Tai Chi motion detection scoring teaching system based on AI

Abstract

The invention provides a Tai Chi motion detection scoring teaching system based on AI, and relates to the technical field of Tai Chi teaching, the Tai Chi motion detection scoring teaching system comprises a video acquisition module, a posture skeleton node detection module, a node calculation module, a motion standard degree scoring module and a feedback module, the video acquisition module is used for acquiring a motion video of a Tai Chi learner; human body postures in the video are detected in real time through a posture skeleton node detection module by using an AI deep learning algorithm, and a node calculation module is based on coordinate information of skeleton nodes; according to the invention, based on a real-time image processing algorithm and an AI scoring algorithm, core functions of real-time identification, real-time tracking, real-time detection, real-time scoring, big data statistical analysis output normative evaluation and guidance opinion correction of human body postures are developed; and therefore, real-time detection, accurate scoring and effective guidance of the actions of the shadowboxing learner can be realized, and the efficiency and action normalization of shadowboxing learning can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of Tai Chi teaching, and particularly to an AI-based Tai Chi movement detection and scoring teaching system. Background Art

[0002] As a traditional Chinese martial art and fitness exercise, Tai Chi features gentle movements, strong coherence, and emphasizes body postures and breath regulation. Many Tai Chi learners fail to achieve the expected fitness effects due to the lack of professional coach guidance, resulting in non-standard movements. Existing Tai Chi teaching methods mostly rely on video teaching, lacking real-time posture correction and movement scoring mechanisms. There is an urgent need for an intelligent system that can detect learners' movement postures in real time, give scores for movement standardization, and provide targeted correction suggestions. Therefore, we propose a Tai Chi movement detection and scoring teaching system that uses real-time image processing algorithms and AI scoring algorithms to solve the above problems. Summary of the Invention

[0003] The purpose of the present invention is to address the deficiencies in the prior art. Existing Tai Chi teaching methods mainly rely on video teaching, lacking real-time posture correction and movement scoring mechanisms. There is an urgent need for an intelligent system that can detect learners' movement postures in real time, give scores for movement standardization, and provide targeted correction suggestions.

[0004] To achieve the above objective, the present invention adopts the following technical solutions: An AI-based Tai Chi movement detection and scoring teaching system includes a video acquisition module, a posture skeleton node detection module, a node calculation module, a movement standardization scoring module, and a feedback module. The video acquisition module is used to collect the movement videos of Tai Chi learners and, through the posture skeleton node detection module, uses AI deep learning algorithms to detect the human postures in the videos in real time and extract the coordinate information of 12 key points of the whole body skeleton. The node calculation module, based on the coordinate information of the skeleton nodes, calculates the angular posture data of each part of the limb in real time, and through the movement standardization scoring module, compares the extracted key point information with a pre-recorded standard movement library, calculates the differences of each key angle, and uses the weighted average method to calculate the standardization score of the overall movement. The feedback module generates movement correction suggestions according to the scoring results and feeds them back to the learner through the display screen and voice prompts.

[0005] Further, the video acquisition module uses a high-definition camera or other video acquisition devices to collect videos, and transmits the movement images of the learners to the posture skeleton node detection module through the video acquisition module.

[0006] Further, the 12 key point coordinate information is the right wrist, right elbow, right shoulder, right hip, right knee, right ankle, left shoulder, left elbow, left wrist, left hip, left knee, and left ankle.

[0007] Further, the pose skeleton node detection module uses pre-trained deep learning models of OpenPose and Mediapipe pose estimation models to process video frames.

[0008] Further, the node calculation module calculates the angle between two line segments in real time. It calculates the vectors of two line segments using the coordinate data of three adjacent key points, calculates the cosine value between the two line segments using the dot product of vectors and the vector modulus (length), and then calculates the angle through the inverse cosine function (arccos) and converts it to degrees to obtain the angle data between three adjacent key points.

[0009] Further, the node calculation module calculates the following eight key angle data, and the three adjacent key points are respectively: 1) Right wrist - right elbow - right shoulder; 2) Right elbow - right shoulder - right hip; 3) Right shoulder - right hip - right knee; 4) Right hip - right knee - right ankle; 5) Left wrist - left elbow - left shoulder; 6) Left elbow - left shoulder - left hip; 7) Left shoulder - left hip - left knee; 8) Left hip - left knee - left ankle.

[0010] Further, the standard action library is a pre-recorded standard Tai Chi action library, including the key point coordinates and angle data of each action.

[0011] Further, the feedback module determines the key points and action details that need to be improved through a scoring structure. The correction suggestions include the direction and amplitude of action adjustment. The feedback module uses a display screen to display the correction suggestions in real time and provides voice prompt support at the same time.

[0012] Compared with the prior art, the beneficial effects of the present invention are: 1. The present invention uses real-time image processing algorithms and AI evaluation algorithms to achieve real-time monitoring and high-precision pose detection, adds a real-time pose correction and action scoring mechanism, and combines an immersive multimedia interactive game experience, enabling learners to continuously correct their pose movements through interaction and experience a more participatory learning process, solving the low efficiency and dullness of the conventional video learning experience.

