Player and ball tracking detection method and system, electronic equipment and storage medium

Abstract

The invention discloses a player and ball tracking detection method and system, electronic equipment and a storage medium, and relates to the technical field of ball game video analysis, and the method comprises the steps: carrying out the preprocessing of a preset ball game video, and obtaining a plurality of target video segments; and analyzing each target video clip to obtain a player tracking detection result and a ball tracking detection result corresponding to each target video clip. According to the invention, the preset ball game video can be automatically analyzed, the dependence on manual operation is reduced, the objectivity and accuracy of the obtained player tracking detection result and the ball tracking detection result are improved, and a coach and a player can conveniently and rapidly track the ball game by analyzing the player tracking detection result and the ball tracking detection result. And optimizing a training strategy and a competition strategy.

Description

Background Art

[0002] Badminton belongs to the category of technical and tactical confrontation. The technical level and tactical application ability of badminton players largely determine whether they can win the game. With the continuous improvement of the competitive level of world badminton, the research on the application of players' techniques and tactics in the game has been paid more and more attention by coaches and players. The research on the characteristics of excellent badminton players' game techniques and tactics and the research on the connection characteristics of each shot in the round have enriched the training methods and means of players to a certain extent. However, there are still many practical problems in the practice of using computer technology and software systems to efficiently analyze badminton techniques and tactics. There are still many deficiencies and they need to be improved and supplemented. Summary of the invention

[0003] The technical problem to be solved by the present invention is to address the deficiencies of the prior art, and specifically provide a method, system, electronic device and storage medium for tracking and detecting a player and a ball, as follows:

[0004] 1) In the first aspect, the present invention provides a method for tracking and detecting a player and a ball, and the specific technical solution is as follows:

[0005] Preprocessing the preset ball sports video to obtain multiple target video clips;

[0006] Each target video clip is analyzed to obtain a player tracking detection result and a ball tracking detection result corresponding to each target video clip.

[0007] The beneficial effects of the player and ball tracking detection method provided by the present invention are as follows:

[0008] It can automatically analyze preset ball sports videos and reduce reliance on manual operations, thereby improving the objectivity and accuracy of the player tracking detection results and the ball tracking detection results, so that coaches and players can optimize training strategies and game strategies by analyzing the player tracking detection results and the ball tracking detection results.

[0009] Based on the above solution, the player and ball tracking and detection method of the present invention can also be improved as follows.

[0010] Furthermore, it also includes:

[0011] Statistical analysis is performed on all player tracking detection results and all ball tracking detection results to obtain the distribution of ball landing points, player hitting technical movements and return ball routes corresponding to each target video clip, and send them to the preset smart terminal.

[0012] The beneficial effect of adopting the above-mentioned further technical solution is that by performing statistical analysis on all player tracking detection results and all ball tracking detection results, more data can be obtained, namely the distribution of ball landing points, player hitting technical movements and return ball routes corresponding to each target video clip, so that coaches, players and other personnel can view and analyze through preset smart terminals, providing more data support for optimizing training strategies and game strategies.

[0013] Furthermore, it also includes:

[0014] Establishing an index relationship between each target video clip and the corresponding ball landing point distribution, player's ball hitting technique and ball return route;

[0015] When a user views any ball landing point distribution, any player's ball hitting technique action or any ball return route through a preset smart terminal, the target video clip corresponding to the ball landing point distribution, the target video clip corresponding to the player's ball hitting technique action or the target video clip corresponding to the ball return route is called through the index relationship.

[0016] The beneficial effect of adopting the above further technical solution is that it is more convenient for coaches, players and other personnel to view relevant target video clips through index relationships, which is highly flexible and practical.

[0017] Further, each target video clip is analyzed, including: analyzing each target video clip based on artificial intelligence.

[0018] The beneficial effect of adopting the above further technical solution is that by using artificial intelligence, player tracking detection and ball tracking detection can be performed accurately and efficiently.

[0019] Furthermore, the process of obtaining the player tracking detection result corresponding to each target video clip includes:

[0020] Using the trained person detection model, perform person tracking detection on each target video clip to obtain the person tracking detection result corresponding to each target video clip;

[0021] According to the field range of the stadium, the human tracking detection results are filtered to obtain the player tracking detection results corresponding to each target video clip.

[0022] The beneficial effect of adopting the above further technical solution is that: the player tracking detection results of non-players are filtered out, and the player tracking detection results corresponding to each target video clip can be accurately obtained.

[0023] Furthermore, the process of obtaining the ball tracking detection result corresponding to each target video clip includes:

[0024] Using the trained ball detection and tracking model, ball detection and tracking are performed on each target video clip to obtain the ball tracking detection result corresponding to each target video clip.

[0025] Further, the preset ball sports videos are: badminton match videos, badminton training videos, table tennis match videos, table tennis training videos, tennis match videos or tennis training videos.

[0026] The beneficial effect of adopting the above further technical solution is that the present invention is applicable to sports videos of various ball games and has strong applicability.

