Billiard auxiliary training system and method based on vision

Abstract

The invention discloses a vision-based auxiliary billiard training system and method, and aims to solve the technical problem that existing billiard training cannot assist in correcting postures and actions. The system comprises a projection module which is used for projecting a white ball, a placing position of a target ball and a ball hitting auxiliary line on a table top of a billiard table, and is also used for projecting an auxiliary standing point on the ground; the image acquisition module is used for acquiring images of the desktop and the top, right ahead and side surfaces of the trainer; the image analysis module is used for analyzing postures and actions of the trainee through a visual technology and comparing the postures and actions with standard actions; and the output module is used for outputting an action diagnosis report and a voice correction prompt. The method has the advantages that postures and actions of the trainee are analyzed through the visual technology, the trainee is helped to correct wrong postures and actions in time, and the training efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of sports training equipment and methods, and in particular to a vision-based billiards auxiliary training system and method. Background Art

[0002] In the early stage of training, billiards novices cannot accurately judge the correctness of the decomposed movements such as standing position, posture, putting the club, moving the club, and hitting the club. Usually, the coach needs to repeatedly demonstrate the standard movements and correct the wrong movements. This training method is cumbersome and has low training efficiency. The existing auxiliary system can feedback the corrective movement through the motion trajectory of the club. For example, the Chinese patent document with publication number CN107635627A records a billiards hitting auxiliary method, a billiards hitting auxiliary system, and a portable electronic device, which makes hitting suggestions by recording the movement of the club and the hitting path. However, this method is suitable for skilled athletes and can assist them in making the best hitting path. However, for novices, the movement of the body when hitting the ball is more important. If the movement is not standard, it is difficult to really learn billiards by only relying on the club and the hitting trajectory to adjust the hitting force. Summary of the invention

[0003] The present invention provides a vision-based billiards auxiliary training system and method, so as to solve the technical problem that the existing billiards training cannot assist in correcting posture and movement.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solutions: Design a vision-based billiards auxiliary training system, including: A projection module is used to project the placement of the white ball, the target ball and the assisting line for hitting the ball on the tabletop of the billiard table, and is also used to project the assisting standing position on the ground; An image acquisition module is used to acquire images of the desktop and the top, front, and sides of the trainee; Image analysis module, used to analyze the trainer's posture and movements through visual technology and compare them with standard movements; The output module is used to output motion diagnosis reports and speech correction reminders.

[0005] Furthermore, the image acquisition module includes seven cameras, namely, one camera located directly above the center point of the billiard table, four cameras located above the extension lines of the diagonal lines of the billiard table, and two cameras located on both sides of the width direction of the billiard table.

[0006] Furthermore, the image analysis module includes a posture estimation algorithm for extracting the coordinates of key points of the human body, and the key points of the human body include the center of gravity foot position, auxiliary foot position, left knee, right knee, grip point, wrist, elbow, and shoulder position of the gripping hand, and hand support point, wrist, elbow, and shoulder position of the auxiliary hand.

[0007] Furthermore, the output module includes a display screen and a speaker, the display screen can obtain the trainee's training video and diagnosis report, and the speaker can broadcast voice prompt information to the trainee.

[0008] Furthermore, it also includes a terminal device for wirelessly receiving and storing training videos and diagnostic reports.

[0009] The present invention also provides a vision-based billiards auxiliary training method, comprising the following steps: S1: Preparation before training. The projection module projects the placement of the white ball and the target ball, as well as the assisting line for hitting the ball, on the billiard table, and projects the assisting standing position. The trainee places the target ball and the white ball correctly according to the projection and voice prompts. S2: Standard action recording, the trainee takes the position according to the auxiliary standing position, and performs the standard action under the guidance of the coach, and the image acquisition module and the image analysis module record the standard action into the system; S3: Posture correction: the trainee gets into position again and takes the posture before hitting the ball. The image analysis module identifies the key points of the posture before hitting the ball and compares it with the standard action recorded in S2. The trainee adjusts the posture to meet the standard action. S4: Batting training. Batting includes releasing the club, moving the club, and releasing the club. During the batting process, the image analysis module analyzes in real time whether the batting action is standard.

