Ball sports analysis method, device and system based on edge computing

Abstract

The present invention discloses a ball sports analysis method, device and system based on edge computing, which relates to the technical field of sports auxiliary equipment. The method includes: obtaining the position data of the target player and the video frame data of the sports field; obtaining the real-time data of the shot based on the position data; inputting the real-time data of the shot into the prediction model to obtain the expected landing point; using the image recognition technology based on the video frame data to obtain the actual landing point of the target ball; performing referee analysis based on the expected landing point and the actual landing point to obtain the referee result; the system includes a position acquisition device, a positioning base station device, a video acquisition device and an edge computing device for implementing the method. The present invention solves the problems of single function, low digitization and intelligence of sports auxiliary equipment in the prior art, and achieves the purpose of intelligent refereeing of ball sports in sports fields based on edge computing.

Description

Technical Field

[0001] The present invention relates to the technical field of sports assisting equipment, and in particular to a ball sports analysis method, device and system based on edge computing. Background Art

[0002] There are many different types of sports, among which ball games like basketball, football, badminton, and table tennis are particularly popular. Currently, traditional scoring and refereeing methods for these ball sports rely on manual card flipping and judging. Some scoring and refereeing methods now rely on manual entry of data into a system, which is then displayed on a screen, or manual review of video footage for judgment. However, these methods still rely heavily on the human factor. Subsequently, sports assistive devices such as electronic scoreboards and intelligent refereeing devices have gradually emerged, but these devices still suffer from limited functionality and a low level of digitalization and intelligence. Summary of the Invention

[0003] The main purpose of the present invention is to provide a ball sports analysis method, device and system based on edge computing, aiming to solve the technical problems in the existing technology that sports auxiliary equipment has single functions and low levels of digitization and intelligence.

[0004] To achieve the above object, the present invention adopts the following technical solutions:

[0005] In a first aspect, the present invention provides a ball motion analysis method based on edge computing, the method comprising:

[0006] Acquire position data of a target player and video frame data of a sports field; wherein the video frame data includes a target ball;

[0007] Obtaining real-time data of a shot according to the position data; wherein the real-time data of a shot includes the shot position data, the shot angle data, and the target player's shot acceleration data;

[0008] Inputting the real-time data of the ball hitting into a prediction model to obtain an estimated landing point of the target ball;

[0009] Obtaining the actual landing point of the target ball using image recognition technology based on the video frame data;

[0010] A referee analysis is performed based on the predicted ball landing point and the actual ball landing point to obtain a referee result.

[0011] Optionally, in the above-mentioned ball sports analysis method based on edge computing, after the step of obtaining the position data of the target player and the video frame data of the sports field, the method further includes:

[0012] According to the video frame data, using human figure recognition technology to obtain the moving object and its identity information;

[0013] If the identity information of the moving object is a non-target player, performing movement trajectory prediction to obtain a first predicted trajectory of the moving object;

[0014] determining whether the first predicted trajectory conflicts with the sports field;

[0015] If the first predicted trajectory conflicts with the sports field, the moving object is marked and a first alarm prompt is issued.

[0016] Optionally, in the above-mentioned ball motion analysis method based on edge computing, after the step of predicting the movement trajectory of the moving object if the identity information of the moving object is a non-target player and obtaining a first predicted trajectory of the moving object, the method further includes:

[0017] If the identity information of the moving object is a target player, performing movement trajectory prediction based on the position data of the target player to obtain a second predicted trajectory of the target player;

[0018] determining whether the first predicted trajectory conflicts with the second predicted trajectory;

[0019] If the first predicted trajectory conflicts with the second predicted trajectory, the moving object is marked and a second alarm is issued.

[0020] Optionally, in the above-mentioned ball sports analysis method based on edge computing, the position data includes UWB positioning data and nine-axis gyroscope data;

[0021] The step of obtaining real-time data of the shot according to the position data comprises:

[0022] Obtaining the target player's hitting acceleration data according to the nine-axis gyroscope data;

[0023] identifying a hitting action according to the hitting acceleration data, and obtaining a hitting time point of the target player;

[0024] Obtaining hitting position data of the target player according to the UWB positioning data and the nine-axis gyroscope data corresponding to the hitting time point;

[0025] According to the angle, specific force and magnetic force of the hitting position data in three-dimensional directions, the relative angle, hitting force and absolute angle of the hitting action are obtained to obtain the hitting angle data of the target player.

[0026] Optionally, in the above-mentioned ball motion analysis method based on edge computing, the step of obtaining the actual landing point of the target ball using image recognition technology according to the video frame data includes:

[0027] According to the video frame data, the target ball and the sports field are identified using image recognition technology to obtain the actual landing point of the target ball and the sideline of the sports field.