[0027] 2) In the second aspect, the present invention also provides a player and ball tracking and detection system, and the specific technical solution is as follows:

[0028] Including video preprocessing module and video analysis module;

[0029] The video preprocessing module is used to: preprocess the preset ball sports video to obtain multiple target video clips;

[0030] The video analysis module is used to analyze each target video segment to obtain a player tracking detection result and a ball tracking detection result corresponding to each target video segment.

[0031] Based on the above solution, the player and ball tracking and detection system of the present invention can also be improved as follows.

[0032] Furthermore, it also includes a data analysis module, which is used to: perform statistical analysis on all player tracking detection results and all ball tracking detection results, obtain the distribution of hitting point, player hitting technical movements and return ball routes corresponding to each target video clip, and send them to a preset smart terminal.

[0033] Furthermore, an index module is included, and the index module is used to:

[0034] Establishing an index relationship between each target video clip and the corresponding ball landing point distribution, player's ball hitting technique and ball return route;

[0035] When a user views any ball landing point distribution, any player's ball hitting technique action or any ball return route through a preset smart terminal, the target video clip corresponding to the ball landing point distribution, the target video clip corresponding to the player's ball hitting technique action or the target video clip corresponding to the ball return route is called through the index relationship.

[0036] Furthermore, the video analysis module is used to analyze each target video clip based on artificial intelligence.

[0037] Furthermore, the video analysis module is also specifically used for:

[0038] Using the trained person detection model, perform person tracking detection on each target video clip to obtain the person tracking detection result corresponding to each target video clip;

[0039] According to the field range of the stadium, the human tracking detection results are filtered to obtain the player tracking detection results corresponding to each target video clip.

[0040] Furthermore, the video analysis module is also specifically used for:

[0041] Using the trained ball detection and tracking model, ball detection and tracking are performed on each target video clip to obtain the ball tracking detection result corresponding to each target video clip.

[0042] Further, the preset ball sports videos are: badminton match videos, badminton training videos, table tennis match videos, table tennis training videos, tennis match videos or tennis training videos.

[0043] 3) In a third aspect, the present invention further provides an electronic device, comprising a processor, the processor being coupled to a memory, the memory storing at least one computer program, the at least one computer program being loaded and executed by the processor, so that the electronic device implements any of the above-mentioned player and ball tracking and detection methods.

[0044] 4) In a fourth aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned player and ball tracking and detection methods.

[0045] It should be noted that the beneficial effects achieved by the technical solutions of the second to fourth aspects of the present invention and the corresponding possible implementation methods can be found in the above-mentioned technical effects of the first aspect and its corresponding possible implementation methods, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments of the present invention:

[0047] Figure 1 A schematic flow chart of a method for tracking and detecting a player and a ball according to an embodiment of the present invention;

[0048] Figure 2 A schematic diagram of a court, players, and non-players in a badminton singles match video;

[0049] Figure 3 is a schematic diagram of the positions of each corner point;

[0050] Figure 4Schematic diagram of the distribution of the positions of key points of the human body;

[0051] Figure 5 A schematic diagram of the net position line;

[0052] Figure 6 Schematic diagram of the network structure of the constructed deep learning model;

[0053] Figure 7 It is a structural schematic diagram of a player and ball tracking and detection system according to an embodiment of the present invention;

[0054] Figure 8 The figure is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The principles and features of the present invention are described below. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0056] The technical solution of the present invention and how the technical solution of the present invention solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present invention will be described below in conjunction with the accompanying drawings.

[0057] like Figure 1 As shown, a method for tracking and detecting a player and a ball according to an embodiment of the present invention comprises the following steps:

[0058] S1. Preprocessing a preset ball sports video to obtain a plurality of target video clips, specifically:

[0059] The perspective switching segments, playback segments and advertising segments in the preset ball sports video are removed, and the remaining parts in the preset ball sports video are split according to the effective rounds to obtain multiple target video segments, each of which includes a effective round.

[0060] Among them, the effective rounds are explained as follows:

[0061] For example, in table tennis and badminton games and training, a round refers to: it refers to the last moment before the ball is intentionally thrown up while it is at rest on the palm of the hand not holding the racket, until the round is judged to be a score or the serve is re-made. In the present invention, a valid round refers to: a round in which a score is judged. In other ball games and training, a valid round can also be considered as: a round in which a score is judged.

[0062] S2. Analyze each target video clip to obtain a player tracking detection result and a ball tracking detection result corresponding to each target video clip.

[0063] The beneficial effects of the player and ball tracking detection method provided by the present invention are as follows:

[0064] It can automatically analyze preset ball sports videos and reduce reliance on manual operations, thereby improving the objectivity and accuracy of the player tracking detection results and the ball tracking detection results, so that coaches and players can optimize training strategies and game strategies by analyzing the player tracking detection results and the ball tracking detection results.