[0013] 2. The present invention detects 12 key points of the whole body in real time, calculates the included angles and pose data based on these points, and provides a more comprehensive pose analysis.

[0014] 3. By comparing the differences between the learner's movements and the standard movements at multiple key angles, the present invention obtains an accurate movement standard score, realizing an intelligent movement evaluation standard for quantitative evaluation of the movement standard degree.

[0015] 4. The present invention generates specific correction suggestions according to the movement standard score and feeds them back to the learner through a display screen or voice prompt, providing personalized guidance.

[0016] 5. Through real-time monitoring and correction, it helps learners quickly discover and correct mistakes in their movements, improving learning efficiency.

[0017] 6. The movement standard score system ensures that learners can gradually meet the requirements of standard movements, improving the practice effect of Tai Chi.

[0018] 7. The present invention has a wide range of application scenarios. In addition to Tai Chi teaching, this system can also be applied to other martial arts, fitness exercises, and dances and other fields that require movement standard evaluation.

[0019] 8. The present invention combines deep learning algorithms and big data analysis to achieve high-precision and high-efficiency pose detection and scoring, providing normative evaluation and correction guidance, with high technical content and practical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a system frame of a Tai Chi movement detection and scoring teaching system based on AI provided by the present invention.

[0021] Legend: 1. Video acquisition module; 2. Pose skeleton node detection module; 3. Node calculation module; 4. Movement standard score module; 5. Feedback module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0023] For the convenience of understanding the present invention, the present invention will be described more comprehensively below with reference to the relevant. Several embodiments of the present invention are given. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0024] It should be noted that when an element is referred to as being "fixed to" another element, it can be directly on the other element or there can also be an intermediate element. When an element is considered to be "connected to" another element, it can be directly connected to the other element or there may be an intermediate element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are only for illustrative purposes.

[0025] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this invention belongs. The terms used herein in the specification of this invention are only for the purpose of describing specific embodiments and are not intended to limit the invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0026] Embodiment 1 As Figure 1 shown, the present invention provides a technical solution: an AI-based Taichi movement detection scoring and teaching system, including a video acquisition module 1, a posture skeleton node detection module 2, a node calculation module 3, a movement standardization scoring module 4 and a feedback module 5. Through the collaborative work of the above modules, real-time posture detection, movement standardization scoring and dynamic correction suggestions for Taichi learners are realized; Specifically: The video acquisition module 1 is used to acquire the movement videos of Taichi learners, and the posture skeleton node detection module 2 uses the AI deep learning algorithm to detect the human postures in the videos in real time, extract the coordinate information of 12 key points of the whole body skeleton of the human body. The node calculation module 3 calculates the angular posture data of each part of the limb in real time based on the coordinate information of the skeleton nodes, and the movement standardization scoring module 4 compares the extracted key point information with the pre-recorded standard movement library, calculates the difference of each key angle, and uses the weighted average method to calculate the standardization score of the overall movement. The feedback module 5 generates movement correction suggestions according to the scoring results and feeds them back to the learners through the display screen and voice prompts.

[0027] In the above embodiment, the more detailed functions, implementation methods and outputs of the video acquisition module 1, the posture skeleton node detection module 2, the node calculation module 3, the movement standardization scoring module 4 and the feedback module 5 are as follows: Explanation of the video acquisition module 1: Function: Real-time acquisition of the movement pictures of Taichi learners; Implementation method: Through a high-definition camera or other video acquisition devices, transmit the movement pictures of the learners to the posture skeleton node detection module 2; Output: The action video of the learner is transmitted to the pose skeleton node detection module 2 through the video acquisition module 1.

[0028] Description of the pose skeleton node detection module 2: Function: Utilize the AI deep learning algorithm to detect the human pose in the video in real time and output the coordinate information of 12 key points of the whole body skeleton.

[0029] Implementation method: Use the pre-trained deep learning models, OpenPose and Mediapipe pose estimation models, to process the video frames; Detect and extract the coordinates of 12 key points, including the right wrist, right elbow, right shoulder, right hip, right knee, right ankle, left wrist, left elbow, left shoulder, left hip, left knee, and left ankle; Output: The coordinate information of 12 key points of the whole body skeleton.

[0030] Description of the node calculation module 3: Function: Based on the coordinate information of the skeleton nodes, calculate the angular pose data of each part of the limb in real time.

[0031] Implementation method: Calculate the angle between two line segments in real time. Use the coordinate data of three adjacent key points to calculate the vectors of the two line segments, use the dot product of the vectors and the vector modulus (length) to calculate the cosine value between the two line segments, and then calculate the angle through the inverse cosine function (arccos) and convert it to degrees to obtain the angle data between three adjacent key points. Calculate the following eight key angle data: 1) Right wrist - right elbow - right shoulder; 2) Right elbow - right shoulder - right hip; 3) Right shoulder - right hip - right knee; 4) Right hip - right knee - right ankle; 5) Left wrist - left elbow - left shoulder; 6) Left elbow - left shoulder - left hip; 7) Left shoulder - left hip - left knee; 8) Left hip - left knee - left ankle.