[0028] Optionally, in the above-mentioned ball sports analysis method based on edge computing, the step of performing referee analysis based on the expected ball landing point and the actual ball landing point to obtain a referee result includes:

[0029] Determining whether the expected ball landing point exceeds the sideline and whether the actual ball landing point exceeds the sideline;

[0030] If the predicted landing point is beyond the sideline, or the actual landing point is beyond the sideline, obtaining a referee result of failure to serve or hit the target ball;

[0031] If the predicted landing point does not exceed the sideline and the actual landing point does not exceed the sideline, a referee result of failure to catch the target ball is obtained.

[0032] Optionally, in the above-mentioned ball sports analysis method based on edge computing, after the step of performing referee analysis based on the expected ball landing point and the actual ball landing point to obtain a referee result, the method further includes:

[0033] Obtaining the score of the target player according to the referee result and pre-designed scoring rules;

[0034] The motion trajectory of the target player is drawn, and the motion trajectory, the real-time data of the shot and the referee result are stored.

[0035] In a second aspect, the present invention provides a ball sports analysis device based on edge computing, the device comprising:

[0036] A data acquisition module, configured to acquire position data of a target player and video frame data of a sports field; wherein the video frame data includes a target ball;

[0037] a shot recognition module, configured to obtain real-time shot data based on the position data; wherein the real-time shot data includes shot position data, shot angle data, and shot acceleration data of the target player;

[0038] A trajectory prediction module, configured to input the real-time data of the ball hitting into a prediction model to obtain an estimated landing point of the target ball;

[0039] An image recognition module, configured to obtain an actual landing point of the target ball using image recognition technology based on the video frame data;

[0040] The referee analysis module is used to perform referee analysis based on the expected landing point and the actual landing point to obtain a referee result.

[0041] In a third aspect, the present invention provides a ball sports analysis system based on edge computing, the system comprising:

[0042] Position acquisition device, used to send positioning signals in real time;

[0043] a positioning base station device for monitoring the positioning signal and generating position data of the target player;

[0044] A video acquisition device for acquiring video frame data of a sports field in real time; and

[0045] An edge computing device is used to implement the above-mentioned edge computing-based ball sports analysis method.

[0046] The above one or more technical solutions provided by the present invention may have the following advantages or at least achieve the following technical effects:

[0047] The present invention proposes a ball sports analysis method, device, and system based on edge computing. By obtaining real-time data on the shot based on the position data of the target player, the real-time data is input into a prediction model to obtain the expected landing point of the target ball. At the same time, based on the video frame data of the sports field, image recognition technology is used to obtain the actual landing point of the target ball. Finally, referee analysis is performed based on the expected landing point and the actual landing point to obtain the referee result, thus achieving the purpose of intelligent refereeing of ball sports in sports fields based on edge computing. The present invention integrates positioning recognition technology and video recognition technology to realize edge referee analysis in sports fields, improves the digitization and intelligence of edge computing devices in digital sports technology, reduces the amount of human participation, correspondingly reduces human errors, and improves the accuracy of sports analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in 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 only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these provided drawings without paying any creative work.

[0049] Figure 1 This is a flow chart of a first embodiment of a ball sports analysis method based on edge computing according to the present invention;

[0050] Figure 2 This is a schematic diagram of the functional modules of the first embodiment of the ball sports analysis device based on edge computing of the present invention;

[0051] Figure 3 This is a connection diagram of the first embodiment of the ball sports analysis system based on edge computing of the present invention;

[0052] Figure 4 for Figure 3 Schematic diagram of the hardware structure of the edge computing device.

[0053] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0054] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only 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 making any creative work shall fall within the scope of protection of the present invention.

[0055] It should be noted that, in the present invention, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "include..." does not exclude the presence of other identical elements in the process, method, article or system comprising the element. In addition, in the present invention, if there is a description involving "first", "second", etc., the description of "first", "second", etc. is only for descriptive purposes and cannot be understood as indicating or implying its relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" or "second" may explicitly or implicitly include at least one of such features.

[0056] In the present invention, the use of suffixes such as "module", "component" or "unit" to represent elements is only for the purpose of facilitating the description of the present invention and has no specific meaning in itself. Therefore, "module", "component" or "unit" can be used interchangeably. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances. In addition, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those of ordinary skill in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0057] In view of the technical problems of existing sports auxiliary equipment such as single functions and low digitalization and intelligence, the present invention provides a ball sports analysis method based on edge computing. The overall idea is as follows:

[0058] Obtain position data of a target player and video frame data of a sports field; wherein the video frame data includes a target ball; obtain real-time hitting data based on the position data; wherein the real-time hitting data includes hitting position data, hitting angle data, and hitting acceleration data of the target player; input the real-time hitting data into a prediction model to obtain an expected landing point of the target ball; obtain an actual landing point of the target ball using image recognition technology based on the video frame data; perform referee analysis based on the expected landing point and the actual landing point to obtain a referee result.