[0010] 7. The vision-based billiards auxiliary training method according to claim 6 is characterized in that in step S2, when it is detected that the direction of the cue coincides with the goal line, the standard action is recorded, and the image analysis module extracts and stores the key points in the standard action.

[0011] Furthermore, in step S3, after the trainee stands in the correct position, the projection module projects the club head position, auxiliary hand support point, wrist, elbow and shoulder positions corresponding to the standard action on the desktop, and the trainee adjusts his posture to meet the standard position.

[0012] Furthermore, in step S4, after the posture correction, the pole is moved 2-6 times, and the pole is stopped for a period of time twice, and the image acquisition module obtains the image of the trainee and starts to store it.

[0013] Furthermore, in step S4, when hitting the ball, the image acquisition module obtains the desktop image, extracts the club movement trajectory through image analysis, compares it with the goal line, and evaluates the result of the shot as "correct", "left-biased" or "right-biased". Compared with the prior art, the beneficial technical effects of the present invention are: The present invention can collect and identify the motion data of billiard trainees, and compare it with standard motions, and correct the hitting motions through projection, voice, display screen and other means, so as to help trainees correct wrong postures and motions in time and improve training efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a structural schematic diagram of the billiards auxiliary training system of the present invention.

[0015] Figure 2 The figure is a schematic diagram of the camera positions of the billiards auxiliary training system of the present invention.

[0016] Figure 3 It is a schematic diagram of the hitting auxiliary line of the billiards auxiliary training system of the present invention. DETAILED DESCRIPTION

[0017] The specific implementation modes of the present invention are described below in conjunction with the accompanying drawings and examples. However, the following examples are only used to illustrate the present invention in detail and do not limit the scope of the present invention in any way.

[0018] Example 1: A vision-based billiards auxiliary training system, see Figures 1 to 3 The system consists of a processor, an image acquisition module, a projection device, and a speaker, and has functions such as motion diagnosis, standard motion recording, and voice correction exercises. The camera is connected to the processor through a switch, which acquires camera images in real time and performs image processing, and recognizes and evaluates the posture and motion of the trainee. The processor is connected to a touch screen, a projection device, and a speaker. The trainee scans the code to log in, selects functions, obtains videos, obtains diagnostic reports, and other operations through the touch screen. The projection device is used to project lines and ball positions onto the billiard table, and the speaker is used to broadcast voice prompts to the trainee.

[0019] The above-mentioned processor has a built-in image analysis module, which analyzes the trainee's posture and movement through visual technology. The visual technology uses the posture estimation algorithm OpenPose to extract the coordinates of the key points of the human body, normalize them, eliminate redundant points, and construct a feature vector that integrates the key point position, distance, joint angle and key point speed according to the characteristics of the movement. The action separator model is trained through a multi-layer perception mechanism.

[0020] Figure 2 Camera 1 in the middle captures the image of the billiard table surface. When the trainee hits the five-point ball along direction 1, camera 6 captures the image directly in front of the trainee, camera 3 captures the image above the trainee's head, and camera 2 captures the image from the side of the trainee. When the trainee hits the five-point ball along direction 2, camera 7 captures the image directly in front of the trainee, camera 4 captures the image above the trainee's head, and camera 5 captures the image from the side of the trainee.

[0021] When using the above-mentioned billiards auxiliary training system for training, the following steps are included: Step 1: The trainee logs into the system by scanning the QR code on the touch screen, authorizes the login (WeChat QR code scanning, mobile phone number verification code, etc.), and establishes databases for different logged-in users to record various data entered and stored in the system.

[0022] Step 2: The processor controls the projection device to project the center line, the tee line, and the line connecting the midpoints of the diagonal pockets on the tabletop. Figure 3 , project the target ball position at the center point of the midline, and project the white ball position at the intersection of the kick-off line and the midpoint of the diagonal pocket. The trainee places the target ball and the white ball correctly according to the projection and voice prompts.

