A long jump analysis and action correction system and method based on artificial intelligence

Abstract

The application discloses a long jump analysis and action correction system and method based on artificial intelligence, belongs to the technical field of artificial intelligence, and comprises a background processing module, which is used for acquiring complete long jump videos of users, performing segmentation processing on target long jump videos, and obtaining frame images processed; an identification and extraction module, which is used for identifying the frame images processed based on a human body bone point identification model, extracting a plurality of key frame images from the frame images in which bone points are identified; an analysis and calculation module, which is used for performing action qualification degree analysis according to the coordinates of the bone points in the plurality of key frame images, and generating an action reference score and a prompt according to an analysis result; and a scoring module, which is used for calculating an actual long jump distance and generating a long jump reference score in combination with a long jump standard score table. The application solves the problems that professional guidance in long jump practice can only be provided by special persons in specific places, and the operation is relatively inconvenient and low in efficiency.

Description

Technical Field

[0001] This invention belongs to the field of artificial intelligence technology, specifically relating to an artificial intelligence-based long jump analysis and motion correction system and method. Background Technology

[0002] In recent years, the Ministry of Education has vigorously promoted comprehensive quality education, attaching great importance to students' overall quality. Physical education has become increasingly important, reflected in the significantly increased proportion of physical education scores in the overall grade at all stages. This increased proportion is crucial for physical health, all-round development, competitive spirit, and future education and employment. Participating in physical education exams can effectively improve physical fitness and enhance physical health. Furthermore, physical education exams not only assess athletic skills but also physical literacy and comprehensive abilities such as coordination, flexibility, and endurance, contributing to the all-round development of one's physical and intellectual abilities. Participation in physical education exams can also cultivate a competitive spirit and ability, laying a foundation for future learning and life.

[0003] The standing long jump is an important event in sports, and its importance is self-evident. With professional guidance and systematic training, one's performance can be significantly improved, leading to better results in physical education exams. However, currently, professional guidance in long jump training can only be provided by specialists in specific venues, which is inconvenient and inefficient.

[0004] Therefore, to address the problems existing in the current technology, there is an urgent need to provide an artificial intelligence-based long jump analysis and motion correction system and method to overcome the aforementioned shortcomings of the existing technology. Summary of the Invention

[0005] The purpose of this invention is to provide an artificial intelligence-based long jump analysis and movement correction system and method to solve the problem that providing professional guidance in the prior art can only be done in a specific venue by a specialist, which is inconvenient and inefficient.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a long jump analysis and motion correction system based on artificial intelligence, including a background processing module, a recognition and extraction module, an analysis and calculation module, and a scoring module; The background processing module is used to receive the user's complete long jump video, detect whether there is a target long jump video at preset time intervals, and if there is, segment and process the target long jump video to obtain the processed frame image. The identification and extraction module is connected to the background processing module. Based on the human skeleton key point recognition model, it identifies the processed frame image and extracts multiple key frame images from the frame image with identified skeleton key points. The analysis and calculation module is connected to the recognition and extraction module. It is used to analyze the qualification of the action based on the coordinates of the skeletal key points in the multiple keyframe images, and generate an action reference score and prompts based on the analysis results. The scoring module is connected to the analysis and calculation module, and is used to calculate the actual long jump distance and generate a long jump reference score by combining it with the long jump standard score table.

[0007] Based on the aforementioned publicly available information, the backend processing module obtains the user's complete long jump video and, together with the analysis and calculation module and the scoring module, processes the long jump video. Multiple keyframe images are extracted from the video, and the qualification level of the action is judged through coordinate analysis and trigonometric function calculations on the multiple keyframe images. Corresponding prompts and action reference scores are given. The actual long jump distance is calculated based on the long jump distance in the keyframe images, and a long jump reference score is generated by combining it with the score table. There is no need for a special person to guide the user in a specific venue. The entire long jump process can be recorded by external photography equipment and uploaded to the long jump analysis and action correction system to obtain professional guidance and corresponding scores.

[0008] As a further technical solution of the present invention, it also includes a video acquisition module, which is used to acquire the user's complete long jump video.