[0059] Through the above technical solution, the real-time data of the shot is obtained based on the position data of the target player, and then the real-time data of the shot is input into the prediction model to obtain the expected landing point of the target ball. At the same time, based on the video frame data of the sports field, the actual landing point of the target ball is obtained using image recognition technology. Finally, based on the expected landing point and the actual landing point, referee analysis is performed to obtain the referee result, achieving the purpose of intelligent refereeing of ball games in sports fields based on edge computing. The present invention integrates positioning recognition technology and video recognition technology to realize edge referee analysis in sports fields, improves the digitization and intelligence of edge computing devices in digital sports technology, reduces the amount of human participation, correspondingly reduces human errors, and improves the accuracy of sports analysis.

[0060] Example 1

[0061] Reference Figure 1The flowchart of the present invention proposes a first embodiment of the ball sports analysis method based on edge computing. The method can be applied to edge computing devices in digital sports venues such as sports fields and indoor sports halls. The device can achieve network connection with a remote or back-end cloud server. While the edge computing device provides proximal services for the sports venue, the cloud server can access the historical data of the edge computing device. The edge computing device can provide users with faster responses, solve needs at the edge, improve processing efficiency, and reduce the load on the cloud server.

[0062] Specifically, the edge computing device refers to a terminal device that can achieve network connection, which can be a computer, tablet computer, portable computer, embedded industrial computer and other terminal devices.

[0063] like Figure 4 FIG2 is a schematic diagram of the hardware structure of an edge computing device, which may include: a processor 1001, such as a CPU (Central Processing Unit), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005.

[0064] Those skilled in the art will understand that Figure 4 The hardware structure shown in the figure does not constitute a limitation on the edge computing device, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0065] Specifically, the communication bus 1002 is used to implement connection and communication between these components;

[0066] The user interface 1003 is used to connect to the client and perform data communication with the client. The user interface 1003 may include an output unit, such as a display screen, a speaker, etc., and an input unit, such as a keyboard, a microphone, etc.;

[0067] The network interface 1004 is used to connect to the backend server and perform data communication with the backend server. The network interface 1004 may include an input / output interface, such as a standard wired interface or a wireless interface, such as a Wi-Fi interface.

[0068] The memory 1005 is used to store various types of data. These data may include, for example, instructions for any application or method in the edge computing device, as well as application-related data. The memory 1005 may be a high-speed RAM memory or a stable memory such as a disk memory. Optionally, the memory 1005 may also be a storage device independent of the processor 1001.

[0069] For details, please refer to Figure 4, the memory 1005 may include an operating system, a network communication module, a user interface module and a computer program, wherein the network communication module is mainly used to connect to the cloud server and perform data communication with the cloud server;

[0070] The processor 1001 is configured to call the computer program stored in the memory 1005 and perform the following operations:

[0071] Acquire position data of a target player and video frame data of a sports field; wherein the video frame data includes a target ball;

[0072] Obtaining real-time data of a shot according to the position data; wherein the real-time data of a shot includes the shot position data, the shot angle data, and the target player's shot acceleration data;

[0073] Inputting the real-time data of the ball hitting into a prediction model to obtain an estimated landing point of the target ball;

[0074] Obtaining the actual landing point of the target ball using image recognition technology based on the video frame data;

[0075] A referee analysis is performed based on the predicted ball landing point and the actual ball landing point to obtain a referee result.

[0076] Based on the above edge computing device, the following Figure 1 The flowchart shown in the figure takes an indoor badminton stadium as an example to describe in detail the edge computing-based ball sports analysis method of this embodiment. The method may include the following steps:

[0077] Step S10: Acquire position data of a target player and video frame data of a sports field; wherein the video frame data includes a target ball.

[0078] Specifically, the location data includes UWB (Ultra Wide Band) positioning data and nine-axis gyroscope data. UWB positioning data includes two-dimensional position data, with an accuracy of 10cm, which can accurately obtain the target player's position data. The nine-axis gyroscope data includes three-dimensional acceleration data and three-dimensional position data. The nine-axis gyroscope can sense the horizontal, vertical, pitch, heading, and angular velocity of an object during movement, enabling accurate acquisition of the target player's three-dimensional posture data.

[0079] During implementation, location data can be acquired using a sports bracelet equipped with a UWB positioning sensor and a nine-axis gyroscope sensor, along with a corresponding UWB positioning base station. This data is then wirelessly transmitted to an edge computing device, which then acquires the target player's location data. The number of UWB positioning base stations is set to correspond to the size of the sports field, facilitating accurate location data acquisition. Video data can also be acquired using webcams installed at both ends of the field. This data is then wirelessly transmitted to an edge computing device, which then acquires video frame data of the field. Specifically, video of ball sports being played within the field is captured, specifically including the target player, target ball, and field. The video may also include aisles around the field, spectator seats, and so on. Location data and video frame data must be acquired synchronously, at the same interval, based on a minimum time interval.

[0080] In this embodiment, each badminton player wears a sports bracelet on the hand that hits the ball, and a high-definition network camera is hung above the outer periphery of the short sidelines at both ends of the badminton court to obtain the badminton player's position data and the badminton court video frame data.

[0081] Step S30: obtaining real-time data of the hitting ball according to the position data; wherein the real-time data of the hitting ball includes the hitting ball position data, the hitting ball angle data and the hitting ball acceleration data of the target player.

