Ball sports training method and ball sports training system

Abstract

A ball sports training method and a ball sports training system. The method comprises: during the process of a training object engaging in ball sports, acquiring a plurality of sports scene images collected by means of an image collection apparatus, and a collection time of each sports scene image; on the basis of the plurality of sports scene images and the collection time of each sports scene image, determining hitting response information of the training object, wherein the hitting response information comprises at least one of hitting implementation information, related-hitting-position information, hitting response time information and hitting speed information; and on the basis of the hitting response information, determining training assessment information for the training object, wherein the training assessment information comprises assessment information about the sports performance of the training object and / or assessment information about the physical health status of the training object. Therefore, the objectivity, accuracy, high efficiency and professionalism of ball sports training are improved, and the labor costs are reduced; and when the solution is applied to rehabilitation training, the effect of rehabilitation training can be significantly improved.

Description

Ball sports training method and ball sports training system

[0001] This application claims priority to U.S. patent application No. 63 / 607,249 filed in the U.S. Patent Office on December 7, 2023, and claims priority to U.S. patent application No. 63 / 682,709 filed in the U.S. Patent Office on August 13, 2024, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The present application relates to the field of sports training technology, and in particular to a ball sports training method and a ball sports training system. Background Art

[0003] Ball sports, especially table sports (e.g., table tennis), offer numerous advantages, including ease of play, high participation, enjoyment, and safety, making them suitable for people of all ages and physiques. Scientific ball sports training can improve a person's physical function. However, existing ball sports training often relies on the coach's personal experience, resulting in high labor costs and a lack of objective training programs. Technical issues

[0004] The purpose of the embodiments of the present application is to provide a ball sports training method, a ball sports training system, a ball sports training control device, a terminal device, a computer-readable storage medium and a computer program product. Technical Solutions

[0005] The technical solution adopted in the embodiment of this application is:

[0006] In a first aspect, a ball sports training method is provided, comprising: acquiring, while a training subject is performing ball sports, a plurality of motion scene images acquired by an image acquisition device and the acquisition time of each motion scene image; determining the training subject's hitting reaction information based on the plurality of motion scene images and the acquisition time of each motion scene image, wherein the hitting reaction information includes at least one of whether the ball is hit information, hitting-related position information, hitting reaction time information, and hitting speed information; the whether the ball is hit information includes first information indicating that the training subject has hit the ball or second information indicating that the training subject has not hit the ball; the hitting-related position information includes hitting position information and / or hitting end position information; the hitting position information is position information of the ball when the training subject hits the ball, and the hitting end position information is position information of the destination of the ball after the training subject hits the ball; and determining training evaluation information for the training subject based on the hitting reaction information, wherein the training evaluation information includes evaluation information of the training subject's sports performance and / or evaluation information of the training subject's physical health status.

[0007] In a second aspect, a ball sports training system is provided, comprising: an image acquisition device for acquiring sports scene images at a preset frequency while a training subject is performing ball sports; and a control device for executing the above-mentioned ball sports training method.

[0008] According to a third aspect, a ball sports training control device is provided, comprising: an acquisition module for acquiring, during a training subject's ball sports, a plurality of motion scene images acquired by an image acquisition device and the acquisition time of each motion scene image; a determination module for determining, based on the plurality of motion scene images and the acquisition time of each motion scene image, the hitting reaction information of the training subject, wherein the hitting reaction information includes at least one of whether the training subject hits the ball, hitting-related position information, hitting reaction time information, and hitting speed information; the whether the hitting information includes first information indicating that the training subject hits the ball or second information indicating that the training subject does not hit the ball; the hitting-related position information includes hitting position information and / or hitting end position information; the hitting position information is position information of the ball when the training subject hits the ball; and the hitting end position information is position information of the destination of the ball after the training subject hits the ball; and an evaluation module for determining, based on the hitting reaction information, training evaluation information of the training subject, wherein the training evaluation information includes evaluation information of the training subject's sports performance and / or evaluation information of the training subject's physical health status.

[0009] In a fourth aspect, a terminal device is provided, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned ball sports training method when executing the computer program.

[0010] In a fifth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned ball sports training method are implemented.

[0011] In a sixth aspect, a computer program product is provided, which, when executed on a terminal device, enables the terminal device to implement the steps in the above-mentioned ball sports training method.

[0012] The beneficial effect of the ball sports training method provided by the embodiments of the present application is that, while a training subject is performing a ball sport, the training subject's hitting reaction information is automatically, quickly, and accurately determined based on multiple motion scene images captured by an image acquisition device and the acquisition time of each motion scene image. Ultimately, based on the hitting reaction information, accurate assessment information of the training subject's athletic performance and / or physical health status is determined. This solution automatically evaluates the training subject through computer vision and data analysis, reducing errors and personal bias introduced by subjective judgment in manual measurement and recording, thereby greatly improving the objectivity, accuracy, and scientific nature of the training evaluation. Furthermore, this solution focuses on the training subject's hitting reaction information, making this sports training solution applicable not only to sports training for healthy people, but also to rehabilitation training for patients with cognitive and motor health issues. On the one hand, applying ball sports training to rehabilitation training can make rehabilitation training more interesting, thereby increasing patients' training enthusiasm and significantly improving the effectiveness of rehabilitation training. Furthermore, by using images captured by computer vision to determine the training subject's batting reaction information and using this information to promptly assess the training subject's limb motor ability or health status, the assessment results can be more accurate, objective, targeted, and real-time. In summary, this solution not only improves the objectivity, accuracy, efficiency, and professionalism of ball sports training, while reducing labor costs, but also significantly enhances the effectiveness of rehabilitation training when applied to it. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or exemplary technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0014] FIG1 is a schematic diagram of a flow chart of a ball sports training method provided in one embodiment of the present application;

[0015] FIG2 is a schematic diagram of a ball sports training scene provided by an embodiment of the present application;

[0016] FIG3a is a schematic diagram of a ball sports training scene in a dynamic serving mode provided by an embodiment of the present application;

[0017] FIG3 b is a schematic diagram of a ball sports training scene in a fixed-point serving mode provided by an embodiment of the present application;

[0018] FIG4 is a schematic diagram of key point detection results provided in an embodiment of the present application;

[0019] FIG5 is a schematic structural diagram of a ball sports training system provided in one embodiment of the present application;

[0020] FIG6 is a schematic structural diagram of a ball sports training control device provided in one embodiment of the present application;

[0021] FIG7 is a schematic diagram of the structure of a terminal device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0023] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0024] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0025] It will also be understood that the term "and / or" used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0026] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0027] References to "one embodiment" or "some embodiments" described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the sentences "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways. In order to illustrate the technical solutions provided by this application, the following detailed description is given in conjunction with specific drawings and embodiments.

[0028] As mentioned above, existing ball sports training programs mostly rely on the coach's personal experience, which not only incurs high labor costs but also lacks objectivity. Furthermore, during training, the trainee's performance needs to be constantly evaluated. However, these evaluations often rely on personal observation and recording, which can lead to a lack of objectivity and inaccuracy in the results. This, in turn, can lead to poor training results.

[0029] In order to at least partially solve the above technical problems, the embodiments of the present application provide a ball sports training method, a ball sports training system, a ball sports training control device, a terminal device, a computer-readable storage medium and a computer program product.

[0030] In a first aspect, a ball sports training method is provided. The ball sports training method uses computer vision methods to automatically, efficiently and accurately evaluate the limb motor ability and physical health status of the training subject. The ball sports training method can be applied to the field of conventional sports training, such as for sports training for healthy elderly people or children to train their motor skills and hand-eye coordination. The ball sports training method can also be used in the field of health technology, especially the field of rehabilitation training technology. For example, the ball sports training method can be applied to rehabilitation treatment for patients such as those suffering from stroke, which can improve the patient's cognitive level and motor level and promote the rehabilitation of their neurocognitive ability and neuromotor ability.

[0031] First, it should be noted that the ball sports in the embodiments of the present application can be a variety of suitable ball sports, including but not limited to tennis, badminton, table tennis, billiards, etc., and in particular, ball sports that can be played on a table, including but not limited to table tennis, billiards, table football, table curling, etc. For simplicity, the following description uses table tennis as an example. The training subjects in the embodiments of the present application can be subjects receiving ball sports training, and the training subjects can be healthy people or patients with certain physical diseases.