[0065] Optionally, in the above technical solution, it also includes:

[0066] S3. Statistically analyze all player tracking detection results and all ball tracking detection results to obtain the distribution of ball landing points, player hitting technical movements and ball return routes corresponding to each target video clip, and send them to a preset smart terminal.

[0067] Optionally, in the above technical solution, it also includes:

[0068] S4, establishing an index relationship between each target video clip and the corresponding ball hitting point distribution, the player's ball hitting technical action and the ball return route;

[0069] S5. When the user views any ball landing point distribution, any player's ball hitting technique action or any ball return route through a preset smart terminal, the target video clip corresponding to the ball landing point distribution, the target video clip corresponding to the player's ball hitting technique action or the target video clip corresponding to the ball return route is called through the index relationship.

[0070] Optionally, in S2, analyzing each target video segment includes: analyzing each target video segment based on artificial intelligence.

[0071] Optionally, in the above technical solution, the process of obtaining the player tracking detection result corresponding to each target video clip includes:

[0072] S20, using the trained person detection model, performing person tracking detection on each target video clip to obtain a person tracking detection result corresponding to each target video clip;

[0073] S21. Filter the human tracking detection results according to the field range of the court 3 to obtain the player tracking detection results corresponding to each target video clip.

[0074] Among them, the process of obtaining a trained person detection model includes: constructing multiple person samples based on historical ball sports videos, the person samples include person images and labels (for example, labeling human images as 1 and non-human images as 0), and then training the first preset network model to obtain a trained person detection model.

[0075] Among them, the first preset network model is R-CNN network or CNN (convolutional neural network), etc., and other network models can also be selected according to actual conditions.

[0076] In another embodiment, the first preset network model may be directly trained with an existing dataset including multiple pedestrian images to obtain a trained human detection model. The existing dataset including multiple pedestrian images may be the UCF-QNRF dataset released by the University of Florida or the CUHK Occlusion dataset released by the Chinese University of Hong Kong.

[0077] Optionally, in the above technical solution, the process of obtaining the ball tracking detection result corresponding to each target video clip includes:

[0078] S22. Using the trained ball detection and tracking model, perform ball detection and tracking on each target video clip to obtain a ball tracking detection result corresponding to each target video clip.

[0079] The process of obtaining the trained ball detection and tracking model includes:

[0080] A plurality of ball position samples are constructed based on historical ball sports videos. The ball position samples include images and labels of the balls. Specifically, the positions of the balls can be used as labels. Then, the second preset network model is trained to obtain a trained ball detection and tracking model.

[0081] Among them, the second preset network model can be an R-CNN network or CNN (convolutional neural network), DNN (deep neural network), etc., and other deep learning networks can also be selected according to actual conditions.

[0082] Optionally, in the above technical solution, the preset ball sports video is: a badminton match video, a badminton training video, a table tennis match video, a table tennis training video, a tennis match video or a tennis training video.

[0083] Among them, the badminton match video can be a badminton singles match video or a badminton match video can be a badminton doubles match video, the badminton training video can be a badminton singles match video or a badminton singles training video can be a badminton doubles training video, the table tennis match video can be a table tennis singles match video or a table tennis doubles match video, the table tennis training video can be a table tennis singles training video or a table tennis doubles training video, the tennis match video can be a tennis singles match video or a tennis doubles match video, and the tennis training video can be a tennis singles training video or a tennis doubles training video.

[0084] Taking a badminton singles match video as an example, a player and ball tracking detection method of the present invention is described, which includes the following steps:

[0085] S101, obtaining a badminton singles match video, specifically:

[0086] A badminton singles match video is shot by a camera in mp4 file or avi format. When shooting the badminton singles match video, refer to the main perspective of the World Badminton Federation match broadcast, place the camera or video camera directly behind one side of the court, and shoot from top to bottom. The lens direction of the camera or video camera is approximately the same as the direction of the court, and the shooting perspective of the camera or video camera is fixed. The shot badminton singles match video will contain a court 3, players and non-players 5, etc., wherein the non-players 5 include referees, coaches and off-court spectators, etc., wherein the number of players participating in the game is 2, respectively recorded as a first player 1 and a second player 2, the first player 1 is a player far away from the camera or video camera, and the second player 2 is a player close to the camera or video camera. In the present invention, the first player 1 can also be referred to as an upper player, and the second player 2 can also be referred to as a lower player, such as Figure 2 shown.

[0087] In another embodiment, the badminton singles game video may also be directly obtained from the network.

[0088] S102, pre-processing the badminton singles match video, specifically:

[0089] S1020. Remove perspective switching clips, playback clips, and advertising clips from the badminton singles match video.

