept the opponent's pass or shot; the offensive player can use the predicted football motion trajectory to select the best passing method or shooting opportunity.

[0065] Implementing the football flight trajectory prediction method based on a motion sensor of this embodiment, obtaining football motion data when the football is kicked within a preset time through the motion sensor; performing feature extraction on the football motion data through a trained football flight trajectory prediction model to obtain the kicking characteristics and initial flight characteristics of the football motion data, and the trained football flight trajectory prediction model is obtained by training a pre-trained football flight trajectory prediction model with a training data set; based on the kicking characteristics of the football motion data and the initial flight characteristics, output the prediction result of the football motion trajectory after a preset time. The present invention performs feature extraction on the football motion data when the football is kicked within a preset time through a trained football flight trajectory prediction model to obtain the kicking characteristics and initial flight characteristics of the football motion data, and uses the kicking characteristics and initial flight characteristics of the football motion data, thereby outputting the prediction result of the football motion trajectory after a preset time, and solves the problem that the existing method is based on manual experience judgment and cannot accurately predict the flight trajectory of the football.

[0066] It is understandable that in the specific embodiments of the present application, relevant data such as user data, control data, and motion data are involved. When the embodiments in the present application are applied to specific products or technologies, user permission or consent needs to be obtained, and the collection, use, and processing of relevant data, as well as the training and use of various models, need to comply with the relevant laws, regulations, and standards of relevant countries and regions.

[0067] Optionally, before the step of extracting features from football motion data through the trained football flight trajectory prediction model to obtain the kicking features and initial flight features of the football motion data, a training data set and a pre-trained football flight trajectory prediction model can also be obtained; the pre-trained football flight trajectory prediction model is trained with sample data in the historical kicking process, and the parameters of the pre-trained football flight trajectory prediction model are adjusted with the corresponding football flight trajectory annotation data. After training is completed, a trained football flight trajectory prediction model is obtained.

[0068] In the embodiments of the present invention, the above-mentioned training data set includes sample data in the historical kicking process and the corresponding football flight trajectory annotation data. The above-mentioned sample data in the historical kicking process can be understood as various data related to kicking recorded in football matches carried out in the past time period, including the force, angle, speed, direction, etc. when kicking, including the initial velocity, direction, and rotation situation when the football is kicked, including the corresponding football flight trajectory, and also including environmental influencing factors, such as wind speed, field conditions and other environmental factors that may affect the football flight trajectory. The above-mentioned football flight trajectory includes the route of the football moving in the air, the magnitude and direction of the football's speed at different time points, and the position where the football finally lands. The above-mentioned corresponding football flight trajectory annotation data is used to verify and optimize the accuracy of the football flight trajectory prediction model.

[0069] The above-mentioned pre-trained football flight trajectory prediction model can be a football flight trajectory prediction model constructed based on machine learning or deep learning, such as a convolutional neural network (CNN), a recurrent neural network (RNN), etc.

[0070] The above-mentioned training can be supervised training. Supervised training can be understood as using a set of data with known labels to train the model, and optimizing the model parameters so that the model can predict the labels of new data or make decisions based on the characteristics of existing data.

[0071] The above-mentioned parameter adjustment can be understood as a processing process for optimizing the performance of the model.

[0072] The above-mentioned trained football flight trajectory prediction model is trained through sample data in the historical kicking process, and can accurately predict the football flight trajectory according to the kicking characteristics when the football is kicked, such as kicking force, angle, wind direction and other information.

[0073] Optionally, the sample data in the historical kicking process includes the kicking data when the football is kicked, the flight data within a preset time, and the flight data after the preset time. The football flight trajectory prediction model includes a first feature extraction network, a second feature extraction network, a third feature extraction network, and a feature fusion network. In the step of training the pre-trained football flight trajectory prediction model through the sample data in the historical kicking process, the first feature extraction network can be used to perform first feature extraction on the kicking data when the football is kicked to obtain the kicking characteristics of the sample data; the second feature extraction network can be used to perform second feature extraction on the flight data within the preset time to obtain the first flight characteristics within the preset time; the third feature extraction network can be used to perform third feature extraction on the flight data after the preset time to obtain the second flight characteristics after the preset time; the feature fusion network can be used to perform fusion processing on the kicking characteristics, the first flight characteristics, and the second flight characteristics to obtain the flight trajectory characteristics of the sample data.