[0009] As a further technical solution of the present invention, it also includes a front-end display module, which includes a chart processing submodule and a data display submodule; The chart processing submodule is connected to the scoring module, and it is used to calculate the actual long jump distance and the reference score, and to create a chart. The data display submodule is connected to the chart processing submodule and is used to display analysis videos, reference scores, and charts on the front end.

[0010] As a further technical solution of the present invention, it also includes a database for storing the long jump standard score table, user information, analysis videos and charts, and the data in the database is updated in real time, and the database is connected to the chart processing submodule.

[0011] As a further technical solution of the present invention, it also includes an administrator module for managing user information, user login, long jump standard score table and / or system access permissions.

[0012] As a further technical solution of the present invention, it also includes an external photography device for capturing the user's complete long jump video.

[0013] Secondly, the present invention provides a long jump analysis and motion correction method based on artificial intelligence, including: Receive the user's complete long jump video; The long jump video is segmented to obtain processed frame images; The processed frame images are identified based on the human skeleton key point recognition model, and multiple key frame images are extracted from the frame images in which the skeleton key points are identified. Based on the coordinates of key points of the skeleton in the multiple keyframe images, the passability of the action is analyzed, and a reference score and prompts are generated based on the analysis results. Calculate the actual long jump distance and generate a reference score for the long jump by combining it with the standard long jump score table.

[0014] Based on the above disclosure, this invention segments and processes the frame images obtained from the long jump video, identifies them based on the human skeleton key point recognition model, extracts multiple key frame images, analyzes the coordinates of key points in different parts of the key frame images, judges the qualification of the action, gives the corresponding reference score and prompts, scores according to the actual long jump distance and the long jump standard score table, and provides users with guidance on non-standard actions and the score obtained for that long jump.

[0015] As a further technical solution of the present invention, the long jump video is segmented to obtain processed frame images, including: Read the long jump video; The long jump video is segmented into a series of consecutive frames; Frame images are extracted frame by frame from each of the consecutive frames; The frame image is preprocessed and saved as a processed frame image.

[0016] As a further technical solution of the present invention, the processed frame image is identified based on a human skeletal key point recognition model, and multiple key frame images are extracted from the frame image with identified skeletal key points, including: Based on the keypointrcnn_resnet50_fpn() model in PyTorch, skeletal keypoints were identified in the processed frame image. The skeletal keypoints include nose keypoint, eye keypoint, ear keypoint, shoulder keypoint, elbow keypoint, wrist keypoint, hip keypoint, knee keypoint, and ankle keypoint. In the processed frame images before the movement of the ankle key point, when the y-axis coordinate value of the wrist key point is the minimum, the frame image corresponding to the minimum y-axis coordinate value of the wrist key point is extracted as the first key frame image; In the processed frame images before the movement of the ankle key point, at the last time when the difference between the y-axis coordinate values of the wrist key point and the knee key point is the maximum, the frame image corresponding to the last time when the difference between the y-axis coordinate values of the wrist key point and the knee key point is the maximum is extracted as the second key frame image; In the processed frame images before the movement of the ankle key point, when the included angle between the line connecting the wrist key point to the ankle key point and the x-axis is the minimum, the frame image corresponding to the minimum included angle between the line connecting the wrist key point to the ankle key point and the x-axis is extracted as the third key frame image; In the processed frame images after the movement of the ankle key point, when the y-axis coordinate value of the wrist key point is the minimum, the frame image corresponding to the minimum y-axis coordinate value of the wrist key point is extracted as the fourth key frame image; In the processed frame images after the movement of the ankle key point, when the y-axis coordinate value of the ankle key point is equal to the y-axis coordinate value of the ankle key point before the movement of the ankle key point, the frame image corresponding to the equality between the y-axis coordinate value of the ankle key point and the y-axis coordinate value of the ankle key point before the movement of the ankle key point is extracted as the fifth key frame image.

