Volleyball tactical identification method and system based on dynamic crab operator analysis

Abstract

The invention discloses a volleyball tactical recognition method and system based on dynamic crab operator analysis, and the method comprises the steps: carrying out the target detection of volleyballs and players on an input video stream through employing a YOLOv8 model, and obtaining detection frame data, which comprises a position coordinate, a category label and a confidence score; analyzing the spatial distribution characteristics of players for the detection frame data based on a static crab operator, and constructing a vector pointing to each player from O by taking the center point of the court as a static crab operator O, the vector comprising various dynamic information; feature fusion is carried out on the dynamic information, and IOU calculation weight is adjusted in combination with an HMIOU algorithm; and dynamically binding crab operators to tactical key points, and calculating a composition angle between the key points and players in real time for tactical analysis. According to the method, the problems of poor player detection and identity tracking continuity, inaccurate tactical formation judgment, insufficient collaboration quantization and the like in the existing volleyball match video can be solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of computer vision group behavior recognition, and in particular relates to a volleyball tactic recognition method and system based on dynamic crab operator analysis. Background Art

[0002] In team sports like volleyball, basketball, and soccer, player positions and formations directly determine tactical deployment and match success. Existing sports video analysis methods primarily rely on traditional target detection and tracking algorithms (such as YOLOv5, SORT, and ByteTrack) to detect and number players in real time. However, in practice, target identity swaps (ID switches) are frequent due to factors such as occlusion, overlap, perspective changes, and high-speed player motion. In particular, traditional multi-target tracking algorithms (such as SOPT and DeepSOPT) rely on performance features, resulting in a high ID swap rate in occluded scenarios. This leads to poor motion trajectory continuity and low tactical recognition accuracy, limiting the application of efficient video-based tactical analysis.

[0003] Existing volleyball tactical analysis often focuses on event labeling (e.g., serving, blocking, spiking), lacking quantitative modeling of spatial-temporal dynamic tactical characteristics such as player position distribution, angular relationships, and movement trend coordination. This makes it unable to meet the needs of high-level technical and tactical evaluation and tactical drill support. For example, in the technical solution with application number CN202010154331, the overall modeling of spatiotemporal dependencies lacks quantitative analysis of the spatial distribution angles of players.

[0004] Therefore, how to reduce the high ID exchange rate and solve the formation judgment errors and collaborative analysis problems is urgent to be solved. Summary of the Invention

[0005] The main purpose of the present invention is to overcome the shortcomings and deficiencies of the existing technology and provide a volleyball tactic recognition method and system based on dynamic crab operator analysis. By combining YOLOv8 multi-target detection, an improved multi-target tracking algorithm and crab operator feature analysis, it can solve the problems of poor continuity of player detection and identity tracking in existing volleyball game videos, inaccurate tactical formation judgment, and insufficient synergy quantification.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a volleyball tactic recognition method based on dynamic crab operator analysis, comprising the following steps:

[0008] S1. Use the YOLOv8 model to detect the volleyball and players in the input video stream and obtain the detection box data, including location coordinates, category labels, and confidence scores.

[0009] S2. Analyze the spatial distribution characteristics of the players based on the detection frame data using the static crab operator. Take the center point of the court as the static crab operator O and construct a vector pointing from O to each player. The vector includes a variety of dynamic information.

[0010] S3, perform feature fusion on dynamic information and adjust the IOU calculation weight in combination with the HMIOU algorithm;

[0011] S4. Dynamically bind the crab operator to tactical key points, and calculate the angle between the key points and players in real time for tactical analysis.

[0012] As a preferred technical solution, step S2 includes: calculating the cosine similarity and the angle feature, as shown in the following formula:

[0013] Angle equation:

[0014] Cosine distance calculation equation:

[0015] cosine distance=1-cosine similarity

[0016]

[0017] Among them, (x, y) is the position of the static crab operator, (x1, y1) is the player position, cosine distance is the cosine distance, and cosine similarity is the cosine similarity. This feature is used to assist in the extraction of the motion direction and position features of the initial target.

[0018] As a preferred technical solution, step S3 includes: estimating the velocity direction of the four corners of the player detection frame at multi-frame time intervals to obtain the angular velocity direction and HMIOU comprehensive weighted matching score; using the comprehensive similarity score to determine whether the target identity is continuous, thereby optimizing the ID matching accuracy.

