Method and device for judging badminton hitting training based on visual technology
By using computer vision technology and deep convolutional neural network models, the problems of forehand and backhand judgment and filtering of ground interference balls in badminton hitting training have been solved, achieving efficient and intelligent hitting assessment and reducing the misjudgment rate.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 恒鸿达(福建)体育科技有限公司
- Filing Date
- 2024-06-12
- Publication Date
- 2026-06-02
AI Technical Summary
Existing technologies lack a complete solution for judging badminton stroke training, especially for judging forehand and backhand strokes, filtering ground interference, and determining effective strokes, resulting in training and assessment that are not intelligent or objective enough.
Using computer vision technology, a deep convolutional neural network model is used to identify the positions of people, balls, rackets, and key points on the human body. Combined with the rectangular bounding box IOU algorithm and the cross product of two-dimensional planar vectors, static ball interference is filtered out, and the hitting position and hand shape are judged to achieve automatic judgment of forehand and backhand shots and the validity of the ball landing.
It improves the intelligence and objectivity of badminton hitting training, reduces the subjectivity of human judgment, lowers the misjudgment rate, and ensures the accuracy and efficiency of hitting assessment.
Smart Images
Figure CN118675102B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of badminton training technology, and in particular to a method and apparatus for judging badminton hitting training based on visual technology. Background Technology
[0002] Badminton requires quick movement, agile reactions, and coordinated actions, effectively improving students' speed, agility, coordination, and endurance, thus contributing to their overall physical development. Including badminton in junior high school physical education exams can increase students' attention to the sport, stimulate their interest, promote badminton, and enrich campus cultural life.
[0003] According to the examination rules, candidates should use forehand high clear and backhand lob techniques consecutively, hitting the shuttlecock in sequence until it lands in the valid area to be considered successful. Current technology lacks a complete training program for badminton hitting, particularly in determining forehand and backhand shots, filtering ground interference, and identifying valid shots. Summary of the Invention
[0004] The technical problem to be solved by the present invention is to provide a method and device for judging badminton hitting training based on visual technology. Based on computer vision technology, the field is pre-marked, and a deep convolutional neural network model is used to identify the positions of people, balls, rackets and key points of the human body. Then, according to logic, the ground interference balls are filtered out, the hitting position is judged, and the forehand and backhand shots are judged, so as to count the number of valid hits of the examinee in real time and effectively.
[0005] In a first aspect, the present invention provides a method for judging badminton hitting training based on visual technology, comprising:
[0006] Regional calibration process:
[0007] The first image of the first half of the court is obtained in real time by the first camera, and the preparation area before hitting the ball is marked.
[0008] The second image of the second half of the court is obtained in real time by the second camera, and the valid area for scoring is marked.
[0009] Pre-hit judgment process:
[0010] The foot rectangle is detected by a deep convolutional neural network model. The IOU between the box and the pre-hit preparation area is calculated. If the IOU is greater than the first threshold, it is determined that the foot is in the preparation area.
[0011] Static ball interference filtering process:
[0012] Initialize a list to store the coordinate information of the badminton detection boxes detected in the most recent N frames. For each subsequent frame, the coordinate information of the badminton detection box is compared with the coordinate information of each badminton detection box in each frame in the list. Based on the comparison result, it is determined whether the ball is static or moving. After the determination is completed, the coordinate information of the badminton detection box of the current frame is saved to the list, and the first frame data in the list is deleted.
[0013] The process of judging forehand and backhand shots:
[0014] The area A of the racket and the area B of the ball are detected by a deep convolutional neural network model, and the intersection I of the areas of A and B is calculated. It is then determined whether I / B is greater than a second threshold. If it is, the ball is hit.
[0015] Detection of thumb, little finger, wrist, and elbow using a deep convolutional neural network model;
[0016] If the thumb is above the wrist and on the designated side of the little finger, it is judged as a forehand shot; if the distance between the wrist and elbow is greater than the third threshold, it is judged as a backhand shot.
