System and method for finding an open space to provide an offensive pass suggestion

Abstract

A system and method for discovering a space to provide an attacking pass suggestion, by training a space analysis model with historical game videos, and using the trained space analysis model to determine a space position in a target game video, determining a pass direction of a ball holder in the target game video, and generating a pass suggestion when the pass direction does not match the space position, can provide an opportunity for players to fully understand and apply tactics, and achieve the technical effect of providing a teaching mode that experiences the game process and provides feedback.

Description

Technical Field

[0001] A basketball passing suggestion system and method, specifically a system and method for extracting image features from basketball game footage to identify openings and provide offensive passing suggestions. Background Technology

[0002] In today's era of rapid digital and technological development, artificial intelligence (AI), big data, and virtual reality (VR) technologies are demonstrating their unique value in many fields, especially in education.

[0003] Currently, the teaching methods used in the education field include basic classroom teaching, supplementary on-site teaching and individual guidance, special multi-grade teaching, and computerized multi-media teaching. However, most of the above teaching methods are only suitable for static teaching and not for dynamic teaching, such as physical education courses like basketball.

[0004] Traditional basketball instruction includes theoretical learning and practical training. Basketball theory is typically taught to players primarily through experience-based instruction. However, basketball is a sport that demands a high level of both technical practice and tactical understanding. Traditional teaching methods face limitations in terms of facilities, resources, and personalized instruction, especially in tactical understanding and practical application. Aside from actual games, it's difficult to provide players with sufficient opportunities for practice. Even during games, coaches can only superficially observe the players' training results. Therefore, there is a lack of a teaching model that allows players to experience tactical processes and receive immediate feedback.

[0005] In conclusion, it is clear that existing technologies have long suffered from the problem of failing to provide players with sufficient opportunities to understand and apply tactics outside of actual matches. Therefore, it is necessary to propose improved technical means to solve this problem. Summary of the Invention

[0006] In view of the problem that existing technologies fail to provide players with sufficient opportunities to fully understand and apply tactics in actual matches, this invention discloses a system and method for identifying gaps to provide offensive passing suggestions, wherein:

[0007] The system disclosed in this invention for identifying gaps and providing offensive passing suggestions includes at least: a memory for storing computer instructions; and a processor for executing the computer instructions to generate the following modules: a data acquisition module for reading gap rules and loading multiple historical game images, each historical game image containing a basketball and multiple players, including offensive and defensive players, and for loading a target game image containing at least one of the multiple players; and an image analysis module for analyzing the multiple historical game images to extract multiple historical image features at different time points from each historical game image, and for analyzing the target game image to extract multiple target image features at different time points from the target game image, and based on... The multiple target image features identify and track the basketball and the on-court positions of each player in the target game image; the image processing module is used to determine, based on the on-court positions of the basketball and the multiple players, when the player with the ball passes the basketball to another player among the multiple players, to acquire the target game scene of the basketball passing in the target game image and determine the direction of the basketball pass; the model training module is used to train a gap analysis model using the multiple historical image features that represent the distance and relative position between the multiple players satisfying the gap rule; the gap identification module is used to analyze the passing gap position in the target game scene using the gap analysis model; the suggestion generation module is used to generate a passing suggestion when it is determined that the passing direction does not match the passing gap position.

[0008] The method for identifying gaps and providing offensive passing suggestions disclosed in this invention includes at least the following steps: loading multiple historical game images, each containing a basketball and multiple players, including offensive and defensive players; reading gap rules; analyzing the multiple historical game images to extract multiple historical image features at different time points from each historical game image; training a gap analysis model using a first machine learning algorithm based on the multiple historical image features indicating that the distance and relative position between the multiple players satisfy the gap rules; loading a target game image, which contains at least one of the multiple players; analyzing the target game image to extract multiple target image features at different time points from the target game image, and identifying and tracking the on-court positions of the basketball and each player in the target game image; determining, based on the on-court positions of the basketball and the multiple players, when the ball-handling player among the multiple players passes the basketball to another player among the multiple players, acquiring the target game scene of the basketball passing in the target game image, and determining the passing direction of the basketball; using the gap analysis model to analyze the passing gap position in the target game scene; and generating a passing suggestion when the passing direction does not match the passing gap position.