[0082] Specifically, step S30 may include:

[0083] Step S31: obtaining the target player's hitting acceleration data according to the nine-axis gyroscope data;

[0084] Step S32: identifying the hitting action according to the hitting acceleration data, and obtaining the hitting time point of the target player;

[0085] Step S33: obtaining the hitting position data of the target player according to the UWB positioning data and the nine-axis gyroscope data corresponding to the hitting time point;

[0086] Step S34: According to the angle, specific force and magnetic force of the hitting position data in the three-dimensional direction, the relative angle, hitting force and absolute angle of the hitting action are obtained to obtain the hitting angle data of the target player.

[0087] In the specific implementation process, the nine-axis gyroscope can directly collect the acceleration of the target player. When the acceleration reaches the maximum value, it is the moment when the target player's hand hits the ball. The action at this moment is the hitting action, and the time point corresponding to this moment is the hitting time point; then the hitting acceleration data (Acc_X, Acc_Y, Acc_Z) of the target player is directly obtained, and the plane position (Pos_X, Pos_Y) of the target player is obtained from the UWB positioning data, and the height position (Pos_X, Pos_Y) of the target player's hitter is obtained from the nine-axis gyroscope data. s_Z), thereby obtaining the target player's hitting position data (Pos_X, Pos_Y, Pos_Z); then, according to the angle corresponding to the position, obtain the relative angle Angle_X of the hitting action, according to the specific force corresponding to the position, obtain the hitting force Angle_Y of the hitting action, and according to the magnetic force corresponding to the position, obtain the absolute angle Angle_Z of the hitting action, thereby obtaining the target player's hitting angle data (Angle_X, Angle_Y, Angle_Z), that is, obtaining the target player's real-time hitting data.

[0088] Step S50: inputting the real-time data of the ball hitting into a prediction model to obtain the expected landing point of the target ball.

[0089] Specifically, the prediction model is a multi-input, multi-output neural network model that utilizes a deep learning algorithm. Prior to step S50, an initial model can be constructed and then trained using sample data or historical data to obtain a trained neural network model, i.e., the prediction model. Subsequently, during each practice session using the edge computing device, the model is continuously trained using real-time ball-hitting data to increase the model's prediction accuracy.

[0090] During the specific implementation process, the hitting position data (Pos_X, Pos_Y, Pos_Z), the hitting angle data (Angle_X, Angle_Y, Angle_Z) and the hitting acceleration data (Acc_X, Acc_Y, Acc_Z) are input into the prediction model together, and the expected landing point of the target ball (Ball_X, Ball_Y, Ball_Z) is output.

[0091] Step S70: Obtain the actual landing point of the target ball using image recognition technology based on the video frame data.

[0092] Specifically, step S70 may include:

[0093] Step S71: Based on the video frame data, the target ball and the sports field are identified using image recognition technology to obtain the actual landing point of the target ball and the sideline of the sports field.

[0094] Specifically, image recognition technology is used to identify the target ball and the edges of the playing field within the video frames transmitted by the camera. Based on multiple frames of real-time video data, when the target ball rebounds from the ground, its trajectory reverses. The reversal point is the actual landing point of the target ball.

[0095] At the same time, after each practice, the predicted landing point can be corrected according to the actual landing point, the prediction model can be updated to improve its prediction accuracy.

[0096] In this embodiment, based on the videos of the badminton court sent by two cameras, image recognition technology is used to identify the actual landing point of the badminton and the sidelines of the badminton court.

[0097] Step S90: performing referee analysis based on the predicted ball landing point and the actual ball landing point to obtain a referee result.

[0098] Specifically, step S90 may include:

[0099] Step S91: determining whether the predicted ball landing point exceeds the sideline, and whether the actual ball landing point exceeds the sideline;

[0100] Step S92: if the predicted landing point is beyond the sideline, or the actual landing point is beyond the sideline, obtaining a referee result of failure to serve or hit the target ball;

[0101] When a player serves, if the predicted landing point is outside the sideline, it means that the target player, i.e. the server, has made an error in serving. Regardless of whether the actual landing point obtained by the video frame data is inside or outside the sideline, the player's serve is judged to have failed. If the predicted landing point does not exceed the sideline, it means that the server has not made an error in serving. However, in actual competition, due to other factors, the predicted landing point may not exceed the sideline, but the actual landing point may still exceed the sideline. In this case, the player's serve is still judged to have failed.