[0017] The process of determining whether the ball has landed is as follows:
[0018] If the vertical coordinate of a moving ball remains unchanged or changes less than a specified value for at least two consecutive frames, it is determined that the ball has landed. The coordinates of the ball after landing are marked as point P. Then, the cross product of vectors in a two-dimensional plane is used to determine whether point P is within the valid scoring area. If so, the ball landing is considered valid.
[0019] Secondly, the present invention provides a device for judging badminton hitting training based on visual technology, comprising:
[0020] The area calibration module is used to acquire the first image of the first half of the court in real time through the first camera and to mark the preparation area before hitting the ball; and to acquire the second image of the second half of the court in real time through the second camera and to mark the valid area for scoring the ball.
[0021] The pre-hit judgment module is used to detect the foot rectangle box through a deep convolutional neural network model, and calculate the IOU between the box and the pre-hit preparation area. If the IOU is greater than the first threshold, it is determined that the ball is in the preparation area.
[0022] The static ball interference filtering module is used to initialize a list to store the coordinate information of the badminton detection boxes detected in the most recent N frames. The coordinate information of the badminton detection box detected in each subsequent frame is compared with the coordinate information of the badminton detection box of each frame in the list. Based on the comparison result, it is determined whether it is a static ball or a moving ball. After the determination is completed, the coordinate information of the badminton detection box of the current frame is saved to the list, and the first frame data in the list is deleted.
[0023] The forehand and backhand hit judgment module is used to detect the area A of the racket and the area B of the ball through a deep convolutional neural network model, calculate the intersection I of the areas of A and B, and determine whether I / B is greater than the second threshold. If it is, then it is a hit.
[0024] Detection of thumb, little finger, wrist, and elbow using a deep convolutional neural network model;
[0025] If the thumb is above the wrist and on the designated side of the little finger, it is judged as a forehand shot; if the distance between the wrist and elbow is greater than the third threshold, it is judged as a backhand shot.
[0026] The ball landing validity judgment module is used to determine that the ball has landed when the vertical coordinate of the ball in motion remains unchanged or changes less than a specified value for at least two consecutive frames. The coordinate of the ball after landing is marked as point P. Then, the cross product of vectors in the two-dimensional plane is used to determine whether point P is within the valid scoring area. If so, the ball landing is considered valid.
[0027] The technical solutions provided in the embodiments of the present invention have at least the following technical effects:
[0028] Based on computer vision technology, including the rectangular bounding box IOU algorithm, 2D vector cross product, badminton shuttlecock landing frame judgment, and filtering of static interference balls on the ground, badminton hitting training and assessment are made more efficient, intelligent, and objective. Forehand and backhand shots are judged for violations using key point coordinates in computer vision technology, reducing the subjectivity of human judgment. For shuttlecock landing point judgment, due to the many interfering balls on the ground, human judgment is likely to lead to misjudgment of which shuttlecock has landed due to distraction; this technology reduces the misjudgment rate.
[0029] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description
[0030] The present invention will be further described below with reference to the accompanying drawings and embodiments.
[0031] Figure 1 This is a flowchart illustrating the overall process of the method in Embodiment 1 of the present invention.
[0032] Figure 2 This is a schematic diagram of the site in Embodiment 1 of the present invention;
[0033] Figure 3 This is a schematic diagram of the second camera calibration in Embodiment 1 of the present invention;
[0034] Figure 4 This is a schematic diagram of the device in Embodiment 2 of the present invention. Detailed Implementation
[0035] This invention provides a method and apparatus for judging badminton hitting training based on visual technology. Based on computer vision technology, the field is pre-calibrated, and a deep convolutional neural network model is used to identify the positions of people, balls, rackets, and key points of the human body. Then, based on logic, ground interference balls are filtered out, the hitting position is judged, and forehand and backhand shots are judged, thereby effectively counting the number of valid hits by the examinee in real time.
[0036] Example 1
[0037] This embodiment provides a method for judging badminton hitting training based on visual technology, such as... Figure 1 As shown, it includes:
[0038] Regional calibration process:
[0039] The first image of the first half of the court is obtained in real time by the first camera, and the preparation area before hitting the ball is marked.