[0009] The system and method disclosed in this invention are as described above. The difference between this invention and the prior art is that this invention trains a gap analysis model using historical game footage, and after using the trained gap analysis model to determine the gap position in the target game footage, it generates a passing suggestion when the passing direction of the ball-handling player in the target game footage does not match the gap position. This solves the problems existing in the prior art and can achieve the technical effect of providing a teaching mode that can provide an experience of the game process and provide feedback. Attached Figure Description

[0010] Figure 1 A schematic diagram of the components of the device for detecting gaps and providing offensive passing suggestions according to the present invention.

[0011] Figure 2 This is a schematic diagram of the module of the system proposed in this invention for detecting gaps and providing offensive passing suggestions.

[0012] Figure 3A The flowchart illustrates the method for identifying gaps and providing offensive passing suggestions as proposed in this invention.

[0013] Figure 3B This is a flowchart of the method for training the gap analysis model proposed in this invention.

[0014] Figure 3C This is a flowchart of the method for displaying target match footage, passing suggestions, and marking open positions, as proposed in this invention.

[0015] Figure 3D This is a flowchart of the method for generating passing suggestions based on offensive tactics proposed in this invention.

[0016] Figure 3E This is a flowchart of the method for training an action prediction model to predict future gap locations, as proposed in this invention.

[0017] Figure 4 This is a schematic diagram illustrating the display of target match images and a tactical board, with gaps marked, as provided in an embodiment of the present invention.

[0018] Explanation of reference numerals in the attached figures:

[0019] 100: Device

[0020] 110: Memory

[0021] 120: Input Module

[0022] 125: Peripheral input device

[0023] 130: Communication Interface

[0024] 140: Storage medium

[0025] 150: Output Module

[0026] 155: Peripheral output device

[0027] 170: Processor

[0028] 190: Bus

[0029] 210: Data Acquisition Module

[0030] 220: Image Analysis Module

[0031] 230: Image Processing Module

[0032] 250: Model Training Module

[0033] 260: Neutral Gear Detection Module

[0034] 270: Tactical Selection Module

[0035] 280: Recommendation to generate modules

[0036] 290: Information display module

[0037] 410: Match footage

[0038] 411~413: Neutral position

[0039] 420: Tactical Board

[0040] 421~423: Neutral position

[0041] 450: Passing Advice

[0042] Step 301: Load historical game footage, which includes basketballs and players. Step 305: Load gap rules.

[0043] Step 310: Analyze historical match footage to extract historical image features

[0044] Step 315: Identify and track the movement trajectory of the basketball and the on-court positions of each player based on historical image features.

[0045] Step 320: Train the action prediction model based on the basketball's trajectory and the players' positions on the court. Step 330: Train the gap analysis model based on historical image features that satisfy the gap rule.

[0046] Step 335: Based on historical video features, determine the success or failure of a pass in an open position that meets the open position rule.

[0047] Step 337: Identify defensive players at the edge of the rule-defined gaps based on historical video features.

[0048] Step 339: Train the gap analysis model based on historical image features, the success or failure of passes in open positions according to rules, and the body parameters of defensive players on the edge of open positions.

[0049] Step 340: Load the target match video

[0050] Step 350: Analyze the target game video to extract target video features, and identify and track the on-court positions of the basketball and each player in the target game video.

[0051] Step 360: Based on the basketball and players' positions on the court, when the player with the ball passes the basketball to another player, obtain the target game screen of the basketball passing and determine the direction of the basketball pass.

[0052] Step 370: Use the gap analysis model to analyze the gap positions in the target game frame.

[0053] Step 375: Based on the basketball and the players' positions on the court, use a motion prediction model to predict the predicted positions of each player at a specific future time.

[0054] Step 377: Based on the predicted positions of each player, use the gap analysis model to analyze the predicted gap positions at a specific time.

[0055] Step 380: When the direction of the pass does not match the open space, generate a passing suggestion.

[0056] Step 381: Determine if the direction of the pass matches the open space.

[0057] Step 383: Count the number of times the gap distribution of defensive players at each position on the field is represented by historical image features when the gap rule is met.