[0019] As a preferred technical solution, step S3 includes:

[0020] The positions of the upper left, lower left, upper right, and lower right within the 3-frame time interval are calculated relative to the velocity direction of the previous time, as follows:

[0021]

[0022] Among them, (x 1j ,y 1j ),(x 2j ,y 2j ),(x 3j ,y 3j ),(x 4j ,y4j ) are the coordinate points of the first time point at the upper left, lower left, upper right and lower right of the detection box, direction ij Refers to the velocity direction of the upper left, lower left, upper right, and lower right of the detection frame at each time interval.

[0023] As a preferred technical solution, step S4 includes: dynamically setting the crab operator at different tactical key points, extracting the angle relationship between the crab operator and the players for tactical analysis; the tactical key points include the center of the net, the setter, the main attacker, and the volleyball position;

[0024] Unified angle calculation, let the tactical key point be point O(x,y), and the positions of two players be A(x1,y1) and B(x2,y2), then the angle is defined as

[0025] If the volleyball is not detected and all players are moving at a low speed or are stationary, the crab counter is at the center of the net. If the ball is within 10-20 frames after being hit, the crab counter is on the volleyball. If the ball is about to reach the setter within 15 frames from the first pass to the second pass's jump, and the setter moves toward the ball to prepare to pass, the crab counter is on the setter. If the main attacker runs to the outside line 15-25 frames after the first pass and before the setter passes the ball, the crab counter is on the main attacker.

[0026] As a preferred technical solution, step S4 includes setting the crab operator at the center of the net, specifically:

[0027] If the three players in the front row are relatively evenly spaced and their angle with the center of the net is [60°, 90°], it is considered a three-player blocking formation;

[0028] If the two front row players are located on either side of the center of the net, with angles between them and the center of the net in the range [50°, 80°], and the angle between the third front row player and the center of the net is greater than 100°, then it is considered a two-person blocking formation.

[0029] As a preferred technical solution, step S4 includes setting the crab operator in the setter, specifically:

[0030] With the setter as the center point, calculate the angles between it and the main and secondary attackers. If the setter is in the front row, and the main and secondary attackers are located on either side of the setter, and the angles between them and the setter are between 60° and 120°, then it is considered a 5-1 formation.

[0031] If the setter is in the back row and the angle between him and the main attacker and the secondary attacker is greater than 90 degrees and is within the range (90 degrees, 150 degrees), it is determined to be a 6-2 formation.

[0032] As a preferred technical solution, step S4 includes setting the crab operator in the case of the main attacker, specifically:

[0033] With the main attacker as the center point, observe the relative directions of the setter and the secondary attacker. If the angle between the setter and the secondary attacker relative to the main attacker is greater than 70 degrees, it is considered an open attack formation.

[0034] If the angle between the secondary attacker and the main attacker is [50°, 70°], and the angle between the setter and the main attacker is [30°, 50°], it is judged as a flat pull.

[0035] As a preferred technical solution, step S4 includes setting the crab operator at the volleyball position, specifically:

[0036] If the angle between the front row players and the volleyball is less than 30 degrees, it is considered a fast break screen;

[0037] If the angle between the back row attacker and the volleyball is [20°, 50°], it is determined to be a back attack tactic.

[0038] In a second aspect, the present invention further provides a volleyball tactic recognition system based on dynamic crab operator analysis, which is applied to the volleyball tactic recognition method based on dynamic crab operator analysis, and includes a multimodal detection module, a spatial distribution calculation module, a trajectory management module, and a tactic analysis module;

[0039] The multimodal detection module uses the YOLOv8 model to detect the volleyball and players in the input video stream and obtain detection box data, including location coordinates, category labels, and confidence scores.

[0040] The spatial distribution calculation module is used to analyze the spatial distribution characteristics of players based on the detection box data using the static crab operator. The center point of the court is used as the static crab operator O, and a vector pointing from O to each player is constructed. The vector includes a variety of dynamic information.

[0041] The trajectory management module is used to fuse dynamic information features and adjust the IOU calculation weight in combination with the HMIOU algorithm;

[0042] The tactical analysis module is used to dynamically bind the crab operator to tactical key points, and calculate the angles between key points and players in real time for tactical analysis.