[0090] S1021, removing invalid video frames:

[0091] like Figure 3As shown, the four corner point positions of the singles hitting area in the badminton singles game video are recorded as the first court corner point position 6, the second court corner point position 7, the third court corner point position 8 and the fourth court corner point position 9 respectively. It should be noted that for the badminton doubles game video, the four corner point positions of the doubles hitting area in the badminton singles game video are recorded as the first court corner point position 6, the second court corner point position 7, the third court corner point position 8 and the fourth court corner point position 9 respectively. Other ball games can be set according to actual conditions, and the four corner point positions of the net 4 are recorded as the first net corner point position 10, the second court corner point position 7, the third court corner point position 8 and the fourth court corner point position 9 respectively. The second net corner point position 11, the third net corner point position 12 and the fourth net corner point position 13 are detected by using the trained court 3 detection model to remove the perspective switching segment, the playback segment and the advertising segment, etc., in each video frame of the badminton game video, and the coordinates of the first court corner point position 6, the coordinates of the second court corner point position 7, the coordinates of the third court corner point position 8, the coordinates of the fourth court corner point position 9, the coordinates of the first net corner point position 10, the coordinates of the second net corner point position 11, the coordinates of the third net corner point position 12 and the coordinates of the fourth net corner point position 13 in each video frame are obtained, as shown in FIG. Figure 3 shown.

[0092] According to the coordinates of the first court corner point position 6, the second court corner point position 7, the third court corner point position 8, the fourth court corner point position 9, the first net corner point position 10, the second net corner point position 11, the third net corner point position 12 and the fourth net corner point position 13 in each two adjacent video frames, the distance between the first court corner point positions 6, the distance between the second court corner point positions 7, the distance between the third court corner point positions 8, the distance between the fourth court corner point positions 9, the distance between the first net corner point positions 10, the distance between the second net corner point positions 11, the distance between the third net corner point positions 12 and the distance between the fourth net corner point positions 13 in each two adjacent video frames are calculated.

[0093] According to the actual situation, preset the distance thresholds corresponding to each first court corner point position 6, second court corner point position 7, third court corner point position 8, fourth court corner point position 9, first net corner point position 10, second net corner point position 11, third net corner point position 12 and fourth net corner point position 13 respectively, and judge whether the distance between the first court corner point positions 6, the distance between the second court corner point positions 7, the distance between the third court corner point positions 8, the distance between the fourth court corner point positions 9, the distance between the first net corner point positions 10, the distance between the second net corner point positions 11, the distance between the third net corner point positions 12 and the distance between the fourth net corner point positions 13 in each two adjacent video frames are all less than the corresponding preset distance thresholds; if not, judge that the latter video frame of the two adjacent video frames is an invalid video frame and remove it.

[0094] The process of obtaining the trained court 3 detection model includes:

[0095] A plurality of court 3 and net 4 samples are constructed based on historical ball sports videos, and then the third preset network model is trained to obtain a trained court 3 detection model.

[0096] Among them, the third preset network model can be an R-CNN network or CNN (convolutional neural network), DNN (deep neural network), etc., and other deep learning networks can also be selected according to actual conditions.

[0097] S1022, obtaining multiple target video clips:

[0098] The badminton singles game video processed by S1020 and S1021 is split according to valid rounds to obtain multiple target video segments, each of which includes a valid round.

[0099] S103, tracking and detecting the players and the ball:

[0100] 1) First, use the trained person detection model obtained by training the R-CNN network to detect each video frame in each target video clip to obtain the person tracking detection result corresponding to each target video clip. The person tracking detection result corresponding to each target video clip includes: the center point position of all human body key points 14 of each person in each video frame in the target video clip. The human body key points 14 include: hands, neck, elbows, shoulders, waist, knees, feet, etc., which can be set according to actual conditions, such as Figure 4 shown.

[0101] Then, according to the field range of the court 3, the person tracking detection result is filtered to obtain the player tracking detection result corresponding to each target video clip, specifically:

[0102] The field range of the court 3 is multiplied by the expansion coefficient δ to obtain the filter area thres_area. The four corner point positions of the filter area thres_area are recorded as the first filter area corner point position, the second filter area corner point position, the third filter area corner point position and the fourth filter area corner point position. Through the position of the net 4, the positions of all human body key points 14 of people outside the filter area thres_area are filtered out to obtain the player tracking detection result corresponding to each target video clip. The expansion coefficient δ can be set according to actual conditions.

[0103] In another embodiment, the boundary line of the field range of the court 3, i.e., the boundary line of the court 3, can be first identified, and then the boundary line of the field range of the court 3, i.e., the boundary line of the court 3, can be used to distinguish between players and non-players 5. At this time, the expansion coefficient δ=1, and then the trained person detection model obtained by training the R-CNN network is used to detect each video frame in each target video clip to obtain the player tracking detection result corresponding to each target video clip.

[0104] The first player 1 and the second player 2 can also be distinguished by the net position line 16 .

[0105] like Figure 5 As shown, the net position line 16 is explained as follows:

[0106] ① After identifying the first net corner point position 10, the second net corner point position 11, the third net corner point position 12 and the fourth net corner point position 13, the straight line formed by projecting the first net corner point 10 and the second net corner point 11 onto the court 3 is the net position line 16, or the straight line formed by projecting the third net corner point 12 and the fourth net corner point 13 onto the court 3 is the net position line 16.