[0102] After the server serves, if the receiving player fails to touch the ball, the target player remains the server. During this process, the receiving player fails to touch the ball, and the target player remains unchanged. If the actual ball landing point captured by the video frame data exceeds the sideline, it indicates that the server has made a serving error, and the serve can be directly determined as a failure. If the receiving player touches the ball, the target player becomes the receiving player, and the receiving action is considered a strike. At this point, if the server's original predicted ball landing point does not exceed the sideline, but the receiving player touches the ball while catching it, the target player has changed, and the process returns to step S10 to continue acquiring position data and video frame data. After the receiving action, the predicted and actual ball landing points are re-acquired accordingly. At this time, if the expected landing point is beyond the sideline, it means that the player of the receiving side has made a catching error. Regardless of whether the actual landing point obtained by the video frame data is inside or outside the sideline, the target player who receives the ball is judged to have failed to hit the ball. If the expected landing point does not exceed the sideline, but the actual landing point exceeds the sideline, it also means that the player of the receiving side has made a catching error, and the target player who receives the ball is judged to have failed to hit the ball.

[0103] Step S93: If the predicted landing point does not exceed the sideline and the actual landing point does not exceed the sideline, a referee result of failure to catch the target ball is obtained.

[0104] If the expected landing point does not exceed the sideline, it means that the shot or serve did not exceed the boundary, and the shot or serve did not make a mistake. Next, it is necessary to determine whether the receiving party has caught the ball based on the actual landing point obtained from the video frame data. If the receiving party accurately hits back the ball, it means that the receiving party has successfully caught the ball, and then returns to step S10 to continue to obtain position data and video frame data; if the receiving party fails to accurately hit back the ball, the ball falls directly into the sports field, that is, the actual landing point does not exceed the sideline, which means that the receiving party has failed to catch the ball.

[0105] The above steps enable the intelligent refereeing function of the edge computing device. This allows for digital and intelligent refereeing of ball sports, obtaining referee results, and displaying live video, providing more convenient and intelligent refereeing services.

[0106] In one embodiment, after step S10, the method may further include:

[0107] Step S21: obtaining the moving object and its identity information using human figure recognition technology according to the video frame data.

[0108] Human figure recognition is performed on the video frame data to obtain moving objects, that is, moving people, and identify the person's identity information, primarily including whether they are the target player. Identity information can be obtained in conjunction with position data. For example, when a person is identified, the corresponding position data is determined. If not, the person is not the target player. If the corresponding position data is present, the person is the target player, thereby obtaining the identity information of all moving objects in the video frame data. Identity information can also be obtained based on preset features. For example, if a moving object wearing a specific color or with a specific marking is pre-set as the target player, feature recognition can be performed on all moving objects in the video frame data to obtain their identity information.

[0109] Step S22: If the identity information of the moving object is a non-target player, a moving trajectory prediction is performed to obtain a first predicted trajectory of the moving object.

[0110] There are generally multiple moving objects in a video. When the identity information of a moving object is recognized as not being the target player, the trajectory of the moving object can be predicted. The technology for predicting the trajectory of moving objects in a video is existing technology and will not be described in detail here. The trajectory can also be predicted by having the moving object wear a UWB bracelet to obtain its position data, thereby obtaining a first predicted trajectory of the moving object.

[0111] Step S23: Determine whether the first predicted trajectory conflicts with the sports field.

[0112] When the camera shoots, the sports field will be completely captured in the video frame data. The user can directly mark the area of ​​the video frame data as the preset sports field area. Based on the preset sports field area in the video frame data, after obtaining the first predicted trajectory, it is determined whether the trajectory conflicts with the preset sports field area, that is, whether the first predicted trajectory will enter the preset sports field area. If so, it is determined that there is a conflict. If not, it is determined that there is no conflict.

[0113] Step S24: If the first predicted trajectory conflicts with the sports field, the moving object is marked and a first alarm is issued.

[0114] If the first predicted trajectory conflicts with the playing surface, it indicates that the moving object is about to or has already entered the playing surface, potentially impacting the players' performance and requiring prompt notification. Video frame data is typically broadcast live at the sports venue, where the moving object can be directly marked in the video. This can also trigger an alarm, such as announcing through a loudspeaker that a non-player has entered the playing surface, prompting the moving object to leave the playing surface immediately to avoid a collision.

[0115] Furthermore, after step S22, the method may further include:

[0116] Step S25: If the identity information of the moving object is a target player, a movement trajectory is predicted based on the position data of the target player to obtain a second predicted trajectory of the target player.

[0117] If the position of the moving object obtained using human figure recognition technology based on the video frame data overlaps with the previously acquired UWB positioning data, the identity information of the moving object indicates that it is the target player. Once the identity information of the moving object is identified as the target player, the trajectory of the target player is directly predicted based on the UWB positioning data to obtain a second predicted trajectory of the target player.

[0118] Step S26: Determine whether there is a conflict between the first predicted trajectory and the second predicted trajectory.

[0119] In a game, there are not only players but also referees and other identities. When step S22 obtains a moving object with the identity of a non-target player, a conflict judgment is performed between the first predicted trajectory corresponding to the moving object and the second predicted trajectory corresponding to the moving object with the identity of the target player obtained in step S25, that is, whether the two predicted trajectories have an intersection.

[0120] Step S27: If there is a conflict between the first predicted trajectory and the second predicted trajectory, the moving object is marked and a second alarm prompt is issued.