[0043] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0044] (1) The present invention proposes a crab operator key point feature modeling method. Based on the dynamic key points (net center, setter, volleyball position) and the positions of each player, the angle and cosine distance are calculated to extract the relative distribution features, thereby constructing a tactical formation model and realizing automatic recognition of dynamic formations based on relative position relationship and angle distribution.

[0045] (2) The present invention designs a multi-feature fusion ID swap repair mechanism, which combines the detection frame momentum, velocity direction, angle deviation, cosine distance, and confidence status to optimize the matching weight, reduce the ID swap probability, and ensure trajectory continuity.

[0046] (3) This invention proposes for the first time to construct a dynamic crab operator tactical recognition method, which calculates the player angle and spatial distribution characteristics in real time according to the crab operator position, and automatically determines tactical formations such as 5-1, 6-2, three-person block, and open attack. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0048] Figure 1 This is a flow chart of a volleyball tactic recognition method based on dynamic crab operator analysis according to an embodiment of the present invention;

[0049] Figure 2 This is a schematic diagram of a static crab operator according to an embodiment of the present invention;

[0050] Figure 3 This is a schematic diagram of a dynamic crab operator according to an embodiment of the present invention;

[0051] Figure 4 Schematic diagram of the structure of a volleyball tactic recognition system based on dynamic crab operator analysis according to an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0053] References to "embodiments" in this application mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described in this application may be combined with other embodiments.

[0054] See also Figure 1 This embodiment provides a volleyball tactic recognition method based on dynamic crab operator analysis, comprising the following steps:

[0055] S1. Use the YOLOv8 model to detect the volleyball and players in the input video stream and obtain the detection box data, including location coordinates, category labels, and confidence scores.

[0056] The YOLOv8 model is used to detect players in videos. Compared to previous models like YOLOv5 and YOLOv7, YOLOv8 boasts higher detection accuracy and inference speed while maintaining real-time performance. Its anchor-free architecture and SimOTA label assignment strategy improve training efficiency, while multi-scale feature fusion enhances the detection of small objects (such as distant players). The model output includes the center point coordinates, category label, and confidence score for each player.

[0057] S2. Analyze the spatial distribution characteristics of the players based on the detection frame data using the static crab operator. Use the center point of the court as the static crab operator O and construct a vector pointing from O to each player. The vector includes a variety of dynamic information.

[0058] In step S2, the center point of the court is defined as the static crab operator O, a vector pointing from O to each player is constructed, and the cosine similarity and angle feature are calculated as follows:

[0059] Angle equation:

[0060] Cosine distance calculation equation:

[0061] cosine distance=1-cosine similarity

[0062]

[0063] Among them, (x, y) is the position of the static crab operator, (x1, y1) is the player position, cosine distance is the cosine distance, and cosine similarity is the cosine similarity. This feature is used to assist in the extraction of the motion direction and position features of the initial target.

[0064] Calculate and output dynamic information such as angle, cosine distance, momentum of the detection frame, and speed direction.

[0065] S3: Fuse the dynamic information with features and adjust the IOU calculation weights in combination with the HMIOU algorithm.

[0066] The angle and cosine distance obtained based on S2 are fused with dynamic information such as the momentum and speed direction of the detection frame, and the IOU calculation weight is adjusted in combination with the HMIOU algorithm.

[0067] Specifically, the method includes the following steps:

[0068] S31. Extract multidimensional vectors (detection box momentum, velocity direction, angle and cosine distance with the static crab operator), normalize each feature dimension to the same scale [0,1], combine each feature vector into a multidimensional vector, and assign different weights to each feature (for example, 0.2 0.2 0.3 0.3) based on experimental experience or training weights.

[0069] S32, estimate the velocity direction of the four corners of the player detection frame at a 3-frame time interval to obtain the angular velocity direction and HMIOU comprehensive weighted matching score, specifically:

[0070] The positions of the upper left, lower left, upper right, and lower right within the 3-frame time interval are calculated relative to the velocity direction of the previous time, as follows:

[0071]

[0072]

[0073] Among them, (x 1j ,y 1j ),(x 2j ,y 2j ),(x 3j ,y 3j ),(x 4j ,y 4j ) are the coordinate points of the first time point at the upper left, lower left, upper right and lower right of the detection box, direction ij Refers to the speed direction of the upper left, lower left, upper right, and lower right of the detection frame at each time interval, 1×10 -6 This is to prevent division by zero errors when the denominator is 0.