[0017] As a still further technical solution of the present invention, the qualification degree of the movement is analyzed according to the coordinate calculation of the bone key points in the plurality of key frame images, a movement reference score is generated and a prompt is given, comprising: Calculating the absolute values of the coordinate differences in the x-axis direction between every two of the wrist key point, shoulder key point, hip key point, knee key point and ankle key point in the first key frame image, if the value obtained by subtracting the first error value from all the absolute values of the coordinate differences is less than a first preset value, the movement is qualified, a full score is given and a prompt is provided at the same time; otherwise, a prompt is given, and corresponding scores are deducted according to the absolute values of the coordinate differences in the x-axis direction between every two of the wrist key point, shoulder key point, hip key point, knee key point and ankle key point; Judging the x-axis coordinate values of the wrist key point, knee key point, ear key point and ankle key point in the second key frame image, if they are arranged in sequence according to the magnitude of the x-axis coordinate values and the order is the wrist key point, the ankle key point, the knee key point and the ear key point, the movement is qualified, a full score is given and a prompt is provided at the same time; otherwise, a prompt is given, and corresponding scores are deducted according to the difference in the order among the wrist key point, ankle key point, knee key point and ear key point; The third keyframe image is used to calculate the angle between the line connecting the wrist key point to the ankle key point and the x-axis. If the angle is 45 degrees, the action is considered qualified, and full marks are given, along with a prompt. Otherwise, a prompt is given, and the corresponding score is deducted based on the absolute value of the difference between the angle and 45 degrees. Calculate the y-axis coordinate difference between the knee key point in the fourth keyframe image and the knee key point in the second keyframe image to obtain the first coordinate difference. If the first coordinate difference is greater than the second preset value, the action is qualified, and full marks are given along with a prompt. If the first coordinate difference is less than or equal to the second preset value, the corresponding score is deducted based on the absolute value of the difference between the second preset value and the first coordinate difference, and a prompt is given. Calculate the difference in x-axis coordinates between the ankle keypoint in the fifth keyframe image and the ankle keypoint in the first keyframe image to obtain the second coordinate difference. If the second coordinate difference is greater than the third preset value, the action is qualified, and full marks are given along with a prompt. If the second coordinate difference is less than or equal to the third preset value, the corresponding score is deducted based on the absolute value of the difference between the third preset value and the second coordinate difference, and a prompt is given.

[0018] As a further technical solution of the present invention, the actual long jump distance is calculated and scored in conjunction with the long jump standard score table, including: The actual long jump distance is calculated using the following formula. :

[0019] in, This is the actual long jump distance; For the user's height; The value of the y-axis coordinate of the eye keypoint in the first keyframe image; The y-axis coordinate value of the ankle key point in the first keyframe image; The x-axis coordinate value of the ankle key point in the first keyframe image; The x-axis coordinate value of the ankle key point in the fifth keyframe image; Based on the actual long jump distance The long jump reference score is calculated by combining the value with the long jump standard score table.

[0020] As a further technical solution of the present invention, before receiving the complete long jump video from the user, it further includes: Enter the standard long jump score table; Enter user information, which includes name, class, height, student ID, and gender.

[0021] Beneficial effects: (1) The background processing module obtains the user's complete long jump video and processes the long jump video together with the analysis and calculation module and the scoring module. It extracts multiple key frame images from the video, and judges the qualification of the action by coordinate analysis and trigonometric function calculation of multiple key frame images, and gives corresponding prompts and action reference scores. The actual long jump distance is calculated based on the long jump distance in the key frame images and the long jump reference score is generated by combining the score table. There is no need for a special person to guide the long jump in a specific venue. The entire long jump process can be recorded by external photography equipment and uploaded to the long jump analysis and action correction system to obtain professional guidance and provide corresponding scores.

[0022] (2) The present invention segments and processes the frame images obtained from the long jump video and identifies them based on the human skeleton key point recognition model, extracts multiple key frame images, analyzes the coordinates of key points in different parts of the key frame images, judges the qualification of the action, gives the corresponding reference score and prompts, scores according to the actual long jump distance and the corresponding long jump standard score table, and provides users with guidance on non-standard actions and the score obtained in this long jump.

[0023] (3) Users can use only an external camera and an AI-based long jump analysis and movement correction system to know the shortcomings of their long jump movements, how to improve them, and receive professional guidance. Furthermore, by viewing the charts displayed on the front end, users can more clearly and intuitively understand their practice results and progress range. The system is easy to operate, saves time, and improves efficiency. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 A schematic diagram of the structure of an artificial intelligence-based long jump analysis and motion correction system provided by the present invention; Figure 2 This is a flowchart illustrating an artificial intelligence-based long jump analysis and motion correction method provided by the present invention. Detailed Implementation

[0026] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. It should be noted that while the description of these embodiments is intended to aid in understanding the invention, it does not constitute a limitation thereof. The specific structural and functional details disclosed herein are only for describing exemplary embodiments of the invention. However, the invention can be embodied in many alternative forms and should not be construed as being limited to the embodiments described herein.

