An immersive-based sports action performance visual analysis method and system

Abstract

The application discloses a kind of based on immersive sports action performance visual analysis method and system, including the following steps: using augmented reality technology to show the completion process of different action completion forms of specific sports action for user, and overall action performance condition;Show the movement condition of different body parts in different time periods in the completion process of any action completion form in front of user, action attribute is associated with action performance by using the way of multi-device cooperation Construction analysis, find the key factor that influences action performance;Key factor is marked in action completion process, user combines key factor to understand the reason that influences action performance, realizes sports action performance visual analysis.The method and system can build the correlation analysis between action attribute and action performance, find the factor that influences action performance and the mechanism that produces influence, realize immersive efficient sports action performance visual analysis.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of visual analysis, and particularly relates to a sports action performance visual analysis method and system based on immersion. BACKGROUND

[0002] In skill-driven projects such as badminton, baseball, golf and gymnastics, the actions of athletes determine whether the athletes can effectively perform the target task and ultimately achieve good results. Sports action performance analysis is an important means to improve the sports performance of athletes, and effective sports action performance analysis can help athletes correct actions, improve abilities and reduce the risk of injury.

[0003] Action performance analysis starts from the action attributes of each part of the body, such as spatial position, speed, angular velocity, etc., to explore the relationship between the action attributes of different body parts and action performance. The traditional action performance analysis method mainly relies on visual charts and statistical analysis techniques to determine the factors affecting sports performance and the correlation between them. A study (Mine K, Milanese S, Jones M A, et al. Pitching mechanics and performance of adult baseball pitchers: A systematic review and meta-analysis for normative data [J]. Journal of Science and Medicine in Sport, 2023: 69-76) first preliminarily determines the factors affecting the pitching performance of pitchers in baseball through literature research, then collects data on target attributes, and finally constructs a correlation analysis with the pitch speed to observe the patterns of action attributes in the form of a line chart. The results show that a larger stride and shoulder internal rotation speed peak of athletes can accelerate the ball speed, which is beneficial to the athletes to win the victory on the court.

[0004] However, the complexity of sports actions increases the difficulty of action performance analysis, and traditional analysis methods are difficult to support the exploration of the performance influencing mechanism of complex actions. Moreover, since the human body parts are connected to each other, the movement of a single part will drive the movement of other body parts, thereby indirectly affecting the action performance. Experts need to draw a large number of charts to find the characteristics of the attributes and possible correlations, then establish hypotheses, and finally perform correlation analysis to verify the hypotheses, which is time-consuming and laborious.

[0005] Meanwhile, the complexity of human structure also makes it difficult for experts to understand the data. Statistical charts abstractly describe the change process of the action, and experts need to make cognitive efforts to imagine the concrete performance of the action to understand the impact mechanism of the action performance. In order to reduce the cognitive burden of experts, biomechanics provides tools to display human actions in the form of human models, such as Xsen, OpenSim and Delsy. Although these tools also provide line charts to describe the change process of the specified attribute, statistical analysis methods are still needed to explore the action performance and the correlation between attributes.

[0006] Recently, the development of immersive analysis provides the possibility for efficient sports action performance analysis. Immersive technology builds a virtual three-dimensional scene in front of the expert, and the three-dimensional object is displayed concretely, which enhances the ability of experts to understand and explore data. Existing action analysis based on immersive visualization (Li A, Liu J, Cordeil M, et al. GestureExplorer: Immersive Visualisation and Exploration of Gesture Data [C] / / Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems. 2023: 1-16) tends to analyze the structural differences of different categories of actions. Like existing tools in sports biomechanics, the action is reproduced by using skeletons or virtual humans, so that experts can intuitively observe the completion of the complete action.

[0007] However, directly integrating statistical analysis into an immersive environment will reduce efficiency, and it will be more efficient to present in combination with two-dimensional charts. Currently, the analysis of sports action performance tends to observe the concrete structure of the action, and lacks analysis tasks related to correlation research. SUMMARY

[0008] In view of the above, the purpose of the present application is to provide an immersive-based sports action performance visual analysis method and system, which uses augmented reality technology to perform three-dimensional visualization of sports actions, and constructs correlation analysis of action attributes and action performance, and understands the implementation mechanism of the action in combination with the key factors affecting the action performance. It is suitable for efficient visual analysis of sports action performance and correlation research.