[0074] S33. Calculation of HMIOU comprehensive weighted matching score.

[0075] HMIOU=αH+βM+γIOU

[0076] Among them, H is the histogram appearance similarity, M is the multi-dimensional fusion vector mentioned above, and IOU is the overlap between the detection frames of two frames. The parameters can be automatically optimized through empirical settings or data-driven methods.

[0077] S34. Use the comprehensive similarity score to determine whether the target identity is continuous and optimize the ID matching accuracy.

[0078] It can be explained that step S34 is to determine whether the targets in the consecutive frames are the same target by integrating the similarities, and whether the trajectories of the two targets are the same person. The specific algorithm is as follows:

[0079] Comprehensive similarity calculation: Assume that we have obtained the velocity direction vector of a corner point of the i-th trajectory in the previous few frames as The corresponding corner point velocity direction of the detected target is Then the angular velocity similarity is:

[0080] Final comprehensive similarity: Sim = 0.6·HMIOU+0.4·sim (i,j) .

[0081] If Sim>0.7, it can be determined that the two targets are the same target, otherwise they are new targets.

[0082] S4. Dynamically bind the crab operator to tactical key points, and calculate the angle between the key points and players in real time for tactical analysis.

[0083] The crab operator is dynamically set at different tactical key points (net center, setter, main attacker or volleyball position), and the angular relationship between it and the players is extracted for tactical analysis.

[0084] To unify the angle calculation, let the key point (crab operator) be point O(x,y), and the positions of two players be A(x1,y1) and B(x2,y2), such as Figure 3 , then the angle is defined as:

[0085]

[0086] I need to explain how to dynamically set the Crab Counter at different tactical key points. At the beginning of the game, when the ball isn't detected and all players are moving slowly or stationary, the Crab Counter is at the center of the net. If the ball is moving quickly within 10-20 frames after being hit, the Crab Counter is on the ball. If the ball is approaching the setter within 15 frames from the first pass to the second pass's jump, and the setter is moving toward the ball to pass, the Crab Counter is on the setter. If the main attacker is running toward the perimeter in the 15-25 frames between the first pass and the second pass, the Crab Counter is on the main attacker. The specific steps for tactical analysis are shown in S41-S44.

[0087] S41. The crab counter is in the center of the net.

[0088] (1) If the three players in the front row are evenly spaced and their angles with the center of the net are between 60° and 90°, it means that the blocking mainly relies on these three players and is judged as a three-player blocking formation.

[0089] (2) If the two front row players are located on either side of the center of the net, with angles between them and the center of the net in the range of [50°, 80°], and the angle between the third front row player and the center of the net is greater than 100° (away from the net defense), then the block depends on two players and is judged to be a two-player block formation.

[0090] S42, the crab counter is on the setter.

[0091] Taking the setter as the center point, calculate the angle between it, the main attacker and the secondary attacker.

[0092] (1) If the setter is in the front row, and the main attacker and secondary attacker are located on the left and right sides of the setter, and the angle between them and the setter is in the range of (60°, 120°), it means that the attack starting point is forward, and it is judged to be a 5-1 formation, which is mainly used for fast attack;

[0093] (2) If the setter is in the back row and the angle between him and the main attacker and the secondary attacker is greater than 90° (i.e. the attack point is further back) and is within the range (90°, 150°), then it is considered a 6-2 formation.

[0094] S43. When the crab counter is on the main attacker.

[0095] With the main attacker as the center point, observe the relative directions of the setter and the secondary attacker.

[0096] (1) If the angle between the setter and the secondary attacker relative to the main attacker is greater than 70°, it means that the three players are widely distributed and the space is wide open, and it is judged as a stretched attack formation.

[0097] (2) If the angle between the secondary attacker and the main attacker is [50°, 70°], and the angle between the setter and the main attacker is [30°, 50°], it means that the coordination is more compact and is judged as a flat spread.

[0098] S44, When the Crab Counter is on the Volleyball

[0099] (1) If the angle between the front row players and the volleyball is less than 30°, it means that the players are concentrated close to the volleyball to perform an offensive action, which is judged as a fast break screen.