[0074] In the embodiment of the present invention, the kicking data when the above-mentioned football is kicked includes the kicking force, angle, direction, etc. The flight data within the preset time includes the initial velocity, initial angle, initial direction, etc. The flight data after the preset time includes information such as the football flight trajectory, landing point, flight time, etc.

[0075] The above-mentioned feature extraction network is responsible for extracting key features from the data; the above-mentioned feature fusion network is responsible for fusing the features output by the feature extraction network to generate a comprehensive feature vector, which contains the flight trajectory features of the football from being kicked, kicked out, and kicked out.

[0076] The above-mentioned feature extraction can be understood as a processing process of extracting useful information from the original data.

[0077] The above-mentioned kicking characteristics include characteristics such as kicking force, angle, speed, direction, etc.

[0078] The above-mentioned first flight characteristics include characteristics such as the speed, acceleration, rotation angle, etc. of the football kicked within the preset time.

[0079] The above-mentioned second flight characteristics include characteristics such as the flight trajectory, landing point position, flight time, etc. of the football after the preset time.

[0080] The above-mentioned preset time period is the time period when the sensor starts to record data. Specifically, it can be within a few seconds before and after the football is kicked, which can be within 1 second, 2 seconds, 3 seconds, etc. After the above-mentioned preset time is after the preset time point. For example, if the preset time is 1 second, then after the preset time can be 1 second later; if the preset time is 2 seconds, after the preset time can be 2 seconds later; if the preset time is 3 seconds, then after the preset time can be 3 seconds later, etc.

[0081] The above-mentioned feature fusion can be understood as a processing process of combining the extracted features to generate a comprehensive feature representation.

[0082] Furthermore, through the first feature extraction network, the kicking force, angle, and direction when the football is kicked are subjected to first feature extraction, and the force feature, angle feature, and direction feature of the kick when kicking the ball in the sample data can be obtained; then, through the second feature extraction network, the initial flight speed, initial angle, and initial direction within the preset time are subjected to feature extraction, and the speed feature, acceleration feature, and rotation angle feature of the football kicked within the preset time are obtained; then, through the third feature extraction network, the football flight trajectory, landing point, and flight time after the preset time are subjected to feature extraction, and the flight trajectory feature, landing point position feature, and flight time feature after the preset time are obtained; finally, through the feature fusion network, the features extracted in the above three stages are fused and processed to generate the flight trajectory feature of the sample data. The above fusion process can be a multi-modal fusion, that is, the features in different time periods and different types (such as when kicking the ball, the first flight, the second flight) are combined to form a comprehensive feature representation.

[0083] Optionally, the kicking data when the football is kicked includes the kicking force, kicking angle, and the rotation angle of the ball. In the step of performing first feature extraction on the kicking data when the football is kicked through the first feature extraction network to obtain the kicking features of the sample data, the first feature extraction network can perform force feature extraction on the kicking force to obtain the kicking force feature; perform angle feature extraction on the kicking angle through the first feature extraction network to obtain the kicking angle feature; perform rotation angle feature extraction on the rotation angle of the ball through the first feature extraction network to obtain the rotation angle feature; based on the kicking force feature, kicking angle feature, and rotation angle feature, obtain the kicking features of the sample data.

[0084] In this embodiment, the above-mentioned kicking force can be understood as the magnitude of the force exerted by the athlete when kicking the football; the above-mentioned kicking angle can be understood as the angle between the football and the ground when it is kicked, and the kicking angle affects the flight trajectory and the final landing point of the ball; the above-mentioned rotation angle of the ball can be understood as the angle at which the ball rotates around a certain axis during flight or movement.