[0009] To achieve the above-mentioned purpose of the application, the technical scheme provided by the present application is as follows:

[0010] In the first aspect, the embodiment of the present application provides an immersive-based sports action performance visual analysis method, comprising the following steps:

[0011] The completion process of different action completion forms of a specific sports action is displayed for the user by using augmented reality technology, and the overall action performance is displayed.

[0012] In the display of the completion process of any action completion form, the movement of different body parts at different time periods is displayed for the user, and the action attributes and action performance are associated and analyzed by using a multi-device cooperative manner to find the key factors affecting the action performance.

[0013] The key factors are marked in the action completion process, and the user understands the reasons affecting the action performance by combining the key factors, thereby realizing visual analysis of sports action performance.

[0014] Preferably, the completion process of different action completion forms of a specific sports action is displayed for the user by using augmented reality technology, which includes: using augmented reality technology to build a display scene, for a specific sports action, there are different action completion forms, the completion processes of different action completion forms are displayed respectively, the display method is to select multiple typical postures for each completion process, divide the typical postures into different groups according to different action forms, and arrange and display the actions in the form of a small multiple figure of a reduced human model to avoid occlusion between actions.

[0015] Preferably, for each posture, the user can select a single posture and enlarge it to the size of a real person, and can move and rotate the posture to observe the action details from different angles.

[0016] Preferably, the overall action performance is the data summary of several action completion processes of the same action performer at different times or different action performers, the summary data is displayed in the form of a column chart to show the action performance of this action completion form, the horizontal coordinate of the column chart is a plurality of performance ranges of this completion form according to the size of the performance value, and the vertical coordinate is the percentage of the data in this performance range, and a column chart is made for each action completion form to realize the display of the overall action performance.

[0017] Preferably, the movement of different body parts at different time periods in the completion process of any action completion form is displayed for the user, which includes: the user selects any posture in any action completion form, the posture is presented in front of the user in the form of a virtual human model with the size of a real person, the user observes the movement of the specific body part at the selected time period by clicking the specific body part, and the movement process of different body parts and the change process of action attributes in the time period are drawn in the three-dimensional trajectory of the body part. The action attributes include spatial position, speed, acceleration, angular velocity and angular acceleration.

[0018] Preferably, the action attributes are correlated with the action performance by using the multi-device cooperation mode to find the key factors affecting the action performance, including:

[0019] The enhanced reality technology is used to support the construction of a visual interactive scene, and a handheld tablet device is used to support the presentation of interactive data. A user selects any action attribute corresponding to any body part at any time by interacting with the handheld tablet. The user can also select an action attribute by interacting with a virtual human model in the augmented reality environment.

[0020] The action attributes and action performance are correlated by constructing a correlation matrix. The horizontal axis of the correlation matrix represents time, and each row of the vertical axis represents the correlation coefficient between a single attribute of a single body part and the action performance during the completion process. Different colors in different cells of the correlation matrix represent the strength of the correlation.

[0021] The handheld tablet is used to display the correlation matrix. A user clicks on any cell in the correlation matrix on the handheld tablet to present the posture of the virtual human represented by the time corresponding to the cell in the augmented reality environment, as well as the distribution of the corresponding action attribute and action performance. The key factors affecting the action performance are recorded.

[0022] Preferably, the correlation coefficient is calculated based on a set of aligned action data extracted from the completion processes of the same action performer at different times or different action performers by using the dynamic time warping algorithm. The correlation between different action attributes and action performance of different body parts at different times is calculated based on the set of aligned action data using the Pearson correlation coefficient.

[0023] Preferably, the distribution of the action attribute and action performance is represented by colored dots drawn on a three-dimensional coordinate system. The three coordinate axes in the three-dimensional coordinate system represent the components of each action attribute in three directions. The farther the dot is from the origin of the coordinate system, the larger the value of the action attribute. The darker the color of the dot, the better the action performance. By analyzing the correlation coefficient and observing the time when the corresponding body part and action attribute occur, the time and action attribute affecting the action performance are marked as characteristic values. The body part, time, and action attribute are recorded as influencing factors of the action performance.

[0024] Preferably, the key factors are marked during the action completion process. The key attributes of the key parts are marked during the action completion process. All key parts are connected according to the time of action occurrence to show the changes of the key parts during the entire completion process to the user.

[0025] In a second aspect, to achieve the above-mentioned object, the embodiment of the present application further provides a sports action performance visual analysis system based on immersion, comprising an action overview module, an action evolution module and an action demonstration module.