[0032] The ball sports training method of the present invention can be executed by any suitable computing device. For example, it can be executed by a processor of a terminal device (such as a mobile phone, computer, tablet computer, laptop computer, electronic watch, etc.). Alternatively, it can be executed by a processor in a server.

[0033] As shown in FIG1 , the ball sports training method provided in the embodiment of the present application includes the following steps S110 , S120 and S130 .

[0034] Step S110: While the training subject is playing a ball game, multiple motion scene images captured by an image capture device and the capture time of each motion scene image are acquired. The visual range of the image capture device may include the entire range of motion of the ball. For example, for table ball sports, the visual range of the image capture device may include at least the surface area of ​​the table where the ball is located.

[0035] In the embodiments of the present application, the image acquisition device can have various forms, as long as it can capture images in a moving scene. For example, the image acquisition device can be a high-speed camera that can capture images at a high frame rate. The number of image acquisition devices can also be arbitrary. One image acquisition device can be set up in the moving scene, or multiple image acquisition devices can be set up. The moving scene image in the embodiments of the present application can be generated by images captured by a single image acquisition device, or by images captured by multiple image acquisition devices.

[0036] In one example, a wide-angle camera can be set up above the table, and the camera can observe the entire sports scene area from a suitable viewing angle. For example, the camera's visual axis can be at a preset angle to the surface normal vector of the table. The preset angle can be less than or equal to the angle threshold. For example, the preset angle is equal to 0, and the camera's visual axis is parallel to the surface normal vector of the table. That is, the camera can shoot a bird's-eye view of the entire table area. When the table tennis ball rolls on the table surface, the camera can accurately observe the position of the table tennis ball on the table surface. For another example, the preset angle is greater than 0. The camera's field of view can include the table surface area and the training subject's body area (such as the upper limb activity area). In this example, the camera can observe both the movement trajectory of the table tennis ball above the table and the training subject's limb activities.

[0037] In another example, multiple image acquisition devices can be set up at different locations around the table, and each of these devices can be used to capture images of the motion scene. The images captured by the different image acquisition devices at the same moment can then be stitched together to produce a motion scene image. In this example, the combined field of view of the multiple image acquisition devices can include the surface area of ​​the table, so the resulting stitched motion scene image can include the entire surface area of ​​the table. For example, at least one of the multiple image acquisition devices can be oriented toward the upper body of the training subject, so that the motion scene image can also include the active area of ​​the subject's upper limbs.

[0038] In the embodiment of the present application, the image acquisition device can be a two-dimensional image acquisition device or a three-dimensional image acquisition device. The motion scene image obtained in this step can be a two-dimensional image or a three-dimensional image.

[0039] With reference to Figure 2, the table tennis scene shown in the figure may include a table tennis table, an image acquisition device and a terminal device. The terminal device may include a display, a host, and an input device (e.g., a keyboard, a mouse). The host may include a memory and a processor. Exemplarily, an application program named "Training Evaluation" may be stored in the memory. The application program may implement the steps of the ball sports training method of the embodiment of the present application when running. For example, before the start of training, the "Training Evaluation" application program may be run on the processor in response to the operation of the input device by the user (e.g., a sports coach or a rehabilitation therapist). For example, the rehabilitation therapist clicks the shortcut of the "Training Evaluation" application program on the display desktop with a mouse, and then the display may display the program interface.

[0040] The terminal device's host computer can be connected to the image acquisition device. The program interface includes a device calibration control. When the user clicks the device calibration control, a device calibration window is displayed. The device calibration window displays the image within the image acquisition device's current field of view. The device calibration window also includes manual and automatic calibration controls. The user can click the manual or automatic calibration controls and follow the prompts in the window to calibrate the image acquisition device. After calibration is complete, training can begin.

[0041] For example, a rehabilitation therapist can first inform the training subject of the rules and safety precautions to be observed during exercise. Once the training subject is ready, training can begin. An operation control such as "Start Training" can be displayed in the program interface window, and the rehabilitation therapist can click this operation control to start training. During the training subject's table tennis training, the image acquisition device can capture images of the motion scene at a preset frequency and transmit the captured images to the terminal device in real time. The terminal device can directly use the received images as motion scene images, or further process the received images to obtain motion scene images. For example, the terminal device's processor can splice images captured by different image acquisition devices to obtain motion scene images. For example, the motion scene images can be displayed in real time on the program interface displayed on the display to facilitate observation by the rehabilitation therapist. The terminal device can also record and store the acquisition time of these motion scene images. For example, the timestamp of each frame in the continuous video frames transmitted by the camera can be associated with the motion scene image obtained after processing. It can be understood that after the training subject exercises for a period of time (for example, after completing a set of training tasks), multiple motion scene images and the timestamp of each motion scene image can be obtained, that is, a time series of motion scene images can be obtained.

[0042] In some embodiments, the image acquisition device may also have a network transmission function. For example, the image acquisition device may be a webcam. The webcam may communicate with a server in real time via a wide area network or a local area network, and may send the real-time acquired images to the server and store them in the server. The server may communicate with at least one terminal device (such as a desktop computer, mobile phone, tablet computer, etc.) via a network. For example, the terminal device may be installed with the aforementioned "training evaluation" application. During table tennis training, the terminal device, while the application is running, may obtain images captured by the image acquisition device in real time through the server to obtain images of the sports scene.

[0043] Step S120 : determining the training subject's hitting reaction information based on the multiple sports scene images and the acquisition time of each sports scene image.

[0044] The shot reaction information includes at least one of: shot-hit information, shot-related position information, shot reaction time information, and shot speed information. The shot-hit information includes: first information indicating that the training subject has hit the ball or second information indicating that the training subject has not hit the ball. The shot-related position information includes shot position information and / or shot destination position information. The shot position information is the position of the ball when the training subject hits the ball. The shot destination position information is the position of the ball after the training subject hits the ball.

[0045] In the embodiment of the present application, the training subject's hitting may include receiving the ball and serving the ball, and may include the training subject's hitting action of touching the table tennis ball with a racket.

[0046] Exemplarily, the batting reaction information may include information on whether the ball was batted. The information on whether the ball was batted may include first information and second information. The first information indicates that the training subject has performed a batting action, and the second information indicates that the training subject has not performed a batting action. Exemplarily, the first information may also be subdivided. For example, the first information may be divided into first touch information and second touch information. The first touch information indicates that the training subject has performed a force batting action and successfully touched the ping-pong ball, and the second touch information indicates that the training subject has performed a force batting action but has not touched the ping-pong ball. In an embodiment of the present application, before starting training, the training subject has been informed of the batting rules. For example, the batting rules may include situations in which batting can be performed and situations in which batting cannot be performed. During ball sports training, the training subject needs to make batting decisions based on these batting rules, and then react to whether to bat.

[0047] For example, when the training subject performs a batting action, the batting reaction information may also include batting-related position information. The batting-related position information may include the position information of the ball when the training subject bats and / or after the batting. Specifically, it may include batting position information and / or batting end position information. The batting position information is the position information of the ball when the training subject bats, that is, the position information of the ball when the training subject's racket touches the ping-pong ball. The batting end position information is the position information of the destination of the ball after the training subject bats. Specifically, it may be the position information of the end position where the ball changes its direction of motion and finally falls back on the table after the training subject bats. The position information in the batting-related position information may be absolute position information. For example, it may be the coordinates of the batting position in a physical coordinate system. The position information in the batting-related position information may also be relative position information. For example, the batting position information is the relative position of the batting position relative to a reference point on the table. For example, a two-dimensional rectangular coordinate system can be established with the center of the table as the origin, the short side of the table surface as the x-axis, and the long side of the table surface as the y-axis, thereby determining the two-dimensional coordinates of the hitting position. For example, the hitting rules taught to the training subject may also include at least one of the following: areas where the ball can be hit, areas where it cannot be hit, and correct hitting endpoints corresponding to different scenarios.

[0048] For example, when the training subject performs a batting action, the batting reaction information may also include the training subject's batting speed information. It is understood that the training subject's racket contacting the ping-pong ball may cause the ball's direction of motion to change, and the batting speed information may be the instantaneous speed of the ping-pong ball after the change in direction. The batting reaction information may also include the training subject's batting reaction time. The batting reaction time may be the duration from the moment the training subject becomes capable of batting to the moment the racket contacts the ball.