[0040] The second image of the second half of the court is obtained in real time by the second camera, and the valid area for scoring is marked.
[0041] like Figure 2 As shown, the first half is Figure 2 The right half of the court is marked with coordinates A, B, C, and D as the candidate's pre-hit preparation area (ABCD). After hitting the ball, the candidate must return to the pre-hit preparation area to prepare for the next hit. The second half of the court is... Figure 2 The left half of the court is marked with the coordinates of four points: E, F, G, and H. This area is the valid scoring zone for hitting the ball. The positions of the first and second cameras can be set according to requirements, as long as they can capture a complete and clear image of the designated area. Preferably, such as... Figure 2 As shown, the first camera (C1) is located on the left side of the left half of the court, facing the examinee; the second camera (C2) is located on the side of the entire court. Because the second camera (C2) is angled to film the effective scoring area EFGH, the effective scoring area is marked as an irregular quadrilateral that is close to a parallelogram, as shown below. Figure 3 As shown.
[0042] Pre-hit judgment process:
[0043] The system detects the foot rectangle using a deep convolutional neural network model, calculates the Intersection over Union (IOU) between the box and the pre-hit preparation area, and determines that the box is within the candidate's preparation area if the IOU is greater than a first threshold (e.g., 0.9).
[0044] Static ball interference filtering process:
[0045] Initialize a list to store the coordinate information of the badminton detection boxes detected in the most recent N frames (e.g., 5 frames). For each subsequent frame, the coordinate information of the badminton detection box is compared with the coordinate information of each badminton detection box in each frame in the list. Based on the comparison result, it is determined whether the ball is static or moving. After the determination is completed, the coordinate information of the badminton detection box of the current frame is saved to the list, and the first frame data in the list is deleted.
[0046] Preferably, the Interchange of Value (IOU) between the detected shuttlecock detection box in each subsequent frame and the coordinate information of each shuttlecock detection box in each frame of the list can be calculated. If the IOU is greater than the fourth threshold (e.g., 0.9), it is identified as a static shuttlecock.
[0047] The process of judging forehand and backhand shots:
[0048] The system uses a deep convolutional neural network model to detect the area A of the racket and the area B of the ball, and calculates the intersection I of the areas of A and B. It then determines whether I / B is greater than a second threshold (e.g., 0.9). If it is, the ball is hit.
[0049] Detection of thumb, little finger, wrist, and elbow using a deep convolutional neural network model;
[0050] If the thumb is above the wrist and on the designated side of the little finger (for a right-handed shot, the thumb should be to the right of the little finger; for a left-handed shot, the thumb should be to the left of the little finger), it is judged as a forehand shot; if the distance between the wrist and elbow is greater than the third threshold (set according to the length of the forearm), it is judged as a backhand shot.
[0051] The process of determining whether the ball has landed is as follows:
[0052] If the vertical coordinate of a moving ball remains unchanged or changes less than a specified value for at least two consecutive frames, it is determined that the ball has landed. The coordinates of the landed ball are marked as point P. Then, the cross product of vectors in a two-dimensional plane is used to determine whether point P is within the valid scoring area. If so, the ball's landing is considered valid. Specifically, this includes:
[0053] Based on the marking of the valid scoring area, the line segments corresponding to each edge are obtained and converted into corresponding vectors. At the same time, the line connecting each endpoint to point P is converted into a vector. Then, the cross product dis of two vectors with the same starting point is calculated. If dis is greater than 0, then point P is to the left of the line segment. If dis is less than 0, then point P is to the right of the line segment. If dis is equal to 0, then point P is on the line segment. Based on the positional relationship between point P and each line segment, it is determined whether point P is within the valid scoring area.
[0054] by Figure 3 Taking the line EH as an example, the coordinates of point E are (x1, y1), the coordinates of point H are (x2, y2), and the spherical coordinates are P(x, y).