[0027] It should be understood that although the terms "first" and "second", etc., may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another. For example, the first object may be referred to as the second object, and similarly, the second object may be referred to as the first object, without departing from the scope of the exemplary embodiments of the invention.

[0028] It should be understood that the term "and / or" that may appear in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, B exists alone, or A and B exist simultaneously. The term " / and" that may appear in this document describes another relationship between related objects, indicating that two relationships can exist. For example, A / and B can mean: A exists alone or A and B exist simultaneously. In addition, the character " / " that may appear in this document generally indicates that the related objects before and after it are in an "or" relationship.

[0029] Example Firstly, such as Figure 1 As shown, this embodiment provides an artificial intelligence-based long jump analysis and motion correction system, including but not limited to a background processing module, an identification and extraction module, an analysis and calculation module, and a scoring module. The background processing module acquires the user's complete long jump video, checks for target long jump videos at preset time intervals, and if a target long jump video is found, segments and processes it to obtain processed frame images. The identification and extraction module, connected to the background processing module, identifies the processed frame images based on a human skeletal key point recognition model, extracting multiple key frame images from the identified skeletal key points. The analysis and calculation module, connected to the identification and extraction module, analyzes the motion qualification level based on the coordinates of the skeletal key points in the multiple key frame images, and generates a motion reference score and prompts based on the analysis results. The scoring module, connected to the analysis and calculation module, calculates the actual long jump distance and generates a long jump reference score by combining it with a long jump standard score table.

[0030] Specifically, the backend processing module uses asynchronous operation to process long jump videos. When users are analyzing long jump videos, they do not need to wait for the upload task to complete on the page. They can perform other tasks while the video is being uploaded, without blocking or delaying the video upload progress.

[0031] In one possible design, the prompts would be recorded into the target long jump video to generate an analysis video for easy viewing by the user.

[0032] In one feasible implementation, a video acquisition module is also included, which is used to acquire a complete video of the user's long jump.

[0033] Furthermore, external devices for capturing long jump videos include, but are not limited to, mobile phones, camcorders, and digital cameras. When using external photography equipment to capture a user's complete long jump video, there are no site requirements or reference points; it is only necessary to shoot at an angle perpendicular to the direction of the long jump.

[0034] In one feasible implementation, a front-end display module is also included, which comprises a chart processing submodule and a data display submodule; The chart processing submodule is connected to the scoring module, and it is used to calculate the actual long jump distance and the reference score, and to create a chart. The data display submodule is connected to the chart processing submodule and is used to display analysis videos, reference scores, and charts on the front end.

[0035] The chart processing submodule generates line charts and bar charts based on the actual long jump distance calculated in the scoring module, combined with reference scores generated from the gender and long jump standard score tables. The data display submodule displays the completed analysis video, reference scores, and line and bar charts on the front-end page, making it easier for users to understand their progress trend.

[0036] In one feasible implementation, a database is also included for storing the long jump standard score table, user information, analysis videos and charts, and the data in the database is updated in real time, and the database is connected to the chart processing submodule.

[0037] Specifically, standard score tables for the long jump include, but are not limited to, score tables for primary schools, middle schools, universities, and sports schools.

[0038] In one possible design, the database is designed using a MySQL database, containing five tables: b_student, b_type, b_video, t_user_audit_hash, and user. The b_student table is the user information table, containing information such as name, class, student ID, height, creation date, and gender. The b_video table is the user long jump analysis table, including analysis videos, long jump distances, and scores, all of which can be queried on the front end. Linear charts are also generated to allow users to see their progress and facilitate subsequent queries.

[0039] In one feasible implementation, an administrator module is also included for managing user information, long jump standard score tables, and user login and / or system access permissions.

[0040] It should be noted that users have the authority to view their own information, change their passwords, upload long jump video analyses, and view their own analysis videos, reference scores, and charts. Administrators can manage user information, reset user passwords, and update the long jump standard score table in the system according to the different long jump standard score tables each year.