[0032] Description of the action standard degree scoring module 4: Output: The angular pose data of each part of the limb; Function: Compare the extracted key point information with the pre-recorded standard action library and calculate the action standard degree score; Implementation method: Pre-record a standard Tai Chi movement library, including the key point coordinates and angle data of each movement; compare the key point coordinates and angle data of the learner's current movement with the standard movement library, calculate the differences of each key angle, and use the weighted average method to calculate the standard score of the overall movement.

[0033] Output: Movement standard score.

[0034] Regarding the description of the feedback module 5: Function: Generate movement correction suggestions according to the scoring results, and feedback them to the learner through the display screen and voice prompts.

[0035] Implementation method: Analyze the movement standard score, determine the key points and movement details that need to be improved, generate specific movement correction suggestions, including the direction and amplitude of movement adjustment, display the correction suggestions in real time through the display screen, and at the same time give the learner guidance through voice prompts Output: Movement correction suggestions and real-time feedback.

[0036] It should be added that the feedback module 5 uses a display screen to display correction suggestions in real time and provides voice prompt support at the same time.

[0037] The working principle (implementation process) of this system: The learner stands in front of the video acquisition device and starts practicing Tai Chi. The video acquisition module 1 captures the movement pictures of the learner in real time. The posture skeleton node detection module 2 processes the video pictures and extracts the coordinate information of 12 key points. The posture skeleton node calculation module 3 calculates the angle posture data of each part of the limb based on the key point coordinates. The movement standard score module 4 compares the current movement data with the standard movement library and calculates the movement standard score. The feedback module 5 generates correction suggestions according to the scoring results and feeds them back to the learner in real time through the display screen and voice prompts to help them correct their movements and improve the practice effect.

[0038] In summary, this invention patent is based on real-time image processing algorithms and AI scoring algorithms, and focuses on several core functions such as real-time recognition, real-time tracking, real-time detection, real-time scoring, big data statistical analysis output of normative evaluation and correction guidance opinions of human postures, so as to realize real-time detection, accurate scoring and effective guidance of the movements of Tai Chi learners, improve the learning efficiency and movement standardization of Tai Chi, and has broad application prospects.

[0039] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirits of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An AI-based Tai Chi movement detection and scoring teaching system, characterized by: It includes a video acquisition module, a posture bone node detection module, a node calculation module, an action standard scoring module and a feedback module. The video acquisition module is used to collect action videos of Tai Chi learners, and the posture bone node detection module uses an AI deep learning algorithm to detect the human posture in the video in real time, and extract the coordinate information of 12 key points of the human skeleton. The node calculation module calculates the angle posture data of each part of the limb in real time based on the coordinate information of the bone nodes, and compares the extracted key point information with the pre-recorded standard action library through the action standard scoring module, calculates the difference of each key angle, and uses the weighted average method to calculate the standard score of the overall action. The feedback module generates action correction suggestions based on the scoring results, and feeds back to the learner by displaying the screen image and voice prompts.

2. The AI-based Tai Chi movement detection and scoring teaching system according to claim 1, characterized in that: The video acquisition module uses a high-definition camera or other video acquisition equipment to acquire video, and transmits the learner's action pictures to the posture skeleton node detection module through the video acquisition module.

3. The AI-based Tai Chi movement detection and scoring teaching system according to claim 1, characterized in that: The coordinate information of the 12 key points are right wrist, right elbow, right shoulder, right waist, right knee, right ankle, left shoulder, left elbow, left wrist, left waist, left knee and left ankle.

4. The AI-based Tai Chi movement detection and scoring teaching system according to claim 1, characterized in that: The posture skeleton node detection module uses OpenPose and Mediapipe posture estimation models to pre-train deep learning models to process video frames.

5. The AI-based Tai Chi movement detection and scoring teaching system according to claim 3 is characterized in that: The node calculation module calculates the angle between two line segments in real time, calculates the vectors of the two line segments using the coordinate data of three adjacent key points, calculates the cosine value between the two line segments using the vector dot product and the vector modulus, and then calculates the angle through the inverse cosine function and converts it into degrees to obtain the angle data between the three adjacent key points.

6. The AI-based Tai Chi movement detection and scoring teaching system according to claim 5, characterized in that: The node calculation module calculates the following eight key angle data, and the three adjacent key points are: 1) Right wrist - right elbow - right shoulder; 2) Right elbow - right shoulder - right waist; 3) Right shoulder - right hip - right knee; 4) Right waist-right knee-right ankle; 5) Left wrist - left elbow - left shoulder; 6) Left elbow - left shoulder - left hip; 7) Left shoulder - left hip - left knee; 8) Left hip-left knee-left ankle.

7. The AI-based Tai Chi movement detection and scoring teaching system according to claim 1, characterized in that: The standard action library is a pre-recorded standard Tai Chi action library, including key point coordinates and angle data of each action.

8. The AI-based Tai Chi movement detection and scoring teaching system according to claim 1, characterized in that: The feedback module determines the key points and action details that need to be improved through a scoring structure, and the correction suggestions include the direction and amplitude of the action adjustment. The feedback module uses a display screen to display the correction suggestions in real time and provides voice prompt support.

Images