[0049] Of course, the hitting reaction information may also include other information that can objectively characterize the hitting reaction of the training subject, for example, the complete movement trajectory of the table tennis ball after it changes its direction of movement may be determined.

[0050] In an embodiment of the present application, the batting reaction information of the training subject can be determined based on multiple motion scene images acquired within a period of time and the acquisition time of each motion scene image. The period of time can be set according to actual needs. The period of time can at least include the time required for a complete batting round. A batting round can be determined based on the actual training task. For example, the training task may include a serving training task and a receiving training task. In the example where the sports training task is a serving training task, a batting round may include the entire process of preparing for the serve, serving, and the ball reaching the end point. In the example where the sports training task is a receiving training task, a batting round may include the entire process of preparing for the serve, serving, receiving, and the ball reaching the end point.

[0051] In one example, the period of time can be manually set. For example, a rehabilitation therapist can manually set the start and end acquisition times. The period of time can be the duration between the start and end acquisition times. The processor of the terminal device can automatically analyze the multiple motion scene images acquired during the period of time and the acquisition time of each motion scene image, and automatically divide the period of time into round times corresponding to multiple batting rounds. The motion scene images and their timestamps within each round time can then be analyzed to obtain the training subject's batting reaction information for each batting round.

[0052] In another example, the time period can also be set automatically. For example, a table tennis training scenario includes an automatic serving device. The automatic serving device can fire ping-pong balls at a preset frequency, and the training subject can practice receiving the balls. The terminal device can connect to the automatic serving device and obtain the time of each serve. This can then automatically determine the duration between two consecutive serves, thereby obtaining the duration of each batting round. In this example, the terminal device's processor can analyze the time series of the sports scene images during each batting round to obtain the training subject's batting reaction information for each batting round.

[0053] In embodiments of the present application, various suitable methods can be employed to determine the subject's batting reaction information during each batting round based on the motion scene images and the acquisition time of each motion scene image. In one example, a machine learning approach can be employed to determine the batting reaction information. For example, the time series of motion scene images during the current batting round can be input into a trained deep learning model, which outputs the subject's batting reaction information during the current batting round. A certain number of training datasets can be pre-collected and used to train the deep learning model, resulting in a high computational accuracy. In another example, image processing and statistical analysis methods can be employed to determine the batting reaction information. For example, the motion scene image is a depth image containing color and depth information. The difference between the color of the ping-pong ball and the color of the table surface is greater than a difference threshold. For example, the table surface is blue, and the ping-pong ball is red, orange, green, yellow, purple, etc. For each motion scene image, a threshold segmentation approach can be employed to segment the imaged area of ​​the table surface and the imaged area of ​​the ping-pong ball. The imaged area of ​​the table surface can be referred to as the table surface area, and the imaged area of ​​the ping-pong ball can be referred to as the ping-pong ball area. A two-dimensional rectangular coordinate system can be established with the center pixel of the table tennis area as the origin, and the coordinates of the center pixel of the table tennis area (e.g., the center pixel) can be calculated. By analogy, the two-dimensional coordinates of the table tennis ball at each moment in the current stroke can be obtained. Then, the movement trajectory of the table tennis ball can be fitted based on the depth information of the center pixel of the table tennis area at each moment. Based on the fitted movement trajectory, various suitable analysis methods can be further used to determine the training subject's shot reaction information. For example, based on the movement direction of the table tennis ball at each moment, the coordinates of the center pixel when the movement direction of the table tennis ball changes from the positive direction to the negative direction can be used as the shot position information.

[0054] Step S130 , determining training evaluation information for the training subject based on the hitting reaction information, wherein the training evaluation information includes evaluation information of the training subject's athletic performance and / or evaluation information of the training subject's physical health status.

[0055] In one example, the evaluation information of the training subject's athletic performance can be determined based on the training subject's batting reaction information in each batting round. For example, for a healthy training subject, the training subject's athletic performance in the batting round can be scored based on the batting position information, batting speed information, and batting end position information, and the level of table tennis performance can be determined based on the score. The division of these levels can be arbitrarily set according to actual needs. For example, the level of table tennis performance can be divided into five categories from low to high: "invalid", "failed", "qualified", "good", and "excellent". Different levels can correspond to different score ranges.

[0056] In another example, the assessment information of the physical health status of the training subject can be determined based on the training subject's batting reaction information in each batting round. For example, for patients who may have cognitive impairment or weak cognitive ability (such as stroke patients), their cognitive ability and motor ability can be evaluated separately based on the patient's batting reaction information. For example, their cognitive ability can be scored based on whether the ball is hit, the batting position information, and the batting end position information. Their athletic performance can be scored based on the batting speed information. The severity of the training subject's stroke or the degree of improvement of rehabilitation training can be determined based on the scoring results. For another example, for some children with cerebral palsy, their hand-eye coordination ability can also be scored based on their batting reaction information in each batting round, and the severity of their condition can be determined based on the scoring results, and then a rehabilitation training plan suitable for the child can be recommended.

[0057] In one example, training evaluation information for a training subject can be determined solely based on the training subject's hitting reaction information. In another example, training evaluation information for the training subject can also be determined based on the training subject's hitting reaction information and combined with first evaluation reference information for the training subject. The first evaluation reference information may include: biomechanical data, physiological data, and psychological data of the training subject monitored at multiple moments during the training subject's exercise, as well as mechanical data of the sports equipment (e.g., a table tennis ball, a racket, etc.) used or operated by the training subject. For example, a ball sports training system may include multiple information acquisition devices (e.g., multiple sensors, multiple data monitoring and processing devices), which can be used to collect and obtain various training data, including: motion scene images collected using image sensors, biomechanical data, physiological data, and psychological data of the training subject, as well as mechanical data of the sports equipment used or operated by the training subject. After the training subject's hitting reaction information is determined using the method of step S120, the determined hitting reaction information (e.g., referred to as the second evaluation reference information) and other training data (the first evaluation reference information) can be combined to accurately determine the training evaluation information for the training subject. The accuracy of training evaluation can be further improved.

[0058] As previously mentioned, existing ball sports training programs mostly rely on the coach's personal experience, which not only incurs high labor costs but also lacks objectivity. However, according to the ball sports training method described in the embodiments of the present application, while a subject is playing a ball sport, the subject's hitting reaction information is automatically, quickly, and accurately determined based on multiple motion scene images captured by an image acquisition device and the acquisition time of each motion scene image. Ultimately, based on the hitting reaction information, accurate assessment information of the subject's athletic performance and / or physical health status is determined. This solution automatically evaluates the subject through computer vision and data analysis, reducing errors and personal bias introduced by subjective judgment in manual measurement and recording, thereby significantly improving the objectivity, accuracy, and scientific nature of training evaluations. Furthermore, this solution focuses on the subject's hitting reaction information, making this sports training program applicable not only to exercise training for healthy individuals but also to rehabilitation training for patients with cognitive and motor health issues. On the one hand, applying ball sports training to rehabilitation training can make rehabilitation training more interesting, thereby increasing patients' training enthusiasm and significantly improving rehabilitation training effectiveness. Furthermore, by using images captured by computer vision to determine the training subject's batting reaction information and using this information to promptly assess the training subject's limb motor ability or health status, the assessment results can be more accurate, objective, targeted, and real-time. In summary, this solution not only improves the objectivity, accuracy, efficiency, and professionalism of ball sports training, while reducing labor costs, but also significantly enhances the effectiveness of rehabilitation training when applied to it.

[0059] In one embodiment, the ball sport is table tennis, and the visual range of the image capture device includes the surface area of ​​the table tennis table where the ball is located. The shot reaction information includes shot-related position information. Before determining the training evaluation information for the training subject in step S130, the ball sport training method provided in this embodiment of the application further includes step S101.

[0060] Step S101 : performing region segmentation on the motion scene image to determine a scoring reference region in the motion scene image, wherein the scoring reference region is located in the ball and table region, which is the imaging region of the ball and table.

[0061] In an embodiment of the present application, the scoring reference area can be set according to actual needs. For example, the ball table area can be divided into a receiving area and a restricted area. The restricted area can indicate an area where receiving the ball is prohibited. For another example, the ball table area can be divided into an end area and a non-end area. The end area can be a valid area where the ball lands on the ball table after the training subject hits the ball. The non-end area can indicate an invalid area where the ball lands on the ball table. In a specific example, the scoring reference area can include a receiving area, a restricted area, and an end area. The training subject can be informed that he is only allowed to receive the ball in the receiving area and is not allowed to receive the ball in the restricted area, and that after hitting the ball, the ball should fall into the corresponding end area.