[0023] Step 3: The trainee performs standard movements under the guidance of the coach. The coach selects "Start Recording" on the touch screen. The processor obtains the image of the billiard table in real time through the image acquisition module, identifies and records the position of the club head and the direction of the club. When the direction of the club coincides with direction 1, the processor reads the images of camera 2, camera 3, and camera 6 to extract the key points of the human body posture. When the direction of the club coincides with direction 2, the processor reads the images of camera 4, camera 5, and camera 7 to extract the key points of the human body posture. The extracted key points of the human body include the position of the center of gravity foot, the position of the auxiliary foot, the left knee, the right knee, the grip point, the wrist, elbow, and shoulder of the grip hand, and the hand support point, wrist, elbow, and shoulder position of the auxiliary hand. The above positions are stored as the key points of the standard movement. After the key points of the standard movement are stored, a voice prompt "Entry Completed" is given.

[0024] Step 4: The trainer selects the "Voice Correction Practice" function on the touch screen. The specific process is as follows: (1) The processor obtains the image of the billiard table in real time through the image acquisition module and identifies the direction of the cue. When the cue is basically aligned with the goal line, the processor announces through the loudspeaker "Practice begins, please get in position". The processor controls the projection device to project the position of the center of gravity foot and auxiliary foot on the ground, and the trainee gets both feet into position.

[0025] (2) The processor obtains the trainee's image through the image acquisition module and extracts the positions of the center of gravity foot and the auxiliary foot. When the trainee is in the correct position, the processor announces through the loudspeaker "Correct position, please assume the pre-hitting posture". At the same time, the processor controls the projection device to project the club head position, auxiliary hand support point, wrist, elbow and shoulder positions corresponding to the standard action on the table. The trainee then adjusts his posture to meet the standard position.

[0026] (3) The processor obtains the trainee's image through the image acquisition module and extracts the positions of the left knee, right knee, grip point, wrist, elbow, and shoulder of the gripping hand, and compares them with the positions of the key points of the standard movement. The processor prompts the trainee to make adjustments through voice and guides the trainee to assume the correct posture. When all the movements are correct, the processor announces "The movement is correct, please stop before hitting the ball."

[0027] (4) After a delay of 1 second, the voice broadcasts “Please start moving the bar 2-6 times, and stop for 1 second the second time”. The processor obtains the image of the trainee through the image acquisition module and starts storing it.

[0028] (5) The processor analyzes the acquired image of the trainee and, after detecting the second forward stop, announces by voice: “Prepare for hitting the ball, pause for 1 second after pulling the rod.”

[0029] (6) The processor acquires the desktop image through the image acquisition module. When it detects the collision between the target ball and the white ball, it announces "Please do not stand up" by voice. After a delay of 1.5 seconds, it announces "Please stand up" by voice. The processor extracts the motion trajectory of the club through image analysis, compares it with the goal line, and evaluates the result of the shot as "correct", "left" or "right", and announces it by voice.

[0030] (7) The processor stops acquiring images and stores the acquired video in the processor.

[0031] In the third step, the system performs the input function: ① The coach teaches and helps the students to take the correct posture and action, and enters the database as a template for subsequent practice and a basis for diagnosis and correction. The input function is activated when the club coincides with the goal line, and the voice broadcast "Input starts, please take the standard action", click to confirm the input option, the voice broadcast "Complete input" and save, and the input function will be activated again when the club coincides with the goal line again.

[0032] ② Place any sphere on the table and record it for storage. When practicing, select the recorded sphere and project the positions of different balls in the sphere onto the table; ③Storage and sharing: diagnostic videos and practice videos can be stored, shared and forwarded; ④ Mark the important joints involved in the goal line and action. For example, it will be green when the position is correct and red when it is wrong. It will turn green after the correct adjustment. In other words, when all nodes are green, it is a correct action.