[0085] Further, the first feature extraction network can perform force feature extraction processing on the magnitude of the force exerted by the athlete when kicking the football, and the magnitude feature of the force exerted by the athlete when kicking the football can be obtained; then, the first feature extraction network can perform angle feature extraction on the angle between the football and the ground when the football is kicked, and the angle feature between the football and the ground when the football is kicked can be obtained; then, the first feature extraction network can perform rotation angle feature extraction on the angle of rotation of the ball around a certain axis during flight or movement, so as to obtain the angle feature of the ball rotating around a certain axis during flight or movement; finally, by comprehensively processing the above three features, the kicking feature of the sample data can be obtained.

[0086] Optionally, the pre-trained football flight trajectory model further includes a trajectory feature extraction network. In the step of adjusting the parameters of the pre-trained football flight trajectory model through the corresponding football flight trajectory annotation data and completing the training to obtain the trained football flight trajectory prediction model, the trajectory feature extraction network can perform trajectory prediction on the flight trajectory feature of the sample data to obtain the predicted trajectory data of the sample data; calculate the loss value between the predicted trajectory data of the sample data and the corresponding football flight trajectory annotation data through a preset loss function, and take minimizing the loss value as the optimization target to adjust the model parameters of the pre-trained football flight trajectory prediction model. After the training is completed, the trained football flight trajectory prediction model is obtained.

[0087] In this embodiment, the above-mentioned trajectory feature extraction network can extract flight trajectory features from sample data.

[0088] The flight trajectory features of the above sample data include the flight trajectory features of the football from being kicked, when kicked, and after being kicked.

[0089] The above-mentioned trajectory prediction can be understood as a processing process of identifying and learning the laws or patterns of football flight from the flight trajectory features of sample data.

[0090] The predicted trajectory data of the above sample data contains information such as the predicted position, speed, and acceleration of the sample data after a preset time.

[0091] The above-mentioned preset loss function can be a cross-entropy loss function, a regression loss function, etc. The loss function is used to measure the difference between the predicted value and the actual value and is used to guide the training of the model.

[0092] Further, the model parameters of the pre-trained football flight trajectory prediction model can be adjusted through backpropagation, and the process of iterative parameter adjustment is carried out until the training of the pre-trained football flight trajectory prediction model is completed, and the trained football flight trajectory prediction model is obtained. The above-mentioned backpropagation is an efficient parameter update method, which updates the model parameters according to the gradient of the loss function.

[0093] Optionally, in the step of extracting features from the football motion data through the trained football flight trajectory prediction model to obtain the kicking features and initial flight features of the football motion data, the first feature extraction network can be used to perform first feature extraction on the football motion data to obtain the kicking features of the football motion data; the second feature extraction network can be used to perform second feature extraction on the football motion data to obtain the initial flight features of the football motion data.

[0094] In this embodiment, the above-mentioned football motion data may include motion data such as the speed, acceleration, rotation angle, and direction change of the football.

[0095] The above-mentioned first feature extraction can be understood as a processing process for extracting kicking features from football motion data. The second feature extraction can be understood as a processing process for extracting flight features within a preset time from football motion data.

[0096] The above-mentioned kicking features include force features, angle features, speed features, direction features, etc. of the kick.

[0097] The above-mentioned initial flight features include initial speed features, initial angle features, initial direction features, etc.

[0098] Optionally, in the step of outputting the football motion trajectory prediction result after a preset time based on the kicking features and initial flight features of the football motion data, the feature fusion network can be used to perform fusion processing on the kicking features and initial flight features of the football motion data to obtain the flight trajectory features of the football motion data; the trajectory feature extraction network can be used to perform trajectory prediction on the flight trajectory features of the football motion data and output the football motion trajectory prediction result after a preset time.

[0099] In this embodiment, the purpose of the above-mentioned feature fusion is to combine the kicking features and initial flight features together to generate the flight trajectory features of the football within a preset time.

[0100] The above-mentioned trajectory prediction can be understood as a processing process for calculating the flight trajectory of the football motion data after a preset time.

[0101] Furthermore, various parameters at the moment when the football is kicked and the initial flight features of the football can be fused through the feature fusion network to obtain the flight trajectory features of the football within a preset time; then, the trajectory feature extraction network can analyze and perform trajectory prediction on the flight trajectory features of the football within a preset time, so as to predict the position and motion state of the football after a preset time.