[0062] For example, each scoring reference area can be further subdivided. For example, the end area can be divided into multiple subareas. For example, the end area is located on the first side of the tee, where the first side is the side away from the training subject. The end area can be divided into three subareas: a left end subarea, a middle end subarea, and a right end subarea. The middle subarea is located at the center of the width of the tee, and the left end subarea and the right end subarea are located on either side of the middle subarea.

[0063] It should be noted that the corresponding positions of these scoring reference areas on the surface of the table tennis table can be marked with different colors or patterns. Alternatively, they can be left unmarked. For example, the table tennis table can be an existing conventional table tennis table tennis table.

[0064] In this step, a variety of suitable methods can be used to perform region segmentation on the motion scene image. For example, the motion scene image can be input into a trained region segmentation model to obtain a scoring reference region.

[0065] Step S130, determining training evaluation information for the training subject based on the hitting reaction information, includes steps S131 and S132.

[0066] Step S131 : determining the exercise score of the training subject according to the relative position relationship between the hitting-related position information and the corresponding scoring reference area.

[0067] For example, the hitting position information corresponds to the receiving area in the above example. Therefore, the training subject's athletic score can be determined based on the relative positional relationship between the hitting position information and the receiving area. Different relative positional relationships can correspond to different athletic scores. For another example, the hitting end position information corresponds to the end area in the above example. During the subdivision of the end area, the end sub-area corresponding to the current hitting end position information can be determined. Then, the training subject's athletic score can be determined based on the relative positional relationship between the hitting end position information and the corresponding end sub-area.

[0068] For example, in this step, each piece of batting reaction information can be scored separately, and a weighted average method can be used to determine the training subject's athletic score. Alternatively, the relative positional relationship between the batting reaction information and the batting position information and the corresponding scoring reference area can be input into a scoring model, which then outputs a final athletic score.

[0069] Step S132: Determine training evaluation information for the training subject according to the exercise score.

[0070] In an embodiment of the present application, different motion scores may correspond to different training evaluation information. The correspondence information between the motion scores and the training evaluation information may be pre-stored in the memory of the terminal device. In this step, after calculating the motion score of the training object, the processor may determine the training evaluation information of the training object by looking up a table. For example, the memory may store a correspondence table of five motion performance levels of "invalid", "failed", "qualified", "good", and "excellent" and five score intervals. In this step, the processor may determine the motion performance level of the training object based on the calculated motion score and the correspondence table. For example, when it is determined that the motion performance level of the training object in the current batting round is "excellent", a congratulatory screen corresponding to "excellent" may be displayed on the display screen, or a clap prompt sound may be output through the connected speaker.

[0071] This approach uses image processing and region segmentation techniques to reduce the impact of human factors on evaluation results, making the assessment more objective and accurate. By determining the relative positional relationship between the scoring reference area and the relevant hitting position information, the training subject's hitting accuracy can be accurately assessed. Furthermore, the comprehensive evaluation of the training subject's athletic performance using a motion score is more intuitive and facilitates comparison and analysis. This evaluation method is simple in execution logic and is also more efficient and accurate.

[0072] In one embodiment, ball sports include different motion modes. Before determining the motion score of the training subject in step S131, the ball sports training method of this embodiment of the present application further includes step S1301. Step S1301 determines, based on the current motion mode, a scoring reference area corresponding to the position information related to the shot.

[0073] In the embodiments of the present application, different sports modes correspond to different hitting rules and scoring criteria. In different sports modes, the scoring reference area corresponding to the hitting-related position information may be different. In the embodiments of the present application, the scoring reference area corresponding to the hitting-related position information is also determined based on the current sports mode.

[0074] In one example, the motion mode may include at least two categories: a receiving mode (in which the training subject only receives the ball and does not serve) and a serving mode (in which the training subject serves). The scoring reference areas corresponding to the same type of hitting position information determined in the receiving mode and the serving mode may be the same or different. For example, the scoring reference areas corresponding to the hitting position information determined in the receiving mode and the serving mode are different. For example, in the serving mode, the hitting position information may not correspond to a scoring reference area; while in the receiving mode, the hitting position information may correspond to the receiving area on the table. For another example, the scoring reference areas corresponding to the hitting end position information determined in the receiving mode and the serving mode are different. For example, the receiving mode may be divided into a static receiving mode and a dynamic receiving mode. In the static receiving mode, the scoring reference area corresponding to the hitting end position information determined may be the middle end sub-area described above; while in the serving mode, the scoring reference area corresponding to the hitting end position information determined may be the left end sub-area or the right end sub-area.

[0075] For example, the terminal device's memory may pre-store information on correspondences between different ball-hitting position information and scoring reference areas under different motion modes. In this step, after determining the subject's ball-hitting reaction information during each ball-hitting round, the scoring reference area corresponding to the ball-hitting position information can be accurately determined based on this correspondence information and the current motion mode. Furthermore, the subject's athletic score can be accurately determined based on the relative positional relationship between the ball-hitting position information and the corresponding scoring reference area under the current motion mode.

[0076] In this approach, the scoring reference area is determined based on the current movement pattern, making the assessment more personalized and enabling customized analysis based on the trainee's specific performance. This approach can also adapt to different movement patterns and training scenarios, adapting to the trainee's performance in different situations, and improving the accuracy of the assessment. This approach can also make exercise training more engaging and motivate the trainee. Furthermore, this approach reduces human interference, improving the consistency and reliability of the assessment.

[0077] In one embodiment, the shot reaction information includes shot position information. The scoring reference area includes a receiving area. The table is provided with at least one serving channel, and the motion mode includes a dynamic serving mode. In the dynamic serving mode, the ball is launched from the serving channel and rolls toward the side where the training subject is located. Step S1301 determines the scoring reference area corresponding to the shot position information based on the current motion mode, including step S1301a. Step S1301a, when the current motion mode is the dynamic serving mode, determines that the scoring reference area corresponding to the shot position information is the receiving area.

[0078] In an embodiment of the present application, the ball receiving area can be set according to actual needs, and it can be set in an area on the surface of the table that is closer to the training object. The serving channel can have any suitable form, which can be a serving channel connected to an automatic serving device, or it can be a mechanical pipe that facilitates manual serving. For example, with reference to Figure 2, the serving channel may include a hose fixed to the first side of the table (the opposite side of the training object). In the dynamic serving mode, the coach or training therapist can manually place the ball into the hose so that the ball rolls out of the hose and falls on the table.

[0079] In one example, the number of serving channels can be one. For example, an automatic serving device can be fixed in the middle of a first side of a table tennis table. The automatic serving device can randomly launch ping-pong balls of different colors at a certain frequency. The training subject can make a catching decision based on the color of the ball and pre-acquired catching rules (e.g., red and green balls can be caught, but blue and yellow balls cannot be caught), and then catch the ball according to the catching decision.

[0080] In another example, there can be multiple serving lanes. For example, as shown in Figure 3a, three serving lanes are provided on the first side of the table. These three serving lanes are located in the middle of the first side (position A0 in the figure) and at two corners of the first side (positions A1 and A2 in the figure). The receiving area on the table can be the area to the left of the linear dividing mark in Figure 3a (near the training subject). The dark gray filled area to the right of the linear dividing mark in Figure 3a can be the restricted area. The training subject is informed in advance that in dynamic receiving mode, they can receive the ball in the receiving area but not in the restricted area. For example, the ping-pong ball can be randomly released from any of the three serving lanes. The three serving lanes can also have different colors. For example, after the ball is released, the training subject can choose to receive or not receive the ball based on the color of the ball and the color of the serving lane, according to pre-known receiving rules. If the training subject decides to receive the ball, they can wait for the ping-pong ball to land in the receiving area. If it does, they can receive the ball. In this example, the coordinates of the ball's center pixel and the coordinate range of the ball receiving area can be used to determine whether the ball falls within the receiving area. In other words, the training subject's athletic score can be determined based on whether the ball falls within the receiving area. For example, if the ball falls within the receiving area, the score corresponding to its receiving position information may be 10 points, while if it falls outside the receiving area, the score corresponding to its receiving position information may be 0 points. Ultimately, the overall athletic score of the training subject can be determined based on the scores of each individual ball-hitting reaction information.