[0041] Secondly, such as Figure 2 As shown, this embodiment provides a long jump analysis and motion correction method based on artificial intelligence, including but not limited to steps S101-S105: Step S101. Receive the user's complete long jump video; It should be noted that the external devices used to collect long jump videos in this embodiment include, but are not limited to, mobile phones, video cameras, and digital cameras. There are no requirements for the venue or reference objects during the collection process.

[0042] In one possible design, before step S101, acquiring the user's complete long jump video, the following is also included: Step S1011. Enter the standard long jump score table; Step S1012. Enter user information, which includes name, class, height, student ID and gender.

[0043] Based on the publicly available information, the administrator needs to enter the standard long jump score table for that year and import the user's relevant information, including name, class, height, student ID, and gender.

[0044] Step S102. Segment and process the long jump video to obtain processed frame images; As one possible design for step S102, the long jump video is segmented to obtain processed frame images, including but not limited to: Step S1021. Read the long jump video; Step S1022. Segment the long jump video into a series of consecutive frames; Step S1023. Extract frame images frame by frame from each of the consecutive frames; Step S1024. Perform preprocessing operations on the frame image and save it as the processed frame image.

[0045] Step S103. Based on the human skeleton key point recognition model, the processed frame image is recognized, and multiple key frame images are extracted from the frame image with recognized skeleton key points; As one possible design for step S103, the processed frame image is identified based on a human skeletal key point recognition model, and multiple key frame images are extracted from the frame images with identified skeletal key points, including but not limited to: Step S1031. Based on the keypointrcnn_resnet50_fpn() model in PyTorch, identify the skeletal key points of the processed frame image, wherein the skeletal key points include the nose key point, eye key point, ear key point, shoulder key point, elbow key point, wrist key point, hip key point, knee key point and ankle key point. Step S1032. Take the upper left corner of the processed frame image as the origin of the coordinate system, the horizontal direction as the x-axis, the vertical direction as the y-axis, the downward direction of the y-axis as the positive direction, and the unit coordinate length as 1 pixel; Step S1033. In the processed frame image before the ankle key point moves, when the y-axis coordinate value of the wrist key point is the smallest, take the frame image corresponding to the minimum y-axis coordinate value of the wrist key point as the first key frame image. Step S1034. In the processed frame image before the ankle key point moves, when the difference between the wrist key point and the knee key point in the y-axis coordinate value is the largest in the last time, the frame image corresponding to the last time the difference between the wrist key point and the knee key point in the y-axis coordinate value is the largest is taken as the second key frame image. Step S1035. In the processed frame image before the ankle key point moves, when the angle between the line connecting the wrist key point to the ankle key point and the x-axis is the smallest, the frame image corresponding to the smallest angle between the line connecting the wrist key point to the ankle key point and the x-axis is taken as the third key frame image. Step S1036. In the processed frame image after the ankle key point has moved, when the y-axis coordinate value of the wrist key point is the smallest, take the frame image corresponding to the minimum y-axis coordinate value of the wrist key point as the fourth key frame image. Step S1037. In the processed frame image after the ankle keypoint has moved, when the y-axis coordinate value of the ankle keypoint is equal to the y-axis coordinate value of the ankle keypoint before the ankle keypoint has moved, the frame image corresponding to the y-axis coordinate value of the ankle keypoint before the ankle keypoint has moved is taken as the fifth keyframe image.

[0046] Step S104. Calculate the coordinates of the key points of the skeleton in the multiple keyframe images, perform an action qualification analysis, and generate an action reference score and prompts based on the analysis results; One possible design for step S104 involves calculating and analyzing the performance level of the action based on the coordinates of key points in the multiple keyframe images, generating an action reference score, and providing prompts, including but not limited to: Step S1041. Calculate the absolute value of the coordinate difference in the x-axis direction between each pair of wrist keypoints, shoulder keypoints, hip keypoints, knee keypoints, and ankle keypoints in the first keyframe image. If the absolute value of all coordinate differences minus the first error value is less than the first preset value, the action is qualified, and full marks are given while a prompt is given; otherwise, a prompt is given, and the corresponding score is deducted according to the absolute value of the coordinate difference in the x-axis direction between each pair of wrist keypoints, shoulder keypoints, hip keypoints, knee keypoints, and ankle keypoints. For example, if the absolute value of the difference between the x-coordinate values ​​of the wrist key point and the shoulder key point minus the error value of 5 pixels is greater than the first preset value, a prompt will be given: "Stretch your hand straight up, stand up straight, and try to align your hand and feet with the red line"; if the absolute value of the difference minus the first error value is less than the first preset value, the action is qualified and a prompt will be given: "You did it very well". The first preset value is entered manually.