[0026] The action overview module is configured to demonstrate the completion process of different action completion forms of a specific sports action and the overall action performance of the specific sports action to a user by using augmented reality technology.

[0027] The action evolution module is configured to demonstrate the movement of different body parts at different time periods in the completion process of any action completion form to the user, and to find the key factors affecting the action performance by using a multi-device collaborative method to construct a correlation analysis between action attributes and action performance.

[0028] The action demonstration module is configured to mark the key factors in the completion process of the action, and the user can understand the reasons affecting the action performance by combining the key factors, thereby realizing the sports action performance visual analysis.

[0029] Compared with the prior art, the present application has at least the following beneficial effects:

[0030] (1) The present application uses augmented reality technology to perform three-dimensional visualization on sports actions, and intuitively demonstrates the sports action performance in an immersive scene, thereby enhancing the interactive experience of the user and reducing the cognitive burden of the user.

[0031] (2) The completion process of a single action completion form is presented in the form of a gesture sequence, and they are sequentially arranged according to the occurrence time, and a correlation analysis between action attributes and action performance is constructed by using a multi-device collaborative method, thereby supporting the user to observe the specific sports action gesture while retaining efficient statistical analysis, determining the factors affecting the action performance, and contributing to sports visual analysis and related action immersive visualization. BRIEF DESCRIPTION OF DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of these drawings.

[0033] Figure 1 is a flowchart of a sports action performance visual analysis method based on immersion provided by the embodiment of the present application;

[0034] Figure 2 is a distribution diagram of each body part of a human body provided by the embodiment of the present application;

[0035] Figure 3 is an effect diagram of the action overview module in the system provided by the embodiment of the present application;

[0036] Figure 4 is an effect diagram of the action evolution module in the system provided by the embodiment of the present application;

[0037] Figure 5 is a three-dimensional coordinate system diagram of the distribution of action attributes and action performances in the action evolution module provided by the embodiment of the present application;

[0038] Figure 6 is an effect diagram of the action demonstration module in the system provided by the embodiment of the present application;

[0039] Figure 7 is a structural diagram of a sports action performance visual analysis system based on immersion provided by the embodiment of the present application. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the protection scope of the present application.

[0041] The inventive concept of the present application is that, in view of the fact that the sports action performance analysis in the prior art lacks efficient research on the correlation between motion attributes and motion performances, the embodiment of the present application provides a sports action performance visual analysis method and system based on immersion, uses augmented reality technology to build an immersive interactive scene, constructs correlation analysis between action attributes and action performances, finds factors affecting action performances and mechanisms for producing effects, and realizes immersive and efficient sports action performance visual analysis.

[0042] Figure 1 is a flowchart of a sports action performance visual analysis method based on immersion provided by the embodiment of the present application. As shown in Figure 1 The embodiment provides a sports action performance visual analysis method based on immersion, which comprises the following steps:

[0043] S1, using augmented reality technology to show the user the completion process of different action completion forms of a specific sports action and the overall action performance.

[0044] Specifically, augmented reality technology is used to construct display scenarios. For a specific sports movement, there are different ways to complete the movement. The completion process of each different movement is displayed separately. The display method is to select multiple typical postures for each completion process, and arrange the typical postures into different groups according to different movement forms. When displaying the movements, they are arranged in the form of miniaturized human models at multiple magnifications to avoid occlusion between movements. For each posture, users can select a single posture and zoom in to life-size, and can move and rotate the posture to observe the movement details from different angles.

[0045] This system aggregates data on the completion of several actions performed by the same person at different times or by different persons for any given action type. The aggregated data is then displayed as a bar chart to illustrate the performance of this action type, including aspects such as shot speed and jump height. The horizontal axis of the bar chart represents several performance ranges for this action type based on performance value, while the vertical axis represents the percentage of data within each performance range. A bar chart is created for each action type to provide a comprehensive overview of overall action performance. Users can compare different completion processes and, by combining the performance data, identify differences between actions with varying performance.

[0046] In this embodiment, after determining the actions to be analyzed and completing the clustering of the action dataset, the system will visualize action data with different completion forms. Each row in the action display layout represents one action completion form, such as... Figure 3 As shown, Form 1, Form 2, and Form 3 represent three different ways of completing a badminton smash. Different colors distinguish the different groups. For each form, a desktop-sized virtual character is used to visualize the details of five typical postures during the completion process, arranged from left to right according to the time of each posture. A bar chart showing the performance of each group is displayed on the far right. The bar chart represents the hitting speed. The horizontal axis shows different performance ranges, and the vertical axis represents the percentage of performance. Form 3 shows the best performance, achieving the highest percentage of performance at the end of the process, indicating that Form 3 achieved a smash at that hitting speed at the end of the completion process.