[0033] In the fourth step (2) and (3), the system performs diagnostic functions, which can be diagnosed from two aspects: posture and stroke: 1. Position and posture ①Whether the center foot is standing on the extension line of the goal line. By default, the center of the foot can be used as the reference point. The left and right deviations can be temporarily set to ±5cm to diagnose whether the stance is biased to the left or right; ② The distance between the sole of the foot and the white ball is approximately equal to the distance between the hand holding the club and the white ball (it can also be the front-to-back distance between the sole of the foot and the extended line of the hand holding the club perpendicular to the ground). It can be initially set to no more than 5cm, which can be diagnosed as a forward or backward stance; ③ The ratio of the horizontal distance between the two feet to the shoulder width can be set to 1:1~1.2 as the correct range by default, which can be diagnosed as the distance between the two feet being too close or too far; ④ Whether the forearm is perpendicular to the foot surface can be diagnosed as the forearm is biased forward or backward; ⑤ Whether the distance between the hand support point and the white ball is between 25 and 30, it can be diagnosed as the hand support is biased forward or backward; ⑥ After lying down, check whether the hand support point, head / chin, wrist / base of thumb, elbow, and shoulder are aligned with the goal line at the same time, and diagnose which joint is biased to the left or right; ⑦ After lying down, whether the angle between the cue and the table is greater than 20° (or whether the vertical distance between the middle of the cue and the edge of the table is greater than 3cm as a reference), it can be diagnosed as lying down too high; ⑧ After lying down, whether the angle between the line connecting the two shoulders and the goal line is between 40° and 50° can be diagnosed as insufficient or excessive separation of the two shoulders; ⑨ After lying down, check whether the angle between the upper arm and the forearm of the pole holder is between 130° and 150°, which can be diagnosed as excessive or insufficient bending of the forearm of the pole holder; 2. Rod straightness ① Release the rod Ⅰ After lying down, whether the club is aligned with the goal line can be diagnosed as the club is not aligned with the goal line, the club head is a few centimeters to the left, and the club position is a few centimeters off; Ⅱ If the distance between the club head and the center of the white ball is greater than 2cm, it can be diagnosed that the club head is too far away from the white ball; Ⅲ Whether the angle between the cue and the table is greater than 20° (theoretically, the closer to zero, the better. In this field, 5°~15° is generally considered to be the standard range, which can be relaxed a little and temporarily use 20° as a reference). This can diagnose that the cue is uneven and the angle between it and the table is too large; (Here, it can also be judged by whether the vertical distance between the middle of the cue and the edge of the table is greater than 2cm); Ⅳ After the body lies down, whether the club is in contact with the chin and chest can be diagnosed as: not in contact with the chin, not in contact with the chest, and not in contact with the chin and chest.

[0034] ②Rod movement ⅠWhether the angle between the club and the table changes during the cueing process can be diagnosed as uneven cueing; ⅡWhether the club is aligned with the goal line during dribbling can be diagnosed as the club is not straight, the club is bent inward or the club is turned outward; III. Whether the number of club moves is between 2 and 6 times, it can be diagnosed as too few or too many club moves; IV. Whether there are three stops in the club movement or the three stops are incorrect, it can be diagnosed as disordered rhythm of club movement (first stop: after climbing down, the club is placed at the hitting point and remains still for 0.5~1 second to determine the hitting point and direction; second stop: after the club movement is completed, the club stops at the starting point of the first stop and remains still for 0.5~1 second. The role is to repeatedly determine the hitting point and direction and relax the muscles; third stop: when the club head is farthest from the white ball, it is also when the angle between the forearm and the upper arm is the largest, and it is kept for 0.5~1 second. The role is to keep focused and accumulate power to shoot.

[0035] ③ Shoot out Ⅰ After the shot (the time interval can be 0.5s after the white ball touches the target ball) whether the club remains aligned with the goal line can be diagnosed as the shot is not straight, the shot deviates to the left, or the shot deviates to the right; Ⅱ After the club is released (0.5s after the white ball collides with the target ball), if the club remains stationary for more than 0.5s, it can be diagnosed as standing up too quickly.