[0047] Step S1042. Determine the x-axis coordinates of the wrist keypoint, knee keypoint, ear keypoint, and ankle keypoint in the second keyframe image. If the x-axis coordinates are arranged in order of size, and the sequence is wrist keypoint, ankle keypoint, knee keypoint, and ear keypoint, then the action is qualified, and full marks are given, along with a prompt. Otherwise, a prompt is given, and corresponding points are deducted based on the different order of the wrist keypoint, ankle keypoint, knee keypoint, and ear keypoint. Step S1043. Calculate the angle between the line connecting the wrist key point to the ankle key point in the third keyframe image and the x-axis. If the angle is 45 degrees, the action is qualified and full marks are given, along with a prompt. Otherwise, a prompt is given, and the corresponding score is deducted based on the absolute value of the difference between the angle and 45 degrees. For example, if a movement is not performed correctly, the prompt could be "Your current angle is 69 degrees. Try to align your body with the red guide line." Since the difference between 69 degrees and 45 degrees is 24 degrees, the score is 100 points minus 24 points, resulting in a final score of 76 points.

[0048] Step S1044. Calculate the y-axis coordinate difference between the knee key point in the fourth keyframe image and the knee key point in the second keyframe image to obtain the first coordinate difference. If the first coordinate difference is greater than the second preset value, the action is qualified and full marks are given, and a prompt is given. If the first coordinate difference is less than or equal to the second preset value, the corresponding score is deducted based on the absolute value of the difference between the second preset value and the first coordinate difference, and a prompt is given. Specifically, the second preset value is entered manually.

[0049] Step S1045. Calculate the coordinate difference on the x-axis between the ankle key point in the fifth keyframe image and the ankle key point in the first keyframe image to obtain the second coordinate difference. If the second coordinate difference is greater than the third preset value, the action is qualified, and full marks are given while a prompt is given; if the second coordinate difference is less than or equal to the third preset value, the corresponding score is deducted based on the absolute value of the difference between the third preset value and the second coordinate difference, and a prompt is given.

[0050] Step S105. Calculate the actual long jump distance and score it according to the long jump standard score table; One possible design for step S105 involves calculating the actual long jump distance and scoring it using a standard long jump scoring table, including: The actual long jump distance is calculated using the following formula. :

[0051] in, This is the actual long jump distance; For the user's height; The y-axis coordinate value of the eye keypoint in the first keyframe image; The y-axis coordinate value of the ankle key point in the first keyframe image; The x-axis coordinate value of the ankle key point in the first keyframe image; The x-axis coordinate value of the ankle key point in the fifth keyframe image; Based on the actual long jump distance The long jump reference score is calculated by combining the value with the long jump standard score table.

[0052] In this embodiment, multiple computers with graphics cards are collected, and the graphics cards are converted into GPU computing power through GUDA (General-Purpose GPU Development Accelerator). GUDA provides a programming model and interface that enables developers to utilize the parallel computing capabilities of graphics cards to perform general computing tasks, not just graphics rendering. GUDA is often used in conjunction with programming languages ​​(such as CUDA), which provide tools and libraries for writing parallel computing code, significantly accelerating video analysis and providing higher computational throughput.