[0047] S2 displays the movement of different body parts at different times during the completion of any action in front of the user. It uses a multi-device collaborative approach to build a correlation analysis between action attributes and action performance to find the key factors affecting action performance.

[0048] Specifically, users can select any pose from any action form, and that pose will be presented to the user as a life-size virtual human model. Users can then click on specific body parts to observe their movement over a selected time period. The distribution of body parts is as follows: Figure 2 As shown, the motion process and changes in action attributes of different body parts during this time period will be plotted in the three-dimensional trajectory of that body part. Action attributes include spatial position, velocity, acceleration, angular velocity, and angular acceleration.

[0049] Augmented reality technology and corresponding equipment support the construction of visual interactive scenes. Handheld tablets support the presentation of interactive data. Users can interact with the handheld tablet to select any action attribute corresponding to any body part at any time. Users can also interact with virtual human models in the augmented reality environment to select action attributes.

[0050] A correlation matrix was constructed to analyze the relationship between action attributes and action performance. The horizontal axis of the correlation matrix represents time, and each row on the vertical axis represents the correlation coefficient between a single attribute of a single body part and its action performance during the completion process. Different colors in the cells of the correlation matrix represent the strength of the correlation. The correlation coefficients were obtained by extracting aligned action data from the completion processes of several actions performed by the same actor at different times or by different actors using a dynamic time-warping algorithm. Based on this aligned action data, the Pearson correlation coefficient was used to calculate the correlation between different action attributes of different body parts and their action performance at different times.

[0051] Using a handheld tablet to display a correlation matrix, users can click on any cell in the matrix to view the posture of a virtual human at the corresponding time in an augmented reality environment, along with the distribution of corresponding action attributes and performance. Key factors influencing action performance are recorded. The distribution of action attributes and performance is represented by colored dots drawn on a three-dimensional coordinate system. The three axes of this system represent the components of each action attribute in three directions. The farther the dot is from the origin, the larger the value of the action attribute; the darker the dot, the better the performance. By analyzing correlation coefficients and observing the timing of corresponding body parts and action attributes, the time and action attributes affecting performance are marked as feature values, and body parts, time, and action attributes are recorded as influencing factors on action performance.

[0052] In the embodiments, such as Figure 4 As shown, the user selects as follows Figure 3 Observe one of the actions and poses in Form 3, which will be presented to the user as a life-size virtual human model, such as... Figure 4As shown in (A), the user clicks the right hand to observe the right hand activity at this stage, the three-dimensional trajectory of the right hand describes the movement process of the right hand and the change process of the action attribute in this time period, and the change of the color in the three-dimensional trajectory intuitively shows the change of the action attribute; as shown in (B), the three-dimensional trajectory of the right hand describes the movement process of the right hand and the change process of the action attribute in this time period, and the change of the color in the three-dimensional trajectory intuitively shows the change of the action attribute; as shown in (C), the hand-held tablet shows the correlation matrix of the action attribute and the action performance of the body part, and the deeper the color of each cell in the correlation matrix, the greater the correlation; the user clicks the cell in the tablet, and the scene as shown in (D) is presented, the three-dimensional coordinates show the value of the movement attribute of the right shoulder corresponding to the cell and the distribution of the action performance of the record, Figure 4 the correlation matrix of the action attribute and the action performance of the body part, and the deeper the color of each cell in the correlation matrix, the greater the correlation; the user clicks the cell in the tablet, and the scene as shown in (D) is presented, the three-dimensional coordinates show the value of the movement attribute of the right shoulder corresponding to the cell and the distribution of the action performance of the record, Figure 4 the correlation matrix of the action attribute and the action performance of the body part, and the deeper the color of each cell in the correlation matrix, the greater the correlation; the user clicks the cell in the tablet, and the scene as shown in (D) is presented, the three-dimensional coordinates show the value of the movement attribute of the right shoulder corresponding to the cell and the distribution of the action performance of the record, Figure 5 the correlation matrix of the action attribute and the action performance of the body part, and the deeper the color of each cell in the correlation matrix, the greater the correlation; the user clicks the cell in the tablet, and the scene as shown in (D) is presented, the three-dimensional coordinates show the value of the movement attribute of the right shoulder corresponding to the cell and the distribution of the action performance of the record, Figure 4 the correlation matrix of the action attribute and the action performance of the body part, and the deeper the color of each cell in the correlation matrix, the greater the correlation; the user clicks the cell in the tablet, and the scene as shown in (D) is presented, the three-dimensional coordinates show the value of the movement attribute of the right shoulder corresponding to the cell and the distribution of the action performance of the record,

[0053] S3, mark the key factors in the action completion process, and the user understands the reasons affecting the action performance by combining the key factors, to realize the visual analysis of sports action performance.