[0107] ② Identify the landing points 15 of the two net posts of the net 4, and the straight line connecting the two landing points 15 is the net position line 16.

[0108] In another embodiment, the positions of all human body key points 14 of each player in each video frame in each target video segment may be used as the player tracking detection result corresponding to the corresponding target video segment.

[0109] 2) Using the trained ball detection and tracking model obtained by training the constructed deep learning model, each video frame in each target video clip is detected to obtain the ball tracking detection result corresponding to each target video clip. The ball tracking detection result corresponding to each target video clip includes: the position of the badminton in each video frame in the target video clip.

[0110] Among them, Figure 6 As shown, the constructed deep learning model includes an input layer (not shown) arranged in sequence, a first Multi-Conv2d Concat17 (a combination of multiple convolutional layers Conv2d and concatenation layers Concat), a first pooling layer 18 (pooling layer: Max Pooling), a second Multi-Conv2d Concat19, second pooling layer 20, third Multi-Conv2dConcat21, third pooling layer 22, first Triple-Conv2d23 (Triple-Conv2d is a combination of multiple convolutional layers), first upsampling layer 24 (upsampling: Upsampling), first Double-Con2d25 (Double-Con2d is a combination of two convolutional layers), first channel attention layer 26, second upsampling layer 27, second Double-Con2d28, second channel attention layer 29, third upsampling layer 30, third Double-Con2d31, third channel attention layer 32, Conv2d+Sigmoid (a combination of a convolutional layer followed by a Sigmoid activation function) and output layer (not shown).

[0111] Among them, the input layer is used to receive video frames of size 512 (length) × 288 (width) × 9 (number of channels); the first Multi-Conv2d Concat17, the first pooling layer 18 (pooling layer: Max Pooling), the second Multi-Conv2dConcat19, the second pooling layer 20 and the third Multi-Conv2d Concat21 constitute an encoder; the first upsampling layer 24, the first Double-Con2d25, the first channel attention layer 26, the second upsampling layer 27, the second Double-Con2d28, the second channel attention layer 29, the third upsampling layer 30, the third Double-Con2d31, the third channel attention layer 32 and Conv2d+Sigmoid constitute a decoder.

[0112] Among them, the fourth channel attention layer 34 is skipped after the first Multi-Conv2d Concat17, and the output data of the first Multi-Conv2d Concat17 is input to the fourth channel attention layer 34. The output data of the fourth channel attention layer 34 is input to the third upsampling layer 30 after the concat (channel number superposition) operation.

[0113] Among them, the fifth channel attention layer 35 is skipped after the second Multi-Conv2d Concat19, and the output data of the second Multi-Conv2d Concat19 is input to the fifth channel attention layer 35. The output data of the fifth channel attention layer 35 is input to the second upsampling layer 27 after the concat (channel number superposition) operation.

[0114] Among them, the third Multi-Conv2d Concat21 is also skip-connected to the sixth channel attention layer 36, and the output data of the third Multi-Conv2d Concat21 is input to the sixth channel attention layer 36. The output data of the sixth channel attention layer 36 is input to the first upsampling layer 24 after the concat (channel number superposition) operation.

[0115] The position of each key point 14 of the human body and the position of the ball can be represented by the coordinates in the preset plane coordinate system. The center point of the court 3 can be used as the origin o, and the two directions along the court 3 can be used as the x-axis direction and the y-axis direction to establish a preset plane coordinate system. Then, The coordinates of different key points 14 of the first player 1 at time t can be expressed as The coordinates of different human body key points 14 of the second player 2 at time t can be represented by, for example, Represents: The coordinates of the first player 1's hand at time t, expressed as It represents: the coordinates of the neck of the second player 1 at time t. Similarly, different i values ​​are set for different human key points 14. Time t is the time corresponding to the timestamp of any video frame in all target video clips.

[0116] Optionally, the center point of the coordinates of all the key points 14 of the first player 1 at time t is taken as the coordinates of the first player 1 at time t, denoted as The center point of the coordinates of all the key points 14 of the second player 2 at time t is taken as the coordinates of the second player 2 at time t, denoted as Moreover, the first court corner point position 6, the second court corner point position 7, the third court corner point position 8, the fourth court corner point position 9, the first net corner point position 10, the second net corner point position 11, the third net corner point position 12 and the fourth net corner point position 13 can also be represented by coordinates in a preset plane coordinate system.

[0117] S104. Statistics:

[0118] 1) Get the distribution of ball landing points:

[0119] After obtaining the position of the badminton in each video frame in each target video clip, the position of the badminton at each moment is obtained to form a badminton position time series, and curve fitting is performed on the badminton position time series. The moment corresponding to the trough position on the fitted curve is used as the player's hitting moment. The trough position of any hitting moment is: the position of the badminton at the hitting moment, which is recorded as the hitting point, thereby obtaining the hitting point distribution in each target video clip (the hitting point at each moment).