[0053] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A long jump analysis and motion correction system based on artificial intelligence, characterized in that, It includes a back-end processing module, an identification and extraction module, an analysis and calculation module, and a scoring module; The background processing module is used to receive the user's complete long jump video, detect whether there is a target long jump video at preset time intervals, and if there is, segment and process the target long jump video to obtain the processed frame image. The identification and extraction module, connected to the backend processing module, identifies the processed frame images based on the human skeletal key point recognition model, and extracts multiple key frame images from the frame images with identified skeletal key points, including: Based on the keypointrcnn_resnet50_fpn() model in PyTorch, skeletal keypoints were identified in the processed frame image. The skeletal keypoints include nose keypoint, eye keypoint, ear keypoint, shoulder keypoint, elbow keypoint, wrist keypoint, hip keypoint, knee keypoint, and ankle keypoint. In the processed frame image before the ankle key point moves, when the y-axis coordinate value of the wrist key point is the smallest, the frame image corresponding to the minimum y-axis coordinate value of the wrist key point is taken as the first key frame image. In the processed frame image before the ankle key point moves, when the difference between the wrist key point and the knee key point on the y-axis coordinate value is the largest, the frame image corresponding to the largest difference between the wrist key point and the knee key point on the y-axis coordinate value is taken as the second key frame image. In the processed frame image before the ankle key point moves, the frame image corresponding to the smallest angle between the line connecting the wrist key point to the ankle key point and the x-axis is the third key frame image. In the processed frame image after the ankle key point has moved, when the y-axis coordinate value of the wrist key point is the smallest, the frame image corresponding to the minimum y-axis coordinate value of the wrist key point is taken as the fourth key frame image. In the processed frame image after the ankle key point moves, when the y-axis coordinate value of the ankle key point is equal to the y-axis coordinate value of the ankle key point before the ankle key point moves, the frame image corresponding to the y-axis coordinate value of the ankle key point before the ankle key point moves is taken as the fifth key frame image. The analysis and calculation module is connected to the recognition and extraction module. It is used to analyze the qualification of the action based on the coordinates of the skeletal key points in the multiple keyframe images, and generate an action reference score and prompts based on the analysis results. The scoring module is connected to the analysis and calculation module, and is used to calculate the actual long jump distance and generate a long jump reference score by combining it with the long jump standard score table.

2. The artificial intelligence-based long jump analysis and motion correction system according to claim 1, characterized in that, It also includes a video acquisition module, which is used to acquire the user's complete long jump video.

3. The artificial intelligence-based long jump analysis and motion correction system according to claim 1, characterized in that, It also includes a front-end display module, which includes a chart processing submodule and a data display submodule; The chart processing submodule is connected to the scoring module, and it is used to calculate the actual long jump distance and the reference score, and to create a chart. The data display submodule is connected to the chart processing submodule and is used to display analysis videos, reference scores, and charts on the front end.

4. The artificial intelligence-based long jump analysis and motion correction system according to claim 3, characterized in that, It also includes a database for storing the long jump standard score table, user information, analysis videos and charts, and the data in the database is updated in real time. The database is connected to the chart processing submodule.

5. The artificial intelligence-based long jump analysis and motion correction system according to claim 1, characterized in that, It also includes an administrator module for managing user information, long jump standard score tables, and user login and / or system access permissions.

6. A long jump analysis and motion correction method based on artificial intelligence, applied to the long jump analysis and motion correction system according to any one of claims 1 to 5, characterized in that, include: Receive the user's complete long jump video; The long jump video is segmented to obtain processed frame images; The processed frame images are identified based on a human skeletal key point recognition model. Multiple key frame images are extracted from the frame images where skeletal key points have been identified, including: Based on the keypointrcnn_resnet50_fpn() model in PyTorch, skeletal keypoints were identified in the processed frame image. The skeletal keypoints include nose keypoint, eye keypoint, ear keypoint, shoulder keypoint, elbow keypoint, wrist keypoint, hip keypoint, knee keypoint, and ankle keypoint. In the processed frame image before the ankle key point moves, when the y-axis coordinate value of the wrist key point is the smallest, the frame image corresponding to the minimum y-axis coordinate value of the wrist key point is taken as the first key frame image. In the processed frame image before the ankle key point moves, when the difference between the wrist key point and the knee key point on the y-axis coordinate value is the largest, the frame image corresponding to the largest difference between the wrist key point and the knee key point on the y-axis coordinate value is taken as the second key frame image. In the processed frame image before the ankle key point moves, the frame image corresponding to the smallest angle between the line connecting the wrist key point to the ankle key point and the x-axis is the third key frame image. In the processed frame image after the ankle key point has moved, when the y-axis coordinate value of the wrist key point is the smallest, the frame image corresponding to the minimum y-axis coordinate value of the wrist key point is taken as the fourth key frame image. In the processed frame image after the ankle key point moves, when the y-axis coordinate value of the ankle key point is equal to the y-axis coordinate value of the ankle key point before the ankle key point moves, the frame image corresponding to the y-axis coordinate value of the ankle key point before the ankle key point moves is taken as the fifth key frame image. Based on the coordinates of key points of the skeleton in the multiple keyframe images, the passability of the action is analyzed, and a reference score and prompts are generated based on the analysis results. Calculate the actual long jump distance and generate a reference score for the long jump by combining it with the standard long jump score table.