[0054] Specifically, the key attributes of the key parts are marked in the action completion process, all the key parts are connected according to the time of the action, and the user observes the time when these influencing factors appear, sees the changes of different body parts affecting the action performance over time in the whole completion process, and understands the reasons affecting the action performance and the realization mechanism of the action by combining the context information and the principle of sports biomechanics.

[0055] In the embodiment, as shown in (A), the completion process of a single completion form of the badminton smash action is shown, these influencing factors are connected by lines, and the changes of the key parts at each time in the completion cycle of the action are statically or dynamically displayed, and the rightmost side shows a histogram, and the action performance percentage is the highest when the time range is the largest in the histogram, indicating that the overall performance of the action is better. Figure 6

[0056] In summary, a kind of visual analysis method of sports action performance based on immersion, three-dimensional visualization of sports action is realized by using augmented reality technology to enhance the user interaction experience, the correlation analysis between action attribute and action performance is constructed by using the way of multi-device cooperation, and the factors affecting the action performance are determined, and the efficient visual analysis of sports action performance is realized by combining visualization and statistical analysis.

[0057] ​Based on the same inventive concept, the embodiment also provides an immersive-based sports action performance visual analysis system 700, as shown in the figure, comprising an action overview module 710, an action evolution module 720 and an action demonstration module 730. Figure 7

[0058] The action overview module 710 is configured to use augmented reality technology to show the user the completion process of different action completion forms of a specific sports action, and the overall action performance.

[0059] The action evolution module 720 is configured to show the user the movement of different body parts at different time periods in the completion process of any action completion form, and to use a multi-device collaborative method to construct a correlation analysis of action attributes and action performance, and to find the key factors affecting the action performance.

[0060] The action demonstration module 730 is configured to mark the key factors in the action completion process, and the user combines the key factors to understand the reasons affecting the action performance, and realizes the visual analysis of sports action performance.

[0061] It should be noted that the above embodiment provides an immersive-based sports action performance visual analysis system, which belongs to the same concept as an immersive-based sports action performance visual analysis method embodiment. The specific implementation process is described in detail in the embodiment of the immersive-based sports action performance visual analysis method. Here, it will not be repeated.

[0062] The specific embodiments described above describe the technical solutions and advantages of the present application. It should be understood that the above description is only the most preferred embodiment of the present application, and is not intended to limit the present application. Any modifications, supplements and equivalent replacements made within the scope of the principles of the present application shall be included in the protection scope of the present application.​

Claims

1. An immersion-based visual analysis method for sports action performance, characterized in that, The method comprises the following steps: The user is shown the completion process of different action completion forms of a specific sports action by using augmented reality technology, and the overall action performance is shown. The movement of different body parts at different time periods in the completion process of any action completion form is shown in front of the user, and the action attributes and action performance are associated and analyzed in a multi-device cooperative manner to find the key factors affecting the action performance, including: The construction of a visual interactive scene is supported by using the corresponding device of augmented reality technology, the presentation of interactive data is supported by using a handheld tablet device, the user selects any action attribute corresponding to any time and any body part by interacting with the handheld tablet, and the user can also select the action attribute by interacting with the virtual human model in the augmented reality environment. The action attributes and action performance are associated and analyzed by constructing a correlation matrix, the horizontal axis of the correlation matrix represents time, each row of the vertical axis represents the correlation coefficient between a single attribute of a single body part and the action performance during the completion process, and different colors in different cells of the correlation matrix represent the strength of the correlation. The correlation matrix is displayed by using the handheld tablet, the user clicks any cell in the correlation matrix of the handheld tablet, the posture of the virtual human represented by the corresponding time in the augmented reality environment is presented, and the distribution of the corresponding action attribute and action performance is presented, and the key factors affecting the action performance are recorded. The key factors are marked in the action completion process, the user understands the reasons affecting the action performance in combination with the key factors, and visual analysis of the sports action performance is realized.