[0102] Optionally, after the step of outputting the predicted result of the football movement trajectory after a preset time based on the football kicking characteristics and the initial flight characteristics, a two-dimensional plane of the ground can be constructed; based on the predicted result of the football movement trajectory, the plane football landing point can be predicted in the two-dimensional plane; based on the mapping relationship between the ground and the two-dimensional plane, the ground football landing point corresponding to the plane football landing point can be determined.

[0103] In this embodiment, the above-mentioned ground football landing point is used to guide the player to move towards the ground football landing point in advance. The above-mentioned ground football landing point can help the player anticipate the landing point of the football in advance. The defensive player can adjust the position in advance to intercept the opponent's pass or shot; the offensive player can use the predicted football movement trajectory to select the best passing method or shooting opportunity.

[0104] The above-mentioned two-dimensional plane of the ground is used to simulate the movement of the football on the ground.

[0105] The above-mentioned mapping relationship between the ground and the two-dimensional plane is preset by the system, which can be understood as a relationship that maps the coordinates on the two-dimensional plane to the coordinates of the ground. The above-mentioned mapping relationship describes the conversion of the position or coordinates on the two-dimensional plane to the corresponding coordinates on the actual ground to ensure that the virtual representation in the two-dimensional plane can accurately reflect the actual ground situation. The mapping relationship between the ground and the two-dimensional plane ensures the accurate conversion from the virtual environment to the real world, so that the predicted result on the two-dimensional plane can accurately reflect on the actual ground.

[0106] Specifically, a two-dimensional plane model corresponding to the ground can be constructed, and then the predicted result of the football movement trajectory is mapped into the two-dimensional plane to predict the football landing point position after a preset time, and according to the mapping relationship between the ground and the two-dimensional plane, the football landing point position on the ground can be determined.

[0107] In the embodiment of the present invention, the present invention effectively helps the players and coaches on the field to make more accurate defensive and offensive strategies by constructing a two-dimensional plane of the ground, mapping the predicted result of the football movement trajectory into the two-dimensional plane, and mapping the plane football landing point predicted in the two-dimensional plane to the actual position on the ground according to the mapping relationship between the ground and the two-dimensional plane.

[0108] The present invention also proposes a football flight trajectory prediction system based on a motion sensor. The football flight trajectory prediction system based on a motion sensor includes: a server, and a motion sensor disposed in the football, and the processor is wirelessly connected to the motion sensor;

[0109] The motion sensor is disposed in the football, and the motion sensor is used to collect the motion data corresponding to the football and transmit the collected motion data to the server;

[0110] When the server executes, it performs the steps of the football flight trajectory prediction method based on a motion sensor provided in the embodiment of the present invention.

[0111] The above-mentioned football flight trajectory prediction system based on a motion sensor may further include: a display device for displaying the football motion trajectory prediction result or the predicted landing position on the ground. The above display device may be a wearable VR device. Of course, the above display device may also be only for the coach. The coach can timely notify the players to move through the football motion trajectory prediction result or the predicted landing position, improving the training efficiency of the players.

[0112] The present invention also proposes a football flight trajectory prediction device based on a motion sensor. Refer to Figure 2 , Figure 2 is a schematic structural diagram of a football flight trajectory prediction device based on a motion sensor in the hardware operating environment involved in the embodiment solution of the present invention.

[0113] The football flight trajectory prediction device based on a motion sensor in the embodiment of the present invention may be a computing device such as a desktop computer, a notebook, a palm computer, and a server. As shown in Figure 2 , the football flight trajectory prediction device based on a motion sensor may include: a processor 1001 (such as a CPU), a network interface 1004, a user interface 1003, a memory 1005, and a communication bus 1002. Among them, the communication bus 1002 is used to realize the connection communication between these components. The user interface 1003 may include a display screen (Display) and an input unit, such as a keyboard (Keyboard). Optionally, the user interface 1003 may further include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory or a stable memory (non-volatile memory), such as a disk memory. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0114] Those skilled in the art can understand that Figure 2 the structure of the football flight trajectory prediction device based on a motion sensor shown in

[0115] does not constitute a limitation on the football flight trajectory prediction device based on a motion sensor, and may include more or fewer components than shown in the figure, or combine some components, or have different component arrangements.