[0121] When two predicted trajectories intersect, that is, there is a conflict between the first predicted trajectory and the second predicted trajectory, it means that the two moving objects are likely to collide with each other. Timely warning can remind the moving objects to stop moving in time to prevent safety accidents.

[0122] The above steps enable the security monitoring function of the edge computing device. It monitors the safety of both players and non-players in the sports venue, issues timely alarms, and prevents safety accidents.

[0123] In another embodiment, after step S90, the method may further include:

[0124] Step S101: Obtain the score of the target player according to the referee result and pre-designed scoring rules.

[0125] Specifically, the scores can also be displayed on-site via a display screen to achieve intelligent scoring.

[0126] Step S102: drawing the motion trajectory of the target player, and storing the motion trajectory, the real-time data of the shot, and the referee result.

[0127] Specifically, the motion trajectory, real-time hitting data, and referee results are stored for subsequent analysis of the target player's scores, reasons for scoring or losing points, and analysis of the player's movement and movement in combination with video playback. The drawn motion trajectory, the real-time hitting data, and the referee results are stored on the edge computing device. The cloud server can directly use this data to obtain a data report, or obtain an analysis report based on some preset data analysis strategies to implement motion analysis of the target player, making it easier for users to understand the reasons why the target player failed to hit the ball, serve, or receive the ball, such as due to an inappropriate hitting position, inappropriate hitting force, or inappropriate hitting angle, etc.

[0128] The above steps enable the edge computing device to analyze the ball, making it easier to review the game process after the match and improve the players' skills.

[0129] The edge computing-based ball sports analysis method provided in this embodiment obtains real-time data on the shot based on the position data of the target player, then inputs the real-time data into the prediction model to obtain the expected landing point of the target ball. At the same time, based on the video frame data of the sports field, the image recognition technology is used to obtain the actual landing point of the target ball. Finally, based on the expected landing point and the actual landing point, referee analysis is performed to obtain the referee result, thereby achieving the purpose of intelligent refereeing of ball sports in the sports field based on edge computing. The present invention integrates positioning recognition technology and video recognition technology to realize edge referee analysis in the sports field, improves the digitization and intelligence of edge computing devices in digital sports technology, reduces the amount of human participation, correspondingly reduces human errors, and improves the accuracy of sports analysis. The edge computing-based ball sports analysis method provided in this embodiment can not only realize intelligent refereeing, but also realize safety monitoring and shot analysis.

[0130] Example 2

[0131] Based on the same invention concept, Figure 2 , the first embodiment of the ball sports analysis device based on edge computing of the present invention is proposed, and the device can be a virtual device. Figure 2 The functional module diagram shown in FIG. 1 is a detailed description of the ball sports analysis device based on edge computing provided in this embodiment. The device may include:

[0132] A data acquisition module, configured to acquire position data of a target player and video frame data of a sports field; wherein the video frame data includes a target ball;

[0133] a shot recognition module, configured to obtain real-time shot data based on the position data; wherein the real-time shot data includes shot position data, shot angle data, and shot acceleration data of the target player;

[0134] A trajectory prediction module, configured to input the real-time data of the ball hitting into a prediction model to obtain an estimated landing point of the target ball;

[0135] An image recognition module, configured to obtain an actual landing point of the target ball using image recognition technology based on the video frame data;

[0136] The referee analysis module is used to perform referee analysis based on the expected landing point and the actual landing point to obtain a referee result.

[0137] Furthermore, the device may further include a first alarm module, which may specifically include:

[0138] A moving object recognition unit, configured to obtain the moving object and its identity information based on the video frame data using human figure recognition technology;

[0139] a first trajectory prediction unit, configured to perform movement trajectory prediction to obtain a first predicted trajectory of the moving object if the identity information of the moving object is a non-target player;

[0140] a first conflict determination unit, configured to determine whether the first predicted trajectory conflicts with the sports field;

[0141] The first alarm prompt unit is used to mark the moving object and issue a first alarm prompt if there is a conflict between the first predicted trajectory and the sports field.

[0142] Furthermore, the device may further include a second alarm module, which may specifically include:

[0143] a second trajectory prediction unit, configured to, if the identity information of the moving object is a target player, perform movement trajectory prediction based on the position data of the target player to obtain a second predicted trajectory of the target player;

[0144] a second conflict determination unit, configured to determine whether there is a conflict between the first predicted trajectory and the second predicted trajectory;

[0145] The second alarm prompting unit is configured to mark the moving object and issue a second alarm prompt if there is a conflict between the first predicted trajectory and the second predicted trajectory.

[0146] Furthermore, the position data includes UWB positioning data and nine-axis gyroscope data; and the shot recognition module may include:

[0147] an acceleration data acquisition unit, configured to obtain the target player's hitting acceleration data based on the nine-axis gyroscope data;

[0148] a hitting time point acquisition unit, configured to identify the hitting action according to the hitting acceleration data and obtain the hitting time point of the target player;

[0149] a hitting position data acquisition unit, configured to obtain the hitting position data of the target player based on the UWB positioning data and the nine-axis gyroscope data corresponding to the hitting time point;

[0150] The hitting angle data acquisition unit is used to obtain the relative angle, hitting force and absolute angle of the hitting action according to the angle, specific force and magnetic force of the hitting position data in the three-dimensional direction, so as to obtain the hitting angle data of the target player.