[0055] Vector EP = (x - x1, y - y1)
[0056] Vector EH = (x2-x1, y2-y1)
[0057] Then calculate the cross product of vectors EP and EH, using the following formula:
[0058] dis=(x-x1)×(y2-y1)-(y-y1)×(x2-x1)
[0059] Similarly, determine the positional relationship between point P and other line segments, such as... Figure 3 As shown, when point P is simultaneously located to the right of EH, to the left of FG, to the left of EF, and to the right of HG, then point P is within the valid scoring area.
[0060] After the area calibration process, the first camera performs the pre-hit judgment process and the forehand / backhand hit judgment process, while the second camera performs the static ball interference filtering process and the ball landing validity judgment process. Combining the judgment results of the first and second cameras, the following examination procedure can be executed (taking right-handed racket hitting as an example):
[0061] 1. Candidates enter the testing area and stand in the middle of the preparation area. The test begins when they hear the instructions from the test administrator.
[0062] 2. The ball machine first serves a ball to the forehand backcourt, then a ball to the backhand net, alternating between serves. Candidates should continuously use forehand high clears and backhand flicks in sequence, hitting the ball into the valid area to succeed.
[0063] 3. After each shot, the candidate should return to the preparation area in the middle of the court, with both feet inside the preparation area and not touching the lines. The candidate should not leave the preparation area before the ball is served by the ball machine.
[0064] Based on the same inventive concept, this application also provides an apparatus corresponding to the method in Embodiment 1, as detailed in Embodiment 2.
[0065] Example 2
[0066] This embodiment provides a device for judging badminton hitting training based on visual technology, such as... Figure 4 As shown, it includes:
[0067] The area calibration module is used to acquire the first image of the first half of the court in real time through the first camera and to mark the preparation area before hitting the ball; and to acquire the second image of the second half of the court in real time through the second camera and to mark the valid area for hitting the ball and scoring.
[0068] The pre-hit judgment module is used to detect the foot rectangle box through a deep convolutional neural network model, and calculate the IOU between the box and the pre-hit preparation area. If the IOU is greater than a first threshold (e.g., 0.9), it is judged to be within the preparation area.
[0069] The static ball interference filtering module initializes a list to store the coordinate information of the badminton detection boxes detected in the most recent N frames (e.g., 5 frames). For each subsequent frame, the coordinate information of the badminton detection box is compared with the coordinate information of each badminton detection box in each frame in the list. Based on the comparison result, it is determined whether the ball is static or moving. After the determination is completed, the coordinate information of the badminton detection box of the current frame is saved to the list, and the first frame data in the list is deleted.
[0070] The forehand and backhand hit judgment module is used to detect the area A of the racket and the area B of the ball through a deep convolutional neural network model, calculate the intersection I of the areas of A and B, and determine whether I / B is greater than a second threshold (e.g., 0.9). If it is, then it is a hit.
[0071] Detection of thumb, little finger, wrist, and elbow using a deep convolutional neural network model;
[0072] If the thumb is above the wrist and on the designated side of the little finger, it is judged as a forehand shot; if the distance between the wrist and elbow is greater than the third threshold, it is judged as a backhand shot.
[0073] The ball landing validity judgment module is used to determine that the ball has landed when the vertical coordinate of the ball in motion remains unchanged or changes less than a specified value for at least two consecutive frames. The coordinate of the ball after landing is marked as point P. Then, the cross product of vectors in the two-dimensional plane is used to determine whether point P is within the valid scoring area. If so, the ball landing is considered valid.
[0074] Preferably, in the process of determining whether the ball has landed effectively, the cross product of vectors in a two-dimensional plane is used to determine whether point P is within the valid scoring area, specifically including:
[0075] Based on the marking of the valid scoring area, the line segments corresponding to each edge are obtained and converted into corresponding vectors. At the same time, the line connecting each endpoint to point P is converted into a vector. Then, the cross product dis of two vectors with the same starting point is calculated. If dis is greater than 0, then point P is to the left of the line segment. If dis is less than 0, then point P is to the right of the line segment. If dis is equal to 0, then point P is on the line segment. Based on the positional relationship between point P and each line segment, it is determined whether point P is within the valid scoring area.