[0100] (2) If the angle between the back row attacker and the volleyball is [20°, 50°], it means that the tactic was initiated by the back row players and is determined to be a back attack tactic.

[0101] It should be noted that, for the sake of convenience, the aforementioned method embodiments are all expressed as a series of action combinations, but those skilled in the art should know that the present invention is not limited to the described order of actions, because according to the present invention, certain steps can be performed in other orders or simultaneously.

[0102] Based on the same concept as the volleyball tactic recognition method based on dynamic crab operator analysis in the above-mentioned embodiment, the present invention also provides a volleyball tactic recognition system based on dynamic crab operator analysis, which can be used to implement the above-mentioned volleyball tactic recognition method based on dynamic crab operator analysis. For ease of explanation, the structural diagram of the embodiment of the volleyball tactic recognition system based on dynamic crab operator analysis only shows the parts relevant to the embodiment of the present invention. Those skilled in the art will understand that the illustrated structure does not constitute a limitation of the device, and it may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0103] See also Figure 4 In another embodiment of the present application, a volleyball tactic recognition system 10 based on dynamic crab operator analysis is provided, the system comprising a multimodal detection module 11, a spatial distribution calculation module 12, a trajectory management module 13 and a tactic analysis module 14;

[0104] The multimodal detection module 11 is used to perform target detection of the volleyball and the players on the input video stream using the YOLOv8 model to obtain detection frame data, including position coordinates, category labels, and confidence scores;

[0105] The spatial distribution calculation module 12 is used to analyze the spatial distribution characteristics of the players based on the detection frame data using the static crab operator. The center point of the court is used as the static crab operator O, and a vector pointing from O to each player is constructed. The vector includes a variety of dynamic information.

[0106] The trajectory management module 13 is used to perform feature fusion on dynamic information and adjust the IOU calculation weight in combination with the HMIOU algorithm;

[0107] The tactical analysis module 14 is used to dynamically bind the crab operator to the tactical key points, and calculate the angle between the key points and the players in real time to perform tactical analysis.

[0108] It should be noted that the volleyball tactics recognition system based on dynamic crab operator analysis of the present invention corresponds one-to-one to the volleyball tactics recognition method based on dynamic crab operator analysis of the present invention. The technical features and beneficial effects described in the above-mentioned embodiment of the volleyball tactics recognition method based on dynamic crab operator analysis are applicable to the embodiment of the volleyball tactics recognition method based on dynamic crab operator analysis. For specific contents, please refer to the description in the embodiment of the method of the present invention. No further details will be given here. This is hereby declared.

[0109] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0110] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be considered as equivalent replacement methods and are included in the scope of protection of the present invention.

Claims

1. A volleyball tactic recognition method based on dynamic crab operator analysis, characterized in that: The steps include: S1. Use the YOLOv8 model to detect the volleyball and players in the input video stream and obtain the detection box data, including location coordinates, category labels, and confidence scores. S2. Analyze the spatial distribution characteristics of the players based on the detection frame data using the static crab operator. Take the center point of the court as the static crab operator O and construct a vector pointing from O to each player. The vector includes a variety of dynamic information. S3, perform feature fusion on dynamic information and adjust the IOU calculation weight in combination with the HMIOU algorithm; S4. Dynamically bind the crab operator to tactical key points, and calculate the angle between the key points and players in real time for tactical analysis.

2. The volleyball tactic identification method based on dynamic crab operator analysis according to claim 1 is characterized in that: The step S2 includes: calculating the cosine similarity and the angle feature, as follows: Angle equation: Cosine distance calculation equation: cosine distance=1-cosine similarity Among them, (x, y) is the position of the static crab operator, (x1, y1) is the player position, cosine distance is the cosine distance, and cosine similarity is the cosine similarity. This feature is used to assist in the extraction of the motion direction and position features of the initial target.

3. The volleyball tactic identification method based on dynamic crab operator analysis according to claim 1 is characterized in that: The step S3 includes: estimating the speed direction of the four corners of the player detection frame at multi-frame time intervals to obtain the angular speed direction and HMIOU comprehensive weight matching score; using the comprehensive similarity score to determine whether the target identity is continuous, thereby optimizing the ID matching accuracy.