[0081] In one embodiment, the shot reaction information includes shot end position information. The scoring reference area includes an end area, which includes multiple end sub-areas. The sports mode includes a fixed-point serve mode. The table is provided with an indicator device configured to provide a user with dynamic prompt information indicating the correct ball landing position. In the fixed-point serve mode, the training subject serves the ball from a preset position according to the dynamic prompt information.

[0082] Step S1301 determines a scoring reference area corresponding to the shot-related position information according to the current motion mode, including step S1301b.

[0083] Step S1301b, when the current motion mode is the fixed-point serve mode, based on the relative positional relationship between the hitting end position information and the end area corresponding to the correct ball landing position, determine that the scoring reference area corresponding to the hitting position information is the end sub-area where the correct ball landing position is located.

[0084] For example, the indicator device can be an indicator light that emits different colors of light. For example, the indicator light can emit red, yellow, or blue light, with each color randomly changing. Different colors of light represent different prompts. This prompt information can be information about the correct ball landing location. Of course, the indicator device can also be other devices, such as a display screen that dynamically plays prompts about the correct ball landing location. Alternatively, the indicator device can be an audio player, and the prompt information about the correct ball landing location can also be audio information.

[0085] Exemplarily, an indicator device for providing visual prompt information can be set on the first side of the table to facilitate user observation. In the case where the indicator device is an indicator light, the number of indicator lights can be one or more. For example, the number of indicator lights can be one, and the indicator light can be located above the teeing channel at position A0 shown in Figure 3a. Exemplarily, the indicator light being red can indicate that the current correct ball landing position is the end sub-area at position A0, that is, the middle end sub-area (the area shown in the dotted box at A0 in Figure 3a); the indicator light being blue can indicate that the current correct ball landing position is the end sub-area at position A1, that is, the right end sub-area (the area shown in the dotted box at A1 in Figure 3a); the indicator light being green can indicate that the current correct ball landing position is the end sub-area at position A2, that is, the left end sub-area (the area shown in the dotted box at A2 in Figure 3a). In another example, the number of indicator devices can be multiple.

[0086] As shown in FIG3b , three indicator devices (indicator lights) may be provided on the first side of the table. The indicator lights may randomly emit red or blue light. In the fixed-point serving mode, the training subject may serve the ball from the cross position in FIG3b . Specifically, the correct ball landing position may be determined based on the color of the light emitted by the three indicator lights at positions A0, A1, and A2 and the pre-known serving rules. For example, the serving rules may include: when the indicator light is blue, the end point sub-area at the indicator light position is not the correct ball landing position; when the indicator light is red, the end point sub-area at the indicator light position is the correct ball landing position. In this example, the correct ball landing position for each serve may be one or more. The training subject may determine the decision about the end point of the serve based on the optical signals observed by the eyes and based on the pre-known serving rules.

[0087] In an embodiment of the present application, when the current motion mode is a fixed-point serve mode, a scoring reference area corresponding to the shot position information can be determined based on the relative positional relationship between the shot end position information and the end area corresponding to the correct ball landing position, combined with the dynamic prompt information currently provided by the indicator device. The scoring reference area corresponding to the training subject's shot position information during the current shot round can be determined based on the correct ball landing position indicated in the dynamic prompt information provided by the indicator device. For example, if the indicator light at position A1 in FIG3b is red and the other two indicator lights are green, the end sub-area at position A1 is determined as the scoring reference area corresponding to the training subject's shot position information. Furthermore, the training subject's motion score can be determined based on whether the shot position information is within the corresponding scoring reference area. For example, if the training subject's shot position information falls within the end sub-area at position A1, the score corresponding to the shot position information can be recorded as a first score X1. If the training subject's shot position information does not fall within the end sub-area at position A1, the score corresponding to the shot position information can be recorded as a first score X2, where X1>X2. For example, if the shot position information is within the corresponding scoring reference area, the first score X1 can be determined based on the position of the shot position information within the scoring reference area. That is, each end point sub-area can be further divided into multiple secondary sub-areas. As shown in Figure 3b, each end point sub-area can be further divided into five secondary sub-areas, with different secondary sub-areas corresponding to different scores. For example, for each of the five secondary sub-areas within the end point sub-area in Figure 3b, darker secondary sub-areas correspond to higher scores. This can further improve the accuracy of the assessment of the training subject's ability status.

[0088] In one embodiment, the shot reaction information includes information on whether the ball was shot and / or information on the end point of the shot. Step S130 determines training evaluation information for the training subject based on the shot reaction information, including step S133. Step S133 determines training evaluation information for the training subject based on the shot reaction information and the shot guidance information. The shot guidance information is guidance information for the training subject to make a shot decision, including whether to shoot and / or determining the shot destination.

[0089] For example, the batting guidance information may include appearance attribute information of the ball (e.g., color, pattern, etc.), and the training subject can make decisions about whether to catch the ball and where to hit the ball based on the appearance attribute information of the ball. For another example, the batting guidance information may include appearance attribute information of the serving lane (e.g., color, pattern, etc.), and the training subject can make decisions about whether to catch the ball and where to hit the ball based on the appearance attribute information of the serving lane. For another example, the batting guidance information may include instruction information provided by an instruction device (e.g., sound, light, displayed image, displayed text, etc.).

[0090] For example, reference reaction information (e.g., correct shot reaction information, such as shot determination information, shot position information, shot end position information, etc.) corresponding to at least a portion of the shot reaction information (e.g., first shot reaction information) can be determined based on the shot guidance information and preset shot rules. Subsequently, training evaluation information for the training subject can be determined based on the difference between the first shot reaction information and the corresponding reference reaction information. Specifically, when the sports training includes multiple motion modes, reference reaction information corresponding to at least a portion of the shot reaction information can be determined based on the shot guidance information, preset shot rules, and the current motion mode.

[0091] In the above scheme, by introducing appropriate batting guidance information into ball sports training, not only can the fun of sports training be improved, but also the cognitive ability of the training subjects can be trained, and a comprehensive assessment of the cognitive level and sports level of the training subjects can be achieved, resulting in better training effects.

[0092] In one embodiment, the batting guidance information includes visual marking information carried by a guidance unit, which includes a ball and / or an indicator device. The indicator device is located within the visual range of the image acquisition device. Before determining the training evaluation information for the training subject in step S133, the ball sports training method provided in this embodiment of the application also includes step S102. In step S102, batting guidance information is determined based on the pixel values ​​of pixels in the guidance area within the sports scene image, where the guidance area includes the imaging area of ​​the guidance unit.

[0093] For example, the batting guidance information may include the color information of the ball. The indicating device may be an indicator light, and the batting guidance information may include the color of the light emitted by the indicator light. The guidance area may include the imaging area of ​​the ball, or may include the imaging area of ​​the indicator light. In an embodiment of the present application, the motion scene image may be first segmented to obtain the guidance area in the image. Then, the batting guidance information may be determined based on the pixel values ​​of the guidance area. For example, in a static serve mode, the pixel values ​​of the pixels in the imaging area of ​​the indicator light at A0 in FIG3b may be extracted, and when the pixel values ​​of these pixels in the red channel are greater than 200, it is determined that the indicator light is lit red. Then, the correct landing position of the ball may be determined to be the end point sub-area at A0. Then, the score corresponding to the batting end point position information may be determined based on the relative positional relationship between the batting end point position information determined in step S120 and the end point sub-area.

[0094] In this approach, introducing visual batting guidance into ball sports training not only increases the interactivity and fun of the training, but also accurately assesses the participants' cognitive level and hand-eye coordination. Determining batting guidance using computer vision further reduces manual intervention and improves the objectivity and accuracy of the assessment.

[0095] In one embodiment, the ball sports training method provided in the embodiment of the present application also includes the following steps: step S141, performing key point detection on the human body area in the motion scene image to determine the position information of multiple upper limb key points; step S142, determining the motion posture information of the training object based on the position information of the upper limb key points; step S143, determining the disease category information of the training object based on the motion posture information.