[0151] Furthermore, the image recognition module is specifically used to:

[0152] According to the video frame data, the target ball and the sports field are identified using image recognition technology to obtain the actual landing point of the target ball and the sideline of the sports field.

[0153] Furthermore, the referee analysis module may include:

[0154] A referee analysis unit, configured to determine whether the predicted ball landing point exceeds the sideline, and whether the actual ball landing point exceeds the sideline;

[0155] a first referee result unit, configured to obtain a referee result of failure to serve or hit the target ball if the expected landing point exceeds the sideline or the actual landing point exceeds the sideline;

[0156] The second referee result unit is used to obtain a referee result of failure to catch the target ball if the expected landing point does not exceed the sideline and the actual landing point does not exceed the sideline.

[0157] Furthermore, the device may further include:

[0158] A score statistics module, configured to obtain the score of the target player according to the referee result and pre-designed scoring rules;

[0159] The player analysis module is used to draw the movement trajectory of the target player and store the movement trajectory, the real-time data of the shot and the referee result, so as to analyze the reasons for the score of the target player to score or lose points.

[0160] It should be noted that the functions that can be realized by each module in the edge computing-based ball sports analysis device provided in this embodiment and the corresponding technical effects achieved can refer to the description of the specific implementation methods in each embodiment of the edge computing-based ball sports analysis method of the present invention. For the sake of brevity of the specification, they will not be repeated here.

[0161] Example 3

[0162] Based on the same invention concept, Figure 3 and Figure 4 , the first embodiment of the ball sports analysis system based on edge computing of the present invention is proposed. Figure 3 The connection diagram shown in FIG. 1 is used to describe in detail the ball sports analysis system based on edge computing provided in this embodiment. The system may include:

[0163] Position acquisition device, used to send positioning signals in real time;

[0164] a positioning base station device for monitoring the positioning signal and generating position data of the target player;

[0165] A video acquisition device for acquiring video frame data of a sports field in real time; and

[0166] An edge computing device is used to implement the edge computing-based ball sports analysis method of the present invention.

[0167] Specifically, the position acquisition device can be a UWB positioning bracelet, which is worn on the player's hand or leg, and the wearing position can be selected according to the actual sports event.

[0168] The positioning base station device can be a UWB positioning base station, which is connected to the UWB positioning bracelet for communication, performs precise positioning, and generates location data to be sent to the edge computing device.

[0169] The video acquisition device can be two high-definition network cameras, which are installed at both ends of the sports field to ensure that the sports field is completely within the camera's shooting range. The two cameras send the video frame data they collect to the edge computing device.

[0170] The edge computing device refers to a terminal device that can achieve network connection, which can be a computer, tablet computer, portable computer, embedded industrial computer and other terminal devices. Figure 4 The figure shows a schematic diagram of the hardware structure of an edge computing device. The device may include a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, all or part of the steps of each embodiment of the ball motion analysis method based on edge computing of the present invention are implemented.

[0171] It can be understood that the device may further include a communication bus, a user interface and a network interface.

[0172] The communication bus is used to realize the connection and communication between these components.

[0173] The user interface is used to connect to the client and communicate data with the client. The user interface may include output units such as a display screen, a speaker, etc., and input units such as a keyboard, a microphone, etc.

[0174] The network interface is used to connect to the backend server and perform data communication with the backend server. The network interface may include an input / output interface, such as a standard wired interface, or a wireless interface, such as a Wi-Fi interface.

[0175] The memory is used to store various types of data, which may include, for example, instructions for any application or method in the edge computing device, as well as application-related data. The memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. Optionally, the memory can also be a storage device independent of the processor.

[0176] The processor is used to call the computer program stored in the memory and execute the ball motion analysis method based on edge computing as described above. The processor can be an application specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a controller, a microcontroller, a microprocessor or other electronic components, used to execute all or part of the steps of each embodiment of the ball motion analysis method based on edge computing as described above.

[0177] The edge computing-based ball sports analysis system provided in this embodiment realizes the purpose of intelligent refereeing of ball sports in sports venues based on sensor equipment and edge computing, as well as preventing sports safety injuries.

[0178] It should be noted that the serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0179] The above descriptions are merely optional embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by utilizing the contents of the present invention specification and drawings under the inventive concept of the present invention, or directly or indirectly applied in other related technical fields, are included in the patent protection scope of the present invention.