[0120] Among them, the curve fitting can be performed in the form of a quadratic polynomial, and the curve fitting can be performed twice or more times on the badminton position time series, which can be set according to the actual situation.

[0121] In another embodiment, a first time period is set, and the distribution of the hitting points in the first time period can be obtained according to the position of the badminton in a plurality of consecutive video frames in the first time period, and through the net position line 16, it can be distinguished whether it is the hitting point in the area of ​​the court 3 where the first player 1 is located or the hitting point in the area of ​​the court 3 where the second player 2 is located.

[0122] 2) Get the player's hitting technique:

[0123] After obtaining the center point positions of the first player 1 and the second player 2 in each video frame in each target video clip, the center point position of the first player 1 at each moment and the center point position of the second player 2 at each moment are obtained, forming a time series of the center point position of the first player 1 and a time series of the center point position of the second player 2. After analyzing the time series of the center point position of the first player 1 and the time series of the center point position of the second player 2, the coach can determine the hitting technical movements of the first player 1 and the second player 2.

[0124] In another embodiment, a second time period is set, and the center point positions of the first player 1 and the second player 2 in each video frame in the second time period are obtained, so that a time series of the center point position of the first player 1 in the second time period and a time series of the center point position of the second player 2 in the second time period can be obtained. After analyzing the time series of the center point position of the first player 1 and the time series of the center point position of the second player 2, the coach can determine the hitting technical actions of the first player 1 and the second player 2 in the second preset time period.

[0125] The first time period and the second time period can be set according to actual conditions.

[0126] 3) Get the ball return line:

[0127] The ball return line is obtained by connecting the ball landing points in every two adjacent video frames.

[0128] 4) Obtain the statistical results of the hitting technique and the line combination:

[0129] According to the number of shots in an effective round, the hitting technical movements and return ball routes of each effective round are classified and counted to obtain the statistical results of the hitting technical movements and route combinations.

[0130] The number of beats can be 3, 4, 5, etc., and can be set according to actual conditions.

[0131] All statistical results (distribution of hitting points, player hitting techniques, return routes, hitting technique statistics and route combination statistics) and each target video clip are stored in a database for easy query and management by users (coaches and players, etc.) through preset smart terminals.

[0132] Among them, the batting technical actions, batting landing point distribution, return ball lines, batting technical action statistics and line combination statistics of each effective round of the game can be plotted into a chart and sent to the display device of the preset intelligent terminal for display, and each effective video time of the index technical statistics is provided to players and coaches to view the corresponding video segment, and an index relationship is established between each target video clip and the corresponding batting landing point distribution, player batting technical actions, return ball lines, batting technical action statistics and line combination statistics. When the user (coach and player, etc.) views any batting landing point distribution, any player batting technical action, any return ball line, batting technical action statistics and line combination statistics through the preset intelligent terminal, the corresponding target video clip is called through the index relationship.

[0133] The beneficial effects of the present invention are as follows:

[0134] 1) The present invention relates to the technical field of badminton match video technical and tactical analysis and automatic statistical recording system, which aims to solve the key technical problems existing in traditional video analysis and recording methods. This project is committed to improving the accuracy and efficiency of technical and tactical data analysis and recording. By using visual and artificial intelligence technologies, it can accurately capture and analyze effective rounds in game videos, and quickly and accurately analyze the technical and tactical application performance of players on both sides. The system aims to automate the recording process of sports performance and reduce dependence on manual operations, thereby improving the consistency, objectivity and accuracy of data collection. This project can also greatly improve the efficiency of technical and tactical analysis, and provide effective feedback to coaches and players through rapid processing, analysis and statistical data, thereby optimizing training and competition strategies.

[0135] 2) The present invention can be applied not only to professional competitive sports, helping coaches and players analyze and improve their performance in competitions and training, but also to school sports and mass sports activities, providing technical support for fans' competition and training analysis. In addition, these systems can also be integrated into sports technology products, such as intelligent sports performance analysis software and intelligent fitness equipment, to provide scientific and accurate sports analysis services for the majority of users.

[0136] 3) The present invention aims at badminton game video analysis and constructs an analysis method based on field detection, net 4 detection, player detection + posture and motion recognition and badminton detection algorithm, which provides real-time motion feedback for players and coaches to further form effective training strategies and game strategies.

[0137] 4) Through this method, coaches can help athletes with targeted training and competition preparation, thereby improving the targetedness of training and on-the-spot tactics in competitions.

[0138] 5) This method brings significant performance improvements to related sports analysis and recording products by integrating advanced visual recognition and artificial intelligence technologies. First of all, the application of this technology greatly improves the accuracy and efficiency of data analysis, ensuring that every technical aspect of the athlete's game can be accurately analyzed. This is not only crucial for the optimization of athlete training and game strategies, but also provides coaches with more detailed and reliable data support.