[0096] In embodiments of the present application, a key point detection model can be used to locate key points on a training subject's upper limbs (e.g., key points on the elbow, shoulder, and wrist). For example, referring to Figure 4 , the key points on the upper limbs detected by the key point detection model may include the left elbow point P1, the right elbow point P2, the left wrist point P3, and the right wrist point P4. The position information of the upper limb key points output by the key point detection model can be input into a trained posture assessment model to obtain the training subject's motion posture information. Specifically, the motion posture information can include motion posture category information. For example, the training subject's posture can be classified into normal posture and abnormal posture. Of course, abnormal posture can also be further subdivided into different abnormal postures. For example, if the training subject's posture information indicates an abnormal posture, the training subject's disease category information can be further determined based on the training subject's motion posture information. For example, motion posture information of patients with different disease categories can be pre-collected to construct a disease classification dataset. The disease classification dataset can then be used to train a disease classification model. In embodiments of the present application, the training subject's motion posture information can be input into a trained disease classification model to obtain the training subject's disease category information.

[0097] In the above scheme, the computer vision method not only realizes the objective and accurate evaluation of the ability status of the training subjects, but also realizes the classification of the diseases of the training subjects, providing strong support for the health management of the training subjects.

[0098] In one embodiment, before determining the training evaluation information of the training object in step S130, the ball sports training method of the embodiment of the present application further includes step S103.

[0099] Step S103: During the process of the training subject performing ball sports, first evaluation reference information is collected and obtained, wherein the first evaluation reference information includes at least one of the following training data: biomechanical data of the training subject, physiological data of the training subject, psychological data of the training subject, and mechanical data of sports equipment used or operated by the training subject.

[0100] In one example, a ball sports training system may include a variety of information acquisition devices. For example, the information acquisition device includes: an image sensor (such as the image sensor in the above-mentioned image acquisition device), a motion sensor, a pressure sensor, a cardiopulmonary function monitoring device, an electroencephalogram (EEG) activity monitoring device, an eye movement assessment device, a balance ability assessment device, a fine motor assessment device, a cognitive assessment device, and a sensory assessment device. The collection of these information acquisition devices can be referred to as an information acquisition device group. Utilizing the various devices in the information acquisition device group, the biomechanical data of the training subject, the physiological data of the training subject, the psychological data of the training subject, and the mechanical data of the sports equipment used or operated by the training subject can be comprehensively collected.

[0101] For example, the subject's biomechanical data may include: motion trajectory data (e.g., motion direction, speed, accuracy, etc.) of the subject's limbs, torso, or fingers. The subject's biomechanical data may also include: pressure data applied by the subject to a monitored location during exercise. For example, the pressure data detected by a pressure sensor on the handle of a tennis racket when the subject grips it.

[0102] For example, the training subject's physiological data may include various physiological data that can reflect the training subject's health status, such as heart rate, blood pressure, respiratory rate, blood oxygen concentration, etc. The training subject's psychological data may also include brain electrical activity data, such as an electroencephalogram (EEG).

[0103] For example, the psychological data of the training subject may include attention evaluation data, sensory evaluation data, and the like.

[0104] For example, the mechanical data of the sports equipment used or operated by the training subject may include the kinematics data of the sports equipment, such as the motion trajectory data (motion position, motion direction, motion speed, etc.) of a ball or racket, and the impact force data of the ball on the racket when hitting the ball.

[0105] Step S130 determines training evaluation information for the training object, including step S134.

[0106] Step S134: Determine training evaluation information for the training subject based on the hitting reaction information and the first evaluation reference information.

[0107] For example, while a training subject is playing a ball game, the information acquisition device group can collect and obtain the aforementioned various training data. In this step, the determined training subject's ball-hitting reaction information and the various training data acquired by the information acquisition device group are combined to accurately determine training evaluation information for the training subject. For example, a deep learning model for training evaluation can be pre-built and trained. In this step, the determined training subject's ball-hitting reaction information and the various training data can be combined and input into the deep learning model, which then outputs training evaluation information.

[0108] In this approach, training evaluations are conducted by integrating the subject's batting reaction information with various evaluation references, resulting in more accurate evaluation results. Based on this, training is then tailored to the subject, making training more personalized and allowing for real-time monitoring and quantification of training progress, significantly improving training effectiveness.

[0109] In one embodiment, the ball sports training method provided in the embodiments of the present application further includes step S144. In step S144, based on the evaluation information of the training subject's sports performance, a sports training recommendation is determined for the training subject. The sports training recommendation includes a sports training course that matches the training subject's sports performance.

[0110] Exemplarily, the level of subsequent sports training can be determined based on the evaluation results of the sports performance of the training object, so that the sports training is more suitable for the training object. Specifically, a sports training course that matches the sports performance of the training object can be determined based on the evaluation results of the sports performance of the training object, and the sports training course can be recommended to the training object. Exemplarily, the sports training course can be a technology-enhanced sports training course, for example, a technology-enhanced table tennis training (Technology enhanced table tennis, referred to as TeTT) ​​course. Courses corresponding to different training intensity levels can be included in this type of course series, such as T1 to T10 levels, and the higher the level, the higher the training intensity of the corresponding course. In this step, after determining the sports score of the training object, a training course of a level that matches its sports score can be recommended to the training object based on the sports score.

[0111] It should be noted that the training content of the training course includes but is not limited to the trainees' (training subjects') hand-eye coordination, executive function, inhibition, operational memory, concentration, flexible thinking, balance, movement coordination, organizational ability, sensory information processing ability and cooperation ability.

[0112] For example, during sports training, the training subject can be evaluated across different ability categories (such as hand-eye coordination, executive function, and inhibition) based on the various training data acquired by the various information acquisition devices. The resulting training evaluation information can include assessment information for the training subject across multiple ability categories. Subsequently, personalized training content can be customized based on the assessment information across these ability categories, reducing ineffective training and significantly improving the relevance, effectiveness, and accuracy of training.

[0113] In the above program, customized training suggestions are provided based on the results of the individual's athletic ability assessment, so that the training is more in line with individual needs and the training effect is improved. Through accurate assessment, ineffective training can be avoided, time waste can be reduced, and training can be made more efficient. Understanding the athletic ability limitations of the training subjects can design safe training plans and reduce the risk of sports injuries. Based on the assessment results, the training subjects' weaknesses can be targeted to promote the overall improvement of skills. For training subjects of different levels, training suggestions of different difficulty levels can be provided to make the training more adaptable. This program improves the targetedness, safety, and effectiveness of training through scientific assessment and personalized suggestions, while also enhancing the participation and satisfaction of the training subjects.

[0114] In one embodiment, the ball sports training method provided in the embodiments of the present application further includes step S145. Step S145 determines health intervention recommendations for the training subject based on the assessment information of the training subject's physical health status. For example, the assessment information of the physical health status of a stroke patient can be used to assess the severity of stroke symptoms. Furthermore, based on the severity, appropriate treatment and intervention recommendations can be determined for the patient, facilitating personalized and scientific health interventions.

[0115] In this solution, by assessing the physical health of the trainee and providing personalized health intervention recommendations, the health needs of different individuals can be met. Health intervention recommendations can help trainees improve their health, and combined with health intervention recommendations, training results can be more effectively enhanced. Through continuous health status assessment, the health status of the trainee can be monitored over the long term, providing continuous data support for sports training and health management. Therefore, this solution also provides trainees with more comprehensive health support.

[0116] In a second aspect, a ball sports training system is provided. As shown in Figure 5 , the ball sports training system 500 provided in an embodiment of the present application includes an image acquisition device 510 and a control device 520. The image acquisition device 510 is used to capture images of the motion scene at a preset frequency while the training subject is performing a ball sport; the control device 520 is used to execute the aforementioned ball sports training method. The control device 520 can be the ball sports training control device provided in any embodiment of the present application.

[0117] The ball sports training system described above not only automatically, efficiently, and accurately objectively assesses the training subject's performance status, but also offers low device cost, ease of implementation, and adaptability to various scenarios. Furthermore, using this ball sports training system as a rehabilitation training system can significantly enhance the enjoyment of rehabilitation training, stimulate patients' initiative and enthusiasm for training, and thus significantly improve the effectiveness of rehabilitation training.

[0118] In one embodiment, the ball sports training system further includes a first information acquisition device connected to the control device, the first information acquisition device being configured to collect and obtain first evaluation reference information while the training subject is performing the ball sports. The control device 520 is further configured to determine training evaluation information for the training subject based on the training subject's hitting reaction information and the first evaluation reference information. The first information acquisition device includes at least one of the following: a motion sensor, a pressure sensor, a cardiopulmonary function monitoring device, an electroencephalogram (EEG) activity monitoring device, an eye movement evaluation device, a balance ability evaluation device, a fine motor skills evaluation device, a cognitive evaluation device, and a sensory evaluation device. The first evaluation reference information includes at least one of the following training data: the training subject's biomechanical data, the training subject's physiological data, the training subject's psychological data, and mechanical data of the sports equipment used or operated by the training subject.