[0115] As shown in Figure 2 , the memory 1005, as a computer storage medium, may include an operating system, a network communication module, a user interface module, and a computer program.

[0116] InFigure 2 In the motion sensor-based football flight trajectory prediction device shown, the network interface 1004 is mainly used to connect to the background server and communicate data with the background server; the user interface 1003 is mainly used to connect to the client (user side) and communicate data with the client; and the processor 1001 can be used to call the computer program stored in the memory 1005. When the computer program is called and executed by the processor 1001, the steps of the above-mentioned motion sensor-based football flight trajectory prediction method are implemented.

[0117] Based on the computer program proposed in the foregoing embodiments, the present invention also proposes a storage medium that stores a computer program. When the computer program is executed by a controller, the motion sensor-based football flight trajectory prediction method described in the foregoing embodiments is implemented.

[0118] For the motion sensor-based football flight trajectory prediction device and storage medium of the present invention, since they can both implement the steps of the above-mentioned motion sensor-based football flight trajectory prediction method, they at least have all the beneficial effects brought by the technical solutions of the above-mentioned motion sensor-based football flight trajectory prediction method embodiments, which will not be elaborated here one by one.

[0119] In the several embodiments provided by the present application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods. For example, multiple modules or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or modules can be in electrical, mechanical or other forms.

[0120] The modules described as separate components may or may not be physically separated. The components displayed as modules may or may not be physical modules, that is, they can be located in one place, or they can be distributed to multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0121] In addition, in each embodiment of the present invention, the functional modules can be integrated in one processing module, or each module can exist physically alone, or two or more modules can be integrated in one module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software functional modules.

[0122] When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.

[0123] The above are only partial or preferred embodiments of the present invention. Neither the text nor the drawings can limit the scope of protection of the present invention. Any equivalent structural transformation made by using the content of the specification and drawings of the present invention under the overall concept of the present invention, or any direct / indirect application in other related technical fields, is included in the scope of protection of the present invention.

Claims

1. A method for predicting the flight trajectory of a football based on a motion sensor, characterized in that: The motion sensor is disposed in the football, and the motion sensor is used to collect motion data corresponding to the football. The method includes: Acquiring the running data of the football when the football is kicked out within a preset time through the motion sensor; Extracting features from the football motion data using a trained football flight trajectory prediction model to obtain kicking features and initial flight features of the football motion data, wherein the trained football flight trajectory prediction model is obtained by training a pre-trained football flight trajectory prediction model using a training data set; Based on the kicking characteristics of the football motion data and the initial flight characteristics, the football motion trajectory prediction result after a preset time is output.

2. The method for predicting a soccer ball's flight trajectory based on a motion sensor according to claim 1, characterized in that: Before extracting features from the football motion data using the trained football flight trajectory prediction model to obtain kicking features and initial flight features of the football motion data, the method further includes: Obtaining a training data set and a pre-trained football flight trajectory prediction model, wherein the training data set includes sample data from a historical football kicking process and corresponding football flight trajectory annotation data; The pre-trained football flight trajectory prediction model is trained by using the sample data from the historical kicking process, and the parameters of the pre-trained football flight trajectory prediction model are adjusted by using the corresponding football flight trajectory annotation data. After the training is completed, a trained football flight trajectory prediction model is obtained.

3. The method for predicting a soccer ball's flight trajectory based on a motion sensor according to claim 2, characterized in that: The sample data in the historical kicking process includes the kicking data when the football is kicked, the flight data within a preset time, and the flight data after the preset time. The football flight trajectory prediction model includes a first feature extraction network, a second feature extraction network, a third feature extraction network, and a feature fusion network. The pre-trained football flight trajectory prediction model is trained by the sample data in the historical extraction process, including: Performing a first feature extraction on the kicking data when the football is kicked by the first feature extraction network to obtain the kicking feature of the sample data; Performing second feature extraction on the flight data within the preset time by the second feature extraction network to obtain a first flight feature of the sample data; Performing a third feature extraction on the flight data after the preset time by the third feature extraction network to obtain a second flight feature of the sample data; The kicking feature, the first flight feature and the second flight feature are fused through the feature fusion network to obtain the flight trajectory feature of the sample data.