[0139] 6) This method improves the user experience of the product. Automated data recording and real-time analysis functions reduce the user's dependence on professional knowledge, making it easy for even non-professionals to use, thereby broadening the market application scope of the product. In addition, through real-time feedback and analysis, users can adjust training methods in a timely manner and effectively improve training results, which increases the attractiveness and market competitiveness of the product.

[0140] In the above embodiments, although the steps are numbered S1, S2, etc., these are only specific embodiments given by the present invention. Those skilled in the art may adjust the execution order of S1, S2, etc. according to actual conditions, which is also within the protection scope of the present invention. It can be understood that in some embodiments, some or all of the above embodiments may be included.

[0141] like Figure 7 As shown, a player and ball tracking detection system 200 according to an embodiment of the present invention includes a video preprocessing module 201 and a video analysis module 202;

[0142] The video preprocessing module 201 is used to: preprocess the preset ball sports video to obtain multiple target video clips;

[0143] The video analysis module 202 is used to analyze each target video segment to obtain a player tracking detection result and a ball tracking detection result corresponding to each target video segment.

[0144] Optionally, the above technical solution also includes a data analysis module, which is used to: perform statistical analysis on all player tracking detection results and all ball tracking detection results, obtain the distribution of ball landing points, player hitting technical movements and return ball routes corresponding to each target video clip, and send them to a preset smart terminal.

[0145] Optionally, in the above technical solution, an index module is further included, and the index module is used to:

[0146] Establishing an index relationship between each target video clip and the corresponding ball landing point distribution, player's ball hitting technique and ball return route;

[0147] When a user views any ball landing point distribution, any player's ball hitting technique action or any ball return route through a preset smart terminal, the target video clip corresponding to the ball landing point distribution, the target video clip corresponding to the player's ball hitting technique action or the target video clip corresponding to the ball return route is called through the index relationship.

[0148] Optionally, in the above technical solution, the video analysis module 202 is used to analyze each target video segment based on artificial intelligence.

[0149] Optionally, in the above technical solution, the video analysis module 202 is further specifically used for:

[0150] Using the trained person detection model, perform person tracking detection on each target video clip to obtain the person tracking detection result corresponding to each target video clip;

[0151] According to the field range of the court 3, the human tracking detection results are filtered to obtain the player tracking detection results corresponding to each target video clip.

[0152] Optionally, in the above technical solution, the video analysis module 202 is further specifically used for:

[0153] Using the trained ball detection and tracking model, ball detection and tracking are performed on each target video clip to obtain the ball tracking detection result corresponding to each target video clip.

[0154] Optionally, in the above technical solution, the preset ball sports video is: a badminton match video, a badminton training video, a table tennis match video, a table tennis training video, a tennis match video or a tennis training video.

[0155] Optionally, ROS (Robot Operating System) is used to provide a system framework for node management and communication for the system of the present invention, and ROS nodes are created for each functional module of the system of the present invention to implement corresponding functions.

[0156] It should be noted that the beneficial effects of the player and ball tracking detection system 200 provided in the above embodiment are the same as the beneficial effects of the player and ball tracking detection method described above, which will not be described in detail here. In addition, when the system provided in the above embodiment realizes its functions, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the system can be divided into different functional modules according to actual conditions to complete all or part of the functions described above. In addition, the system and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be described in detail here.

[0157] Among them, the player and ball tracking and detection system of the present invention can be a computer program (including program code) running in a computer device. For example, the player and ball tracking and detection system of the present invention is an application software that can be used to execute the corresponding steps in the player and ball tracking and detection method of the present invention.

[0158] In some embodiments, the player and ball tracking and detection system of the present invention can be implemented by a combination of software and hardware. As an example, the player and ball tracking and detection system of the present invention can be a processor in the form of a hardware decoding processor, which is programmed to execute the player and ball tracking and detection method of the present invention. For example, the processor in the form of a hardware decoding processor can adopt one or more application specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field programmable gate arrays (FPGAs), or other electronic components.

[0159] The modules involved in the embodiments of the present invention may be implemented in software or hardware. The name of a module does not limit the module itself in some cases.

[0160] An electronic device according to an embodiment of the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, any of the above-mentioned methods for tracking and detecting a player and a ball is implemented. That is to say, an electronic device according to an embodiment of the present invention may include but is not limited to: a processor and a memory; the memory is used to store the computer program; the processor is used to execute the method for tracking and detecting a player and a ball shown in any embodiment of the present invention by calling the computer program.

[0161] In an alternative embodiment, an electronic device is provided, such as Figure 8 As shown, Figure 8 The electronic device 4000 shown includes: a processor 4001 and a memory 4003. The processor 4001 and the memory 4003 are connected, such as through a bus 4002. Optionally, the electronic device 4000 may also include a transceiver 4004, which may be used for data interaction between the electronic device and other electronic devices, such as data transmission and / or data reception. It should be noted that in actual applications, the transceiver 4004 is not limited to one, and the structure of the electronic device 4000 does not constitute a limitation on the embodiments of the present invention.