4. The volleyball tactic identification method based on dynamic crab operator analysis according to claim 3 is characterized in that: The step S3 comprises: The positions of the upper left, lower left, upper right, and lower right within the 3-frame time interval are calculated relative to the velocity direction of the previous time, as follows: Among them, (x 1j ,y 1j ),(x 2j ,y 2j ),(x 3j ,y 3j ),(x 4j ,y 4j ) are the coordinate points of the first time point at the upper left, lower left, upper right and lower right of the detection box, direction ij Refers to the velocity direction of the upper left, lower left, upper right, and lower right of the detection frame at each time interval.

5. The volleyball tactic identification method based on dynamic crab operator analysis according to claim 1 is characterized in that: The step S4 includes: dynamically setting the crab operator at different tactical key points and extracting the angle relationship between the crab operator and the players for tactical analysis; the tactical key points include the center of the net, the setter, the main attacker, and the volleyball position; Unified angle calculation, let the tactical key point be point O(x,y), and the positions of two players be A(x1,y1) and B(x2,y2), then the angle is defined as If the volleyball is not detected and all players are moving at a low speed or are stationary, the crab counter is at the center of the net. If the ball is within 10-20 frames after being hit, the crab counter is on the volleyball. If the ball is about to reach the setter within 15 frames from the first pass to the second pass's jump, and the setter moves toward the ball to prepare to pass, the crab counter is on the setter. If the main attacker runs to the outside line 15-25 frames after the first pass and before the setter passes the ball, the crab counter is on the main attacker.

6. The volleyball tactic identification method based on dynamic crab operator analysis according to claim 5 is characterized in that: The step S4 includes setting the crab operator at the center of the net, specifically: If the three players in the front row are relatively evenly spaced and their angle with the center of the net is [60°, 90°], it is considered a three-player blocking formation; If the two front row players are located on either side of the center of the net, with angles between them and the center of the net in the range [50°, 80°], and the angle between the third front row player and the center of the net is greater than 100°, then it is considered a two-person blocking formation.

7. The volleyball tactic identification method based on dynamic crab operator analysis according to claim 5 is characterized in that: The step S4 includes setting the crab operator in the setter, specifically: With the setter as the center point, calculate the angles between it and the main and secondary attackers. If the setter is in the front row, and the main and secondary attackers are located on either side of the setter, and the angles between them and the setter are between 60° and 120°, then it is considered a 5-1 formation. If the setter is in the back row and the angle between him and the main attacker and the secondary attacker is greater than 90 degrees and is within the range (90 degrees, 150 degrees), it is determined to be a 6-2 formation.

8. The volleyball tactics identification method based on dynamic crab operator analysis according to claim 5 is characterized in that: The step S4 includes setting the crab operator in the case of the main attacker, specifically: With the main attacker as the center point, observe the relative directions of the setter and the secondary attacker. If the angle between the setter and the secondary attacker relative to the main attacker is greater than 70 degrees, it is considered an open attack formation. If the angle between the secondary attacker and the main attacker is [50°, 70°], and the angle between the setter and the main attacker is [30°, 50°], it is judged as a flat pull.

9. The volleyball tactics identification method based on dynamic crab operator analysis according to claim 5 is characterized in that: The step S4 includes setting the crab operator at the volleyball position, specifically: If the angle between the front row players and the volleyball is less than 30 degrees, it is considered a fast break screen; If the angle between the back row attacker and the volleyball is [20°, 50°], it is determined to be a back attack tactic.

10. A volleyball tactic recognition system based on dynamic crab operator analysis, characterized in that: A volleyball tactic recognition method based on dynamic crab operator analysis applied to any one of claims 1-9, comprising a multimodal detection module, a spatial distribution calculation module, a trajectory management module, and a tactical analysis module; The multimodal detection module uses the YOLOv8 model to detect the volleyball and players in the input video stream and obtain detection box data, including location coordinates, category labels, and confidence scores. The spatial distribution calculation module is used to analyze the spatial distribution characteristics of players based on the detection box data using the static crab operator. The center point of the court is used as the static crab operator O, and a vector pointing from O to each player is constructed. The vector includes a variety of dynamic information. The trajectory management module is used to fuse dynamic information features and adjust the IOU calculation weight in combination with the HMIOU algorithm; The tactical analysis module is used to dynamically bind the crab operator to tactical key points, and calculate the angles between key points and players in real time for tactical analysis.

Images