[0119] In embodiments of the present application, motion sensors can be used to collect motion data from target parts of a training subject. Target parts can include upper limbs, lower limbs, torso, and fingers. Motion data can include various motion trajectory data, such as motion speed, direction, and accuracy. In embodiments of the present application, pressure sensors can be used to monitor the pressure applied by the training subject to various appropriate locations during training. For example, pressure data applied by various parts of the training subject to the ground, a seat, or contact objects can be collected. Cardiopulmonary function monitoring devices can include heart rate monitors and blood pressure monitors. EEG activity monitoring devices can include near-infrared spectroscopy (fNIRS) and electroencephalogram (EEG) acquisition devices. Eye movement assessment devices can include various types of eye movement testing devices and can be used to assess eye movement control and attention. Balance assessment devices can be used to assess a training subject's balance ability. Fine motor skills assessment devices can be used to evaluate a training subject's fine motor control, such as bilateral hand coordination and dexterity. Cognitive assessment devices can be used to assess an individual's executive function skills and attention abilities. The sensory evaluation device can be used to evaluate the regulatory ability of the sensory nerves of the training subject.

[0120] For example, the ball sports training system in the embodiment of the present application may include a device group consisting of multiple information acquisition devices. The device group may include at least an image acquisition device (for ease of understanding, it can be referred to as a second information acquisition device), and may also include a motion sensor, a pressure sensor, a cardiopulmonary function monitoring device, an EEG activity monitoring device, an eye movement assessment device, a balance ability assessment device, a fine motor assessment device, a cognitive assessment device, a sensory assessment device, etc. For example, the ball sports training system may also include a data collection platform that can collect in real time the training data collected and processed by the various information acquisition devices in the device group. And these training data can be uploaded to a cloud server. In one example, the cloud server may include the above-mentioned control device 520. After the control device obtains these training data, it can integrate these training data and determine the training evaluation information for the training object in real time. The cloud server can send the training evaluation information to the connected terminal device in real time and can display the training evaluation results on the display screen of the terminal device. In another example, the cloud server can be used only to store these training data. The terminal device connected to the cloud server has a "training evaluation" application installed. When the terminal device's processor runs this application, it can obtain this training data in real time and determine the training evaluation information for the training object based on this training data. The training evaluation results can also be displayed on the display in real time.

[0121] The various information acquisition devices (including the first information acquisition device) in the above-mentioned device group can be devices with low cost and good compatibility. Not only can the device group be used to monitor training data during various ball sports training to complete ball sports training, but the device group can also be used to monitor training data during other types of sports (non-ball sports, such as cycling training) training. Before training begins, training evaluators (such as coaches, rehabilitation trainers) can select appropriate information acquisition devices from the device group according to actual needs to monitor appropriate spatial areas or body parts. For example, when conducting table tennis training, you can choose an image acquisition device (to collect images of the table area and the upper body), a motion sensor (to monitor the exercise data of the training subject's arms), a cardiopulmonary function monitoring device (to monitor heart rate, blood pressure, respiratory rate, blood oxygen concentration), an EEG activity monitoring device (to monitor EEG activity signals at specific locations in the brain), an eye movement assessment device (to monitor the activity data of the training subject's eyes and evaluate their eye attention data) and a cognitive assessment device (to monitor and evaluate the training subject's executive function and evaluate their concentration data). During the training process, the data collection platform can collect the training data collected by the selected information acquisition device in real time and upload it to the cloud server.

[0122] It is understandable that ball sports training has a good effect in improving people's sports skills, cognitive abilities and physical functions. However, for the elderly, children and people with cognitive, physical and psychological disorders, learning new sports skills may face more challenges. The teaching methods required by these groups are significantly different from those of healthy peers. In addition, different training subjects have different personal abilities and different exercise requirements during sports training. During the training process, it is necessary to integrate the training data of the training subjects in multiple aspects to achieve the safety, scientificity and effectiveness of the training process, as well as the adaptability to the personal abilities of the training subjects. Traditional ball sports training equipment often only focuses on the training data of the training subjects in biomechanics, and lacks training equipment that can integrate the physiological data, biomechanical data and psychological data of the training subjects. Therefore, the training effect, accuracy and training safety of existing ball sports training equipment are all poor. According to the above-mentioned ball sports training system of the embodiment of the present application, the various information acquisition devices in the device group can monitor the training data of the training subjects in multiple aspects (physiology, biomechanics and psychology) in real time, and can also monitor the mechanical data of sports equipment. Through comprehensive monitoring of training data, the ball sports training system can achieve timely and accurate training evaluation of training subjects, and can accurately provide matching sports training suggestions.

[0123] Furthermore, this ball sports training system can meet the needs of different user groups, including those with disabilities or users across age groups. Through integrated data analysis, it can more accurately provide matching training recommendations for each training subject, making training activities more precise and efficient. This ball sports training system can optimize rehabilitation training methods, fully meeting the sports training needs of training subjects with health conditions, and can also meet the training needs of normal athletes, improving their athletic performance. Furthermore, this ball sports training system has good usability for different target users. For example, athletes can use this ball sports training system to improve their physical function; trainers and healthcare professionals can use this ball sports training system to provide professional training and training recommendations to trainees, and can also more conveniently monitor their training progress. Furthermore, researchers can use this ball sports training system for data collection and subsequent product development, thereby improving the efficiency of research and development.

[0124] In a specific example, the above-mentioned ball sports training system can be used to conduct rehabilitation training for stroke patients, and real-time monitoring and evaluation of the training status of stroke patients can be achieved. The evaluation results can be sent to the mobile phones of the stroke patient's attending physician, family members and rehabilitation trainers, so that they can understand the training progress and rehabilitation status of the stroke patient in real time and accurately, thereby significantly improving the efficiency, accuracy and personalization level of rehabilitation training.

[0125] In one embodiment, the ball sports training system further includes a social robot. The social robot includes an output device. The social robot is configured to output exercise prompt information, including exercise suggestion information and / or exercise motivation information, during the training subject's ball sports performance.

[0126] In an embodiment of the present application, the social robot may include a voice output device (such as a speaker) and a display device (such as a display screen). The interaction between the social robot and the training subject may be implemented based on artificial intelligence technology. In one example, during the training process of the training subject, the social robot can interact with the patient and provide suggestions for various professional exercises and physical and cognitive training activities to help the training subject achieve better training results. In another example, the social robot can also simulate training movements to help the training subject better understand and master training techniques and improve training results. In another example, the social robot can also monitor the training subject's movements, postures, and technical details in real time through the sensors and cameras it carries, and provide feedback. In another example, the social robot can also identify the training subject's identity and can promptly obtain the training subject's historical training records. During the training process, the social robot can select a training method that matches the training subject to assist the training subject (for example, as the training subject's opponent, playing table tennis with the training subject). In another example, the social robot also communicates with the cloud server in real time. During sports training (such as rehabilitation training for patients through table tennis), it can obtain the patient's physiological data in real time. When the physiological data is abnormal, the social robot can promptly alert the patient and call a professional caregiver when necessary. The social robot can prompt the patient through voice to moderately increase or decrease the intensity of training and provide encouragement. In another example, the social robot can improve the training motivation, learning, confidence and enthusiasm of the training subject through gamification tasks. The social robot can play the role of a competitive partner or coach, and improve the patient's participation and enthusiasm in training through social interaction. In one example, the social robot can also play the role of a rehabilitation trainer, and by using rich visual expressions such as language, gestures, and expressions, and natural conversational voice to comfort, guide and motivate patients, it can improve credibility and expectations, and thus improve the patient's psychological state and rehabilitation effect during the rehabilitation process. Alternatively, the social robot can be remotely controlled by a real coach or rehabilitation therapist, and the voice and language of the social robot can be the real voice and language of the coach or rehabilitation therapist.

[0127] Setting up a social training robot in the ball sports training system can provide users with more personalized training plans, and has a good effect in accompanying and social interaction with the training subjects, thereby better motivating the training subjects to actively participate in sports training and further improving the training effect.

[0128] In a third aspect, a ball game training control device is provided. As shown in FIG6 , the ball game training control device 600 provided in an embodiment of the present application includes an acquisition module 610 , a determination module 620 , and an evaluation module 630 .