Claims

1. A ball sports analysis method based on edge computing, characterized in that: An edge computing device applied to ball sports events, the method comprising: Acquire position data of a target player and video frame data of a sports field; wherein the video frame data includes a target ball; Obtaining real-time data of a shot according to the position data; wherein the real-time data of a shot includes the shot position data, the shot angle data, and the target player's shot acceleration data; Inputting the real-time data of the ball hitting into a prediction model to obtain an estimated landing point of the target ball; Obtaining the actual landing point of the target ball using image recognition technology based on the video frame data; Performing referee analysis based on the predicted ball landing point and the actual ball landing point to obtain a referee result; The position data includes UWB positioning data and nine-axis gyroscope data; the step of obtaining real-time data of hitting the ball based on the position data includes: Obtaining the target player's hitting acceleration data according to the nine-axis gyroscope data; identifying a hitting action according to the hitting acceleration data, and obtaining a hitting time point of the target player; Obtaining hitting position data of the target player according to the UWB positioning data and the nine-axis gyroscope data corresponding to the hitting time point; According to the angle, specific force and magnetic force of the hitting position data in three-dimensional directions, the relative angle, hitting force and absolute angle of the hitting action are obtained to obtain the hitting angle data of the target player.

2. The ball sports analysis method based on edge computing according to claim 1, characterized in that: After the step of obtaining the position data of the target player and the video frame data of the sports field, the method further includes: According to the video frame data, using human figure recognition technology to obtain the moving object and its identity information; If the identity information of the moving object is a non-target player, performing movement trajectory prediction to obtain a first predicted trajectory of the moving object; determining whether the first predicted trajectory conflicts with the sports field; If the first predicted trajectory conflicts with the sports field, the moving object is marked and a first alarm prompt is issued.

3. The ball sports analysis method based on edge computing according to claim 2, characterized in that: After the step of predicting a movement trajectory and obtaining a first predicted trajectory of the moving object if the identity information of the moving object is a non-target player, the method further includes: If the identity information of the moving object is a target player, performing movement trajectory prediction based on the position data of the target player to obtain a second predicted trajectory of the target player; determining whether the first predicted trajectory conflicts with the second predicted trajectory; If the first predicted trajectory conflicts with the second predicted trajectory, the moving object is marked and a second alarm is issued.

4. The ball sports analysis method based on edge computing according to claim 1, characterized in that: The step of obtaining the actual landing point of the target ball by using image recognition technology based on the video frame data includes: According to the video frame data, the target ball and the sports field are identified using image recognition technology to obtain the actual landing point of the target ball and the sideline of the sports field.

5. The ball sports analysis method based on edge computing according to claim 4, characterized in that: The step of performing referee analysis based on the predicted ball landing point and the actual ball landing point to obtain a referee result includes: Determining whether the expected ball landing point exceeds the sideline and whether the actual ball landing point exceeds the sideline; If the predicted landing point is beyond the sideline, or the actual landing point is beyond the sideline, obtaining a referee result of failure to serve or hit the target ball; If the predicted landing point does not exceed the sideline and the actual landing point does not exceed the sideline, a referee result of failure to catch the target ball is obtained.

6. The ball sports analysis method based on edge computing according to any one of claims 1 to 5, characterized in that: After the step of performing referee analysis based on the predicted ball landing point and the actual ball landing point to obtain a referee result, the method further includes: Obtaining the score of the target player according to the referee result and pre-designed scoring rules; The motion trajectory of the target player is drawn, and the motion trajectory, the real-time data of the shot and the referee result are stored.

7. A ball sports analysis device based on edge computing, characterized in that: An edge computing device for ball sports analysis includes: A data acquisition module, configured to acquire position data of a target player and video frame data of a sports field; wherein the video frame data includes a target ball; a shot recognition module, configured to obtain real-time shot data based on the position data; wherein the real-time shot data includes shot position data, shot angle data, and shot acceleration data of the target player; A trajectory prediction module, configured to input the real-time data of the ball hitting into a prediction model to obtain an estimated landing point of the target ball; An image recognition module, configured to obtain an actual landing point of the target ball using image recognition technology based on the video frame data; A referee analysis module, configured to perform referee analysis based on the predicted ball landing point and the actual ball landing point to obtain a referee result; The position data includes UWB positioning data and nine-axis gyroscope data; the shot recognition module includes: an acceleration data acquisition unit, configured to obtain the target player's hitting acceleration data based on the nine-axis gyroscope data; a hitting time point acquisition unit, configured to identify the hitting action according to the hitting acceleration data and obtain the hitting time point of the target player; a hitting position data acquisition unit, configured to obtain the hitting position data of the target player based on the UWB positioning data and the nine-axis gyroscope data corresponding to the hitting time point; The hitting angle data acquisition unit is used to obtain the relative angle, hitting force and absolute angle of the hitting action according to the angle, specific force and magnetic force of the hitting position data in the three-dimensional direction, so as to obtain the hitting angle data of the target player.

8. A ball sports analysis system based on edge computing, characterized in that: The system comprises: Position acquisition device, used to send positioning signals in real time; a positioning base station device for monitoring the positioning signal and generating position data of the target player; A video acquisition device for acquiring video frame data of a sports field in real time; and An edge computing device for implementing the ball sports analysis method based on edge computing as described in any one of claims 1 to 6.

Images