4. The method for predicting a soccer ball's flight trajectory based on a motion sensor according to claim 3, characterized in that: The kicking data of the football when it is kicked includes the kicking force, the kicking angle, and the rotation angle of the ball. The first feature extraction is performed on the kicking data of the football when it is kicked by the first vital sign extraction network to obtain the kicking features of the sample data, including: Extracting the force feature of the kicking force through the first feature extraction network to obtain the kicking force feature; Extracting angle features of the kicking angle through the first feature extraction network to obtain a kicking angle feature; Extracting rotation angle features of the rotation angle of the ball through the first feature extraction network to obtain rotation angle features; The kicking feature of the sample data is obtained based on the kicking force feature, the kicking angle feature and the rotation angle feature.

5. The method for predicting a soccer ball's flight trajectory based on a motion sensor according to claim 3, characterized in that: The pre-trained football flight trajectory model also includes a trajectory feature extraction network. The pre-trained football flight trajectory prediction model is parameter adjusted by the corresponding football flight trajectory annotation data. After the training is completed, a trained football flight trajectory prediction model is obtained, including: Performing trajectory prediction on the flight trajectory features of the sample data through the trajectory feature extraction network to obtain predicted trajectory data of the sample data; The loss value of the predicted trajectory data of the sample data and the corresponding football flight trajectory annotation data is calculated by a preset loss function, and the model parameters of the pre-trained football flight trajectory prediction model are adjusted with minimizing the loss value as the optimization goal. After the training is completed, a trained football flight trajectory prediction model is obtained.

6. The method for predicting a soccer ball flight trajectory based on a motion sensor according to any one of claims 1 to 5, characterized in that: The feature extraction of the football motion data by using the trained football flight trajectory prediction model to obtain the kicking features and initial flight features of the football motion data includes: Performing first feature extraction on the football motion data through the first feature extraction network to obtain a kicking feature of the football motion data; The second feature extraction network is used to perform second feature extraction on the football motion data to obtain initial flight features of the football motion data.

7. The method for predicting a soccer ball's flight trajectory based on a motion sensor according to claim 6, characterized in that: The step of outputting a prediction result of the football motion trajectory after a preset time based on the kicking feature and the initial flight feature of the football motion data includes: The kicking feature of the football motion data and the initial flight feature are fused through the feature fusion network to obtain the flight trajectory feature of the football motion data; The trajectory feature extraction network is used to perform trajectory prediction on the flight trajectory features of the football motion data, and outputs the football motion trajectory prediction result after a preset time.

8. The method for predicting a soccer ball flight trajectory based on a motion sensor according to any one of claims 1 to 7, characterized in that: After outputting the football motion trajectory prediction result after a preset time based on the kicking feature of the football motion data and the initial flight feature, the method further includes: Construct a two-dimensional plane of the ground; Based on the football motion trajectory prediction result, predicting the landing point of the plane football in the two-dimensional plane; Based on the mapping relationship between the ground and the two-dimensional plane, the ground football landing point corresponding to the plane football landing point is determined, and the ground football landing point is used to guide the player to move to the ground football landing point in advance.

9. A football flight trajectory prediction system, characterized in that: The soccer ball flight trajectory prediction system comprises: a server, and a motion sensor arranged in the soccer ball, wherein the processor is wirelessly connected to the motion sensor; The motion sensor is arranged in the football, and is used to collect motion data corresponding to the football, and transmit the collected motion data to the server; The server executes the steps of the method for predicting a football flight trajectory based on a motion sensor as described in any one of claims 1 to 8.

10. A soccer ball flight trajectory prediction device based on a motion sensor, characterized in that: The motion sensor-based football flight trajectory prediction device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of the motion sensor-based football flight trajectory prediction method as described in any one of claims 1 to 8 are implemented.

Images