[0036] The present invention has been described in detail above with reference to the accompanying drawings and embodiments. However, those skilled in the art will appreciate that, without departing from the spirit of the present invention, the specific parameters in the above embodiments may be changed to form a plurality of specific embodiments, which are all within the common variation range of the present invention and will not be described in detail herein.

Claims

1. A vision-based billiards auxiliary training system, characterized in that: include: A projection module is used to project the placement of the white ball, the target ball and the assisting line for hitting the ball on the tabletop of the billiard table, and is also used to project the assisting standing position on the ground; The image acquisition module is used to capture images of the desktop and the top, front, and sides of the trainee. The image analysis module is used to analyze the trainee's posture and movements through visual technology and compare them with standard movements. The output module is used to output movement diagnosis reports and voice correction reminders.

2. The vision-based billiards auxiliary training system according to claim 1, characterized in that: The image acquisition module includes seven cameras, which are located at one camera directly above the center of the billiard table, four cameras above the extended lines of the diagonal of the billiard table, and two cameras on both sides of the width direction of the billiard table.

3. The vision-based billiards auxiliary training system according to claim 1, characterized in that: The image analysis module includes a posture estimation algorithm for extracting coordinates of key points of the human body. The key points of the human body include the center of gravity foot position, auxiliary foot position, left knee, right knee, grip point, wrist, elbow, and shoulder position of the gripping hand, and hand support point, wrist, elbow, and shoulder position of the auxiliary hand.

4. The vision-based billiards auxiliary training system according to claim 1, characterized in that: The output module includes a display screen and a speaker. The display screen can obtain the trainee's training video and diagnosis report, and the speaker can broadcast voice prompt information to the trainee.

5. The vision-based billiards auxiliary training system according to claim 1, characterized in that: It also includes a terminal device for wirelessly receiving and storing training videos and diagnostic reports.

6. A vision-based billiards assisted training method, using the vision-based billiards assisted training system according to claim 1, characterized in that: The following steps are involved: S1: Preparation before training. The projection module projects the placement of the white ball and the target ball, as well as the assisting line for hitting the ball, on the billiard table, and projects the assisting standing position. The trainee places the target ball and the white ball correctly according to the projection and voice prompts. S2: Standard movement recording. The trainee takes position according to the auxiliary standing position and performs standard movements under the guidance of the coach. The image acquisition module and the image analysis module record the standard movements into the system. S3: Posture correction. The trainee takes position again and assumes the pre-hitting posture. The image analysis module identifies the key points of the pre-hitting posture and compares it with the standard movement recorded in S2. The trainee adjusts the posture to match the standard movement. S4: Batting training. Batting includes releasing the club, moving the club, and striking the club. During the batting process, the image analysis module analyzes in real time whether the batting movement is standard.

7. The vision-based billiards auxiliary training method according to claim 6, characterized in that: In step S2, when it is detected that the direction of the club coincides with the goal line, the standard action is recorded, and the image analysis module extracts and stores the key points in the standard action.

8. The vision-based billiards auxiliary training method according to claim 6, characterized in that: In step S3, after the trainee stands in the correct position, the projection module projects the club head position, auxiliary hand support point, wrist, elbow and shoulder positions corresponding to the standard action on the desktop, and the trainee adjusts his posture to meet the standard position.

9. The vision-based billiards auxiliary training method according to claim 6, characterized in that: In step S4, after the posture correction, the bar is moved 2-6 times, and the trainer stops forward for a period of time. The image acquisition module obtains the trainee's image and starts storing it.

10. The vision-based billiards auxiliary training method according to claim 6, characterized in that: In step S4, when hitting the ball, the image acquisition module obtains the desktop image, extracts the movement trajectory of the club through image analysis, compares it with the goal line, and evaluates the result of the shot as "correct", "left-biased" or "right-biased".

Images