[0162] Processor 4001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of the present invention. Processor 4001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0163] The bus 4002 may include a path to transmit information between the above components. The bus 4002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 4002 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 8In the figure, only one thick line is used to represent the bus 4002, but this does not mean that there is only one bus or one type of bus.

[0164] The memory 4003 can be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited to this.

[0165] The memory 4003 is used to store the application code (computer program) for executing the solution of the present invention, and the execution is controlled by the processor 4001. The processor 4001 is used to execute the application code stored in the memory 4003 to implement the content shown in the above method embodiment.

[0166] Among them, the electronic device can also be a terminal device, and the terminal device can be any device that can install applications, including at least one of a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, a smart TV, and a smart car device.

[0167] It should be noted that Figure 8 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0168] A computer-readable storage medium according to an embodiment of the present invention stores a computer program, and when the computer program is executed by a processor, any of the above-mentioned methods for tracking and detecting a player and a ball is implemented.

[0169] Optionally, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc (CD-ROM), a magnetic tape, a floppy disk, an optical data storage device, and the like.

[0170] In an exemplary embodiment, a computer program product or a computer program is also provided, the computer program product or the computer program includes computer instructions, the computer instructions are stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the electronic device performs any of the above-mentioned player and ball tracking detection methods.

[0171] Computer program code for performing the operations of the present invention may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0172] It should be understood that the flow charts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the module, the program segment, or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a sequence different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0173] The computer-readable storage medium provided by the embodiment of the present invention may be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, device, or component.

[0174] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device executes the method shown in the above embodiment.

[0175] The above description is only a preferred embodiment of the present invention and an explanation of the technical principles used. Those skilled in the art should understand that the disclosure scope involved in the present invention is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present invention (but not limited to) by each other.

[0176] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects and represent the definition of a specific order or sequence. The order of use of similar objects can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than the order shown or described.

[0177] Those skilled in the art know that the present invention can be implemented as a system, method or computer program product. Therefore, the present invention can be specifically implemented in the following forms, namely: it can be complete hardware, it can be complete software (including firmware, resident software, microcode, etc.), or it can be a combination of hardware and software, which is generally referred to as "circuit", "module" or "system" herein. In addition, in some embodiments, the present invention can also be implemented in the form of a computer program product in one or more computer-readable media, and the computer-readable medium contains computer-readable program code.

[0178] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A method for tracking and detecting a player and a ball, characterized in that: include: Preprocessing the preset ball sports video to obtain multiple target video clips; Each target video clip is analyzed to obtain a player tracking detection result and a ball tracking detection result corresponding to each target video clip.

2. A player and ball tracking detection method according to claim 1, characterized in that: Also includes: Statistical analysis is performed on all player tracking detection results and all ball tracking detection results to obtain the distribution of ball landing points, player hitting technical movements and return ball routes corresponding to each target video clip, and send them to the preset smart terminal.

3. A player and ball tracking detection method according to claim 2, characterized in that: Also includes: Establishing an index relationship between each target video clip and the corresponding ball landing point distribution, player's ball hitting technique and ball return route; When the user views any ball landing point distribution, any player's ball hitting technical action or any ball return route through the preset smart terminal, the target video clip corresponding to the ball landing point distribution, the target video clip corresponding to the player's ball hitting technical action or the target video clip corresponding to the ball return route is called through the index relationship.

4. A method for tracking and detecting a player and a ball according to any one of claims 1 to 3, characterized in that: Analyze each target video clip, including: analyze each target video clip based on artificial intelligence.

5. A player and ball tracking detection method according to claim 4, characterized in that: The process of obtaining the player tracking detection results corresponding to each target video clip includes: Using the trained person detection model, perform person tracking detection on each target video clip to obtain the person tracking detection result corresponding to each target video clip; According to the field range of the stadium, the person tracking detection result corresponding to each target video clip is filtered to obtain the player tracking detection result corresponding to each target video clip.

6. A player and ball tracking detection method according to claim 4, characterized in that: The process of obtaining the ball tracking detection result corresponding to each target video clip includes: Using the trained ball detection and tracking model, ball detection and tracking are performed on each target video clip to obtain the ball tracking detection result corresponding to each target video clip.

7. A method for tracking and detecting a player and a ball according to any one of claims 1 to 3, characterized in that: The preset ball sports videos are: badminton match videos, badminton training videos, table tennis match videos, table tennis training videos, tennis match videos or tennis training videos.

8. A player and ball tracking detection system, characterized in that: Including video preprocessing module and video analysis module; The video preprocessing module is used to: preprocess the preset ball sports video to obtain multiple target video clips; The video analysis module is used to analyze each target video segment to obtain a player tracking detection result and a ball tracking detection result corresponding to each target video segment.

9. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, a method for tracking and detecting a player and a ball as claimed in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for tracking and detecting a player and a ball according to any one of claims 1 to 7 is implemented.

Images