[0076] Preferably, in the static ball interference filtering process, the coordinate information of each badminton detection box in each frame of the list is compared. Specifically, the IOU of the detected badminton detection box in each subsequent frame with the coordinate information of each badminton detection box in each frame of the list is calculated. If the IOU is greater than the fourth threshold (e.g., 0.9), it is identified as a static ball.
[0077] Since the apparatus described in Embodiment 2 of the present invention is an apparatus used to implement the method of Embodiment 1 of the present invention, those skilled in the art can understand the specific structure and variations of the apparatus based on the method described in Embodiment 1 of the present invention, and therefore will not be described again here. All apparatuses used in the method of Embodiment 1 of the present invention fall within the scope of protection of the present invention.
[0078] This invention, based on computer vision technology, includes a rectangular bounding box IOU algorithm, two-dimensional vector cross product, badminton shuttlecock landing frame judgment, and filtering of static interference balls on the ground, making badminton hitting training and assessment more efficient, intelligent, and objective. Forehand and backhand shots are judged for violations using key point coordinates in computer vision technology, reducing the subjectivity of human judgment. For shuttlecock landing point judgment, due to the many interfering balls on the ground, human judgment is likely to lead to misjudgment of which shuttlecock has landed due to distraction; this invention reduces the misjudgment rate.
[0079] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0080] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0081] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0082] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0083] While specific embodiments of the present invention have been described above, those skilled in the art should understand that the specific embodiments described are merely illustrative and not intended to limit the scope of the present invention. Equivalent modifications and variations made by those skilled in the art in accordance with the spirit of the present invention should be covered within the scope of protection of the claims of the present invention.
Claims
1. A method for judging badminton hitting training based on visual technology, characterized in that, include: Regional calibration process: The first image of the first half of the court is obtained in real time by the first camera, and the preparation area before hitting the ball is marked. The second image of the second half of the court is obtained in real time by the second camera, and the valid area for scoring is marked. After the valid scoring area for hitting the ball is marked, the first camera performs the pre-hit judgment process and the forehand and backhand hit judgment process, and the second camera performs the static ball interference filtering process and the ball landing valid judgment process. Pre-hit judgment process: The foot rectangle is detected by a deep convolutional neural network model. The IOU between the box and the pre-hit preparation area is calculated. If the IOU is greater than the first threshold, it is determined that the foot is in the preparation area. Static ball interference filtering process: Initialize a list to store the coordinate information of the badminton detection boxes detected in the most recent N frames. For each subsequent frame, the coordinate information of the badminton detection box is compared with the coordinate information of each badminton detection box in each frame in the list. Based on the comparison result, it is determined whether the ball is static or moving. After the determination is completed, the coordinate information of the badminton detection box of the current frame is saved to the list, and the first frame data in the list is deleted. The process of judging forehand and backhand shots: The area A of the racket and the area B of the ball are detected by a deep convolutional neural network model, and the intersection I of the areas of A and B is calculated. It is then determined whether I / B is greater than a second threshold. If it is, the ball is hit. Detection of thumb, little finger, wrist, and elbow using a deep convolutional neural network model; If the thumb is above the wrist and on the designated side of the little finger, it is judged as a forehand shot; if the distance between the wrist and elbow is greater than the third threshold, it is judged as a backhand shot. The process of determining whether the ball has landed is as follows: When the vertical coordinate of a moving ball remains unchanged or changes less than a specified value for at least two consecutive frames, it is determined that the ball has landed. The coordinates of the ball after landing are marked as point P. Then, the cross product of vectors in the two-dimensional plane is used to determine whether point P is within the valid scoring area. If so, the ball landing is considered valid. In the static ball interference filtering process, the coordinate information of each badminton detection box in each frame of the list is compared. Specifically, the IOU of the detected badminton detection box in each subsequent frame with the coordinate information of each badminton detection box in each frame of the list is calculated. If the IOU is greater than the fourth threshold, it is identified as a static ball.