7. The long jump analysis and motion correction method based on artificial intelligence according to claim 6, characterized in that, The long jump video is segmented to obtain processed frame images, including: Read the long jump video; The long jump video is segmented into a series of consecutive frames; Frame images are extracted frame by frame from each of the consecutive frames; The frame image is preprocessed and saved as a processed frame image.

8. The long jump analysis and motion correction method based on artificial intelligence according to claim 6, characterized in that, Analyzing the action qualification degree based on the calculated coordinates of the skeletal key points in the multiple key frame images, generating an action reference score and giving a prompt, comprising: Calculating the absolute value of the coordinate difference in the x-axis direction between every two of the wrist key point, shoulder key point, hip key point, knee key point and ankle key point in the first key frame image; if the value obtained by subtracting the first error value from all the absolute values of the coordinate differences is less than a first preset value, the action is qualified, a full score is given and a prompt is output; otherwise, a prompt is output, and corresponding points are deducted according to the absolute value of the coordinate difference in the x-axis direction between every two of the wrist key point, shoulder key point, hip key point, knee key point and ankle key point; Judging the x-axis coordinate values of the wrist key point, knee key point, ear key point and ankle key point in the second key frame image; if the x-axis coordinate values are arranged in descending or ascending order as the wrist key point, the ankle key point, the knee key point and the ear key point, the action is qualified, a full score is given and a prompt is output; otherwise, a prompt is output, and corresponding points are deducted according to the difference in the sequence among the wrist key point, ankle key point, knee key point and ear key point; Calculating the angle between the line connecting the wrist key point and the ankle key point in the third key frame image and the x-axis; if the angle is 45 degrees, the action is qualified, a full score is given and a prompt is output; otherwise, a prompt is output, and corresponding points are deducted according to the absolute value of the difference between the angle and 45 degrees; Calculating the coordinate difference in the y-axis between the knee key point in the fourth key frame image and the knee key point in the second key frame image to obtain a first coordinate difference; if the first coordinate difference is greater than a second preset value, the action is qualified, a full score is given and a prompt is output; if the first coordinate difference is less than or equal to the second preset value, corresponding points are deducted and a prompt is output based on the absolute value of the difference between the second preset value and the first coordinate difference; Calculating the coordinate difference in the x-axis between the ankle key point in the fifth key frame image and the ankle key point in the first key frame image to obtain a second coordinate difference; if the second coordinate difference is greater than a third preset value, the action is qualified, a full score is given and a prompt is output; if the second coordinate difference is less than or equal to the third preset value, corresponding points are deducted and a prompt is output based on the absolute value of the difference between the third preset value and the second coordinate difference.

9. The long jump analysis and motion correction method based on artificial intelligence according to claim 6, characterized in that, Calculating the actual long jump distance and performing scoring in combination with a long jump standard score table, comprising: The actual long jump distance is calculated using the following formula. : in, This is the actual long jump distance; For the user's height; The y-axis coordinate value of the eye keypoint in the first keyframe image; The y-axis coordinate value of the ankle key point in the first keyframe image; The x-axis coordinate value of the ankle key point in the first keyframe image; The x-axis coordinate value of the ankle key point in the fifth keyframe image; Based on the actual long jump distance The long jump reference score is calculated by combining the value with the long jump standard score table.

10. The long jump analysis and motion correction method based on artificial intelligence according to claim 6, characterized in that, Before receiving the complete long jump video of the user, the method further comprises: Inputting the long jump standard score table; Inputting user information, wherein the user information includes name, class, height, student number and gender.

Images