[0129] The acquisition module 610 is used to acquire a plurality of sports scene images captured by the image acquisition device and the acquisition time of each sports scene image during the process of the training subject playing ball games.

[0130] The determination module 620 is used to determine the training subject's batting reaction information based on multiple motion scene images and the acquisition time of each motion scene image, wherein the batting reaction information includes at least one of batting information, batting-related position information, batting reaction time information, and batting speed information. The batting information includes first information indicating that the training subject batted the ball or second information indicating that the training subject did not bat. The batting-related position information includes batting position information and / or batting end position information. The batting position information is the position information of the ball when the training subject bats the ball, and the batting end position information is the position information of the destination reached by the ball after the training subject bats the ball.

[0131] The evaluation module 630 is used to determine training evaluation information for the training subject based on the hitting reaction information, wherein the training evaluation information includes evaluation information of the training subject's sports performance and / or evaluation information of the training subject's physical health status.

[0132] In a fourth aspect, a terminal device is provided. As shown in FIG7 , a terminal device 700 provided in an embodiment of the present application includes: at least one processor 710 (only one processor is shown in FIG7 , but this is not limiting), a memory 720, and a computer program 730 stored in the memory 720 and executable on the at least one processor 710. When the processor 710 executes the computer program 730, the steps of the ball sports training method described above are implemented.

[0133] In the embodiments of the present application, the terminal device may include, but is not limited to, a processor and a memory. FIG7 is merely an example of a terminal device and does not constitute a limitation on the terminal device. The terminal device may include more or fewer components than shown, or a combination of certain components, or different components. The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0134] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / modules are based on the same concept as the method embodiment of this application. Their specific functions and technical effects can be found in the method embodiment section and will not be repeated here.

[0135] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The functional modules in the embodiment can be integrated into a processing module, or each module can exist physically alone, or two or more modules can be integrated into one module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software functional modules. In addition, the specific names of the functional modules are only for the convenience of distinguishing each other and are not used to limit the scope of protection of this application. The specific working process of the modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0136] In a fifth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned ball sports training method are implemented.

[0137] In a sixth aspect, a computer program product is provided, which, when executed on a terminal device, enables the terminal device to implement the steps in the above-mentioned ball sports training method.

[0138] The above are merely optional embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included within the scope of the claims of the present application.

Claims

1. A ball sports training method, characterized in that: include: During the process of the training subject playing ball sports, acquiring a plurality of sports scene images acquired by the image acquisition device and the acquisition time of each sports scene image; Determine the batting reaction information of the training subject according to the multiple motion scene images and the acquisition time of each motion scene image, wherein the batting reaction information includes at least one of batting information, batting-related position information, batting reaction time information, and batting speed information, the batting information includes first information indicating that the training subject batted the ball or second information indicating that the training subject did not bat, the batting-related position information includes batting position information and / or batting end position information, the batting position information is position information of the ball when the training subject batted the ball, and the batting end position information is position information of the destination reached by the ball after the training subject batted the ball; According to the batting reaction information, training evaluation information for the training subject is determined, wherein the training evaluation information includes evaluation information of the training subject's athletic performance and / or evaluation information of the training subject's physical health status.

2. The ball sports training method according to claim 1, characterized in that: The ball sport is table tennis, the visual range of the image acquisition device includes the surface area of ​​the table where the table tennis ball is located, the hitting reaction information includes the hitting-related position information, and before determining the training evaluation information for the training object, the method further includes: Performing region segmentation on the motion scene image to determine a scoring reference region in the motion scene image, wherein the scoring reference region is located in a ball case region, and the ball case region is an imaging region of the ball case; The step of determining the training evaluation information for the training object according to the hitting reaction information includes: Determining the exercise score of the training subject according to the relative position relationship between the ball-hitting related position information and the corresponding scoring reference area; According to the exercise score, training evaluation information for the training object is determined.

3. The ball sports training method according to claim 2, characterized in that: The ball sports include different sports modes. Before determining the sports score of the training object, the method further includes: According to the current motion mode, a scoring reference area corresponding to the ball-hitting related position information is determined.

4. The ball sports training method according to claim 3, characterized in that: The hitting reaction information includes the hitting position information, the scoring reference area includes a receiving area, the table is provided with at least one serving channel, the motion mode includes a dynamic serving mode, in which the ball is sent out from the serving channel and rolls toward the side where the training subject is located, and determining the scoring reference area corresponding to the hitting-related position information according to the current motion mode includes: When the current motion mode is the dynamic serving mode, the scoring reference area corresponding to the hitting position information is determined as the receiving area.

5. The ball sports training method according to claim 4, characterized in that: The batting reaction information includes the batting end position information, the scoring reference area includes an end area, the end area includes a plurality of end sub-areas, the motion mode includes a fixed-point serving mode, the table is provided with an indicator device, the indicator device is used to provide a user with dynamic prompt information of a correct ball landing position, in the fixed-point serving mode, the training subject serves from a preset position according to the dynamic prompt information, and determining the scoring reference area corresponding to the batting-related position information according to the current motion mode includes: When the current motion mode is the fixed-point serving mode, the scoring reference area corresponding to the hitting position information is determined to be the end point sub-area where the correct ball landing position is located.

6. The ball sports training method according to any one of claims 1 to 5, characterized in that: The hitting reaction information includes the hitting information and / or the hitting end position information. The determining of the training evaluation information for the training object according to the hitting reaction information includes: Based on the batting reaction information and the batting guidance information, training evaluation information for the training subject is determined, wherein the batting guidance information is guidance information for the training subject to make a batting decision, and the batting decision includes a decision on whether to bat and / or a decision on determining a batting destination.

7. The ball sports training method according to claim 6, characterized in that: The hitting guidance information includes visual marking information carried by a guidance unit, the guidance unit includes the ball and / or an indication device, the indication device is located within the visual range of the image acquisition device, and the method further includes: The hitting guidance information is determined according to pixel values ​​of pixels in a guidance area in the sports scene image, wherein the guidance area includes an imaging area of ​​the guidance portion.

8. The ball sports training method according to any one of claims 1 to 5, characterized in that: Before determining the training evaluation information for the training object, the method further includes: During the process of the training subject playing ball sports, first evaluation reference information is collected and obtained, wherein the first evaluation reference information includes at least one of the following training data: biomechanical data of the training subject, physiological data of the training subject, psychological data of the training subject, and mechanical data of sports equipment used or operated by the training subject; The determining of the training evaluation information for the training object includes: The training evaluation information for the training object is determined according to the hitting reaction information and in combination with the first evaluation reference information.

9. The ball sports training method according to any one of claims 1 to 5, characterized in that: The method further comprises: Determining sports training suggestions for the training subject based on the evaluation information of the training subject's sports performance, wherein the sports training suggestions include sports training courses that match the training subject's sports performance; and / or Determine health intervention recommendations for the training subject based on assessment information of the training subject's physical health status.

10. The ball sports training method according to any one of claims 1 to 5, characterized in that: The method further comprises: Performing key point detection on a human body region in the motion scene image to determine position information of a plurality of upper limb key points; Determining the motion posture information of the training subject according to the position information of the upper limb key points; The disease category information of the training object is determined according to the motion posture information.

11. A ball sports training system, characterized in that: include: An image acquisition device, used to acquire sports scene images at a preset frequency while the training subject is performing ball sports; A control device for executing the ball sports training method as described in any one of claims 1-9.

12. The ball sports training system according to claim 11, characterized in that: The ball sports training system further comprises a first information acquisition device connected to the control device, the first information acquisition device being used to collect and obtain first evaluation reference information during the process of the training subject performing ball sports; The control device is also used to determine the training evaluation information of the training object based on the batting reaction information of the training object and in combination with the first evaluation reference information; wherein the first information acquisition device includes at least one of the following: a motion sensor, a pressure sensor, a cardiopulmonary function monitoring device, an electroencephalogram activity monitoring device, an eye movement evaluation device, a balance ability evaluation device, a fine motor evaluation device, a cognitive evaluation device, and a sensory evaluation device.

13. The ball sports training system according to claim 11, characterized in that: The ball sports training system also includes a social robot, which includes an output device. The social robot is used to output sports prompt information using the output device during the process of the training object playing ball sports, and the sports prompt information includes sports suggestion information and / or sports motivation information.

Images