2. The immersion-based sports action performance visual analysis method of claim 1, wherein, The user is shown the completion process of different action completion forms of a specific sports action by using augmented reality technology, and the overall action performance is shown. The display scene is built by using augmented reality technology, different action completion forms exist for a specific sports action, the completion processes of different action completion forms are respectively shown, the typical postures of each completion process are selected, the typical postures are arranged in different groups according to different action forms, and the actions are arranged and displayed in the form of a small multiple of a human model to avoid occlusion between actions.

3. The immersion-based sports action performance visual analysis method of claim 2, wherein, For each posture, the user can select a single posture and enlarge it to the size of a real person, and can move and rotate the posture to observe the action details from different angles.

4. The immersion-based sports action performance visual analysis method of claim 1, wherein, The overall action performance is shown by data collection of a plurality of action completion processes of the same action performer at different times or different action performers, the collected data is shown in the form of a column chart to show the action performance of this action completion form, the horizontal axis of the column chart is a plurality of performance ranges of this completion form according to the size of the performance value, and the vertical axis is the percentage of the data distributed in the performance range, and a column chart is made for each action completion form to realize the display of the overall action performance.

5. The immersion-based sports action performance visual analysis method of claim 1, wherein, The movement of different body parts at different time periods in the completion process of any action completion form is shown in front of the user, and the action attributes and action performance are associated and analyzed in a multi-device cooperative manner to find the key factors affecting the action performance, including: The user selects any gesture in any action completion form, and a virtual human model of real size is presented in front of the user. The user observes the movement of a specific body part at a selected time period by clicking the specific body part. The movement process and the change process of action attributes of different body parts in the time period are plotted in the three-dimensional trajectory of the body part. The action attributes include spatial position, velocity, acceleration, angular velocity, and angular acceleration.

6. The immersion-based sports action performance visual analysis method of claim 1, wherein, The correlation coefficient is obtained by using the dynamic time warping algorithm to extract a set of aligned action data from the action completion processes of the same action performer at different times or different action performers. The correlation between different action attributes of different body parts at different times and action performance is calculated based on the set of aligned action data using the Pearson correlation coefficient.

7. The immersion-based sports action performance visual analysis method of claim 1, wherein, The distribution of the action attributes and the action performance is represented by colored dots plotted on a three-dimensional coordinate system. The three coordinate axes in the three-dimensional coordinate system represent the components of each action attribute in three directions. The farther the dot is from the origin of the coordinate system, the greater the value of the action attribute. The darker the color of the dot, the better the action performance. By analyzing the correlation coefficient and observing the time when the corresponding body part and action attribute occur, the time and action attribute that affect the action performance are marked as characteristic values. The body part, time, and action attribute are recorded as influencing factors of the action performance.

8. The immersion-based sports action performance visual analysis method of claim 1, wherein, The key factors are marked in the action completion process. All key parts are connected according to the time of action occurrence, and the user is shown the changes of the key parts in the entire completion process.

9. An immersive-based sports action performance visual analysis system, comprising an action overview module, an action evolution module, and an action demonstration module; The action overview module is used to show the user the completion process of different action completion forms of a specific sports action and the overall action performance using augmented reality technology. The action evolution module is used to show the user the movement of different body parts at different time periods in the completion process of any action completion form. The action attributes and action performance are correlated and analyzed using a multi-device collaborative method to find the key factors that affect the action performance, including: The corresponding device of the augmented reality technology supports the construction of a visual interactive scene. The handheld tablet device supports the presentation of interactive data. The user selects any action attribute corresponding to any body part at any time by interacting with the handheld tablet. The user can also select the action attribute by interacting with the virtual human model in the augmented reality environment. The action attributes and action performance are correlated and analyzed by constructing a correlation matrix. The horizontal coordinate of the correlation matrix represents time, and each row of the vertical coordinate represents the correlation coefficient between a single attribute of a single body part and the action performance in the completion process. Different colors in different cells of the correlation matrix represent the strength of the correlation. The handheld tablet is used to display the correlation matrix, and any cell in the correlation matrix is clicked to present the posture of the virtual person at the corresponding time, and the distribution of the corresponding action attribute and action performance in the augmented reality environment, and the key factors affecting the action performance are recorded; The action demonstration module is used to mark the key factors in the action completion process, and the user combines the key factors to understand the reasons affecting the action performance, and realizes the visual analysis of the sports action performance.

Images