2. The method according to claim 1, characterized in that: In the process of determining whether the ball has landed effectively, the cross product of vectors in a two-dimensional plane is used to determine whether point P is within the valid scoring area. Specifically, this includes: Based on the marking of the valid scoring area, the line segments corresponding to each edge are obtained and converted into corresponding vectors. At the same time, the line connecting each endpoint to point P is converted into a vector. Then, the cross product dis of two vectors with the same starting point is calculated. If dis is greater than 0, then point P is to the left of the line segment. If dis is less than 0, then point P is to the right of the line segment. If dis is equal to 0, then point P is on the line segment. Based on the positional relationship between point P and each line segment, it is determined whether point P is within the valid scoring area.
3. The method according to claim 1, characterized in that: The fourth threshold is 0.
9.
4. The method according to claim 1, characterized in that: The first threshold is 0.9, the second threshold is 0.9, and the third threshold is set according to the length of the forearm. The value of N is 5.
5. A device for judging badminton hitting training based on visual technology, characterized in that, include: The area calibration module is used to acquire the first image of the first half of the court in real time through the first camera and to mark the preparation area before hitting the ball. The second image of the second half of the court is obtained in real time by the second camera, and the valid area for scoring is marked. After the valid scoring area for hitting the ball is marked, the first camera performs the pre-hit judgment process and the forehand and backhand hit judgment process, and the second camera performs the static ball interference filtering process and the ball landing valid judgment process. The pre-hit judgment module is used to detect the foot rectangle box through a deep convolutional neural network model, and calculate the IOU between the box and the pre-hit preparation area. If the IOU is greater than the first threshold, it is determined that the ball is in the preparation area. The static ball interference filtering module is used to initialize a list to store the coordinate information of the badminton detection boxes detected in the most recent N frames. The coordinate information of the badminton detection box detected in each subsequent frame is compared with the coordinate information of the badminton detection box of each frame in the list. Based on the comparison result, it is determined whether it is a static ball or a moving ball. After the determination is completed, the coordinate information of the badminton detection box of the current frame is saved to the list, and the first frame data in the list is deleted. The forehand and backhand hit judgment module is used to detect the area A of the racket and the area B of the ball through a deep convolutional neural network model, calculate the intersection I of the areas of A and B, and determine whether I / B is greater than the second threshold. If it is, then it is a hit. Detection of thumb, little finger, wrist, and elbow using a deep convolutional neural network model; If the thumb is above the wrist and on the designated side of the little finger, it is judged as a forehand shot; if the distance between the wrist and elbow is greater than the third threshold, it is judged as a backhand shot. The ball landing validity judgment module is used to determine that the ball has landed when the vertical coordinate of the ball in motion remains unchanged or changes less than a specified value for at least two consecutive frames. The coordinate of the ball after landing is marked as point P. Then, the cross product of vectors in the two-dimensional plane is used to determine whether point P is within the valid scoring area. If so, the ball landing is considered valid. In the static ball interference filtering process, the coordinate information of each badminton detection box in each frame of the list is compared. Specifically, the IOU of the detected badminton detection box in each subsequent frame with the coordinate information of each badminton detection box in each frame of the list is calculated. If the IOU is greater than the fourth threshold, it is identified as a static ball.
6. The apparatus according to claim 5, characterized in that: In the process of determining whether the ball has landed effectively, the cross product of vectors in a two-dimensional plane is used to determine whether point P is within the valid scoring area. Specifically, this includes: Based on the marking of the valid scoring area, the line segments corresponding to each edge are obtained and converted into corresponding vectors. At the same time, the line connecting each endpoint to point P is converted into a vector. Then, the cross product dis of two vectors with the same starting point is calculated. If dis is greater than 0, then point P is to the left of the line segment. If dis is less than 0, then point P is to the right of the line segment. If dis is equal to 0, then point P is on the line segment. Based on the positional relationship between point P and each line segment, it is determined whether point P is within the valid scoring area.
7. The apparatus according to claim 5, characterized in that: The fourth threshold is 0.
9.
8. The apparatus according to claim 5, characterized in that: The first threshold is 0.9, the second threshold is 0.9, and the third threshold is set according to the length of the forearm. The value of N is 5.