[0058] Step 387: Select offensive tactics based on the defensive players' on-court positions and number of open looks indicated by the target image features.

[0059] Step 389: Generate passing suggestions based on offensive tactics Detailed Implementation

[0060] The features and implementation methods of the present invention will be described in detail below with reference to the accompanying drawings and embodiments. The content is sufficient to enable any person skilled in the art to easily and fully understand the technical means used by the present invention to solve the technical problem and to implement it accordingly, thereby achieving the effects that the present invention can achieve.

[0061] This invention can analyze open spaces in basketball game footage and generate passing suggestions for the player with the ball based on those open spaces.

[0062] The apparatus for implementing this invention can be a computing device, which includes, but is not limited to, one or more processors, one or more memories, and buses connecting different hardware components (including memories and processors). Through the included hardware components, the computing device can load and execute an operating system, allowing the operating system to run on the computing device, and can also execute software or programs. The computing device also includes a housing, within which the aforementioned hardware components are disposed.

[0063] The bus of the computing device proposed in this invention may include one or more types, such as a data bus, address bus, control bus, expansion bus, and / or local bus. The bus of the computing device includes, but is not limited to, Industry Standard Architecture (ISA) bus, Peripheral Component Interconnect (PCI) bus, Video Electronics Standards Association (VESA) local bus, Universal Serial Bus (USB), and PCI Express (PCI-E / PCIe) bus.

[0064] The processor of the computing device proposed in this invention is coupled to a bus. The processor includes a set of registers or a register space, which may be entirely located on the processor's processing chip, or wholly or partially located outside the processing chip and coupled to the processing chip via dedicated electrical connections and / or via a bus. The processor may be a central processing unit, a microprocessor, or any suitable processing element. If the computing device is a multiprocessor device, that is, the computing device contains multiple processors, then the processors contained in the computing device are identical or similar and are coupled and communicate via a bus. In some embodiments, the processor can interpret a computer instruction or a series of multiple computer instructions to perform specific operations or calculations, such as mathematical operations, logical operations, data comparison, copying / moving data, etc., thereby driving other hardware components in the computing device or running an operating system or executing various programs and / or modules. Computer instructions can be assembly language instructions, instruction set architecture instructions, machine instructions, machine-dependent instructions, microinstructions, firmware instructions, or source code or object code written in any combination of one or more programming languages. Computer instructions can be executed entirely on a single computing device, partially on a single computing device, or partially on one computing device and partially on another connected computing device. The aforementioned programming languages ​​include object-oriented programming languages ​​such as Common Lisp, Python, C++, Objective-C, Smalltalk, Delphi, Java, Swift, C#, Perl, Ruby, etc., as well as conventional procedural programming languages ​​such as C or other similar programming languages.

[0065] Computing devices typically include one or more chipsets. The processor of the computing device can be coupled to the chipset or electrically connected to the chipset via a bus. A chipset consists of one or more integrated circuits (ICs), including a memory controller and peripheral input / output (I / O) controllers, etc. That is, the memory controller and I / O controllers can be contained within a single IC or implemented using two or more ICs. Chipsets typically provide I / O and memory management functions, as well as multiple general-purpose and / or special-purpose registers, timers, etc., which can be accessed or used by one or more processors coupled to or electrically connected to the chipset. In some embodiments, the chipset may also be part of the processor.

[0066] The processor of a computing device can also access data in the memory and mass storage areas installed on the computing device through the memory controller. The aforementioned memory includes any type of volatile memory and / or non-volatile memory (NVRAM), such as Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Read-Only Memory (ROM), Flash memory, etc. The aforementioned mass storage areas can contain any type of storage device or storage medium, such as hard disks, optical discs, flash drives, memory cards, solid-state drives (SSDs), or any other storage device. In other words, the memory controller can access data in static random access memory, dynamic random access memory, flash memory, hard disk drives, and solid-state drives.

[0067] The processor of a computing device can also connect and communicate with peripheral devices or interfaces such as peripheral output devices, peripheral input devices, communication interfaces, and various data or signal receiving devices via a peripheral input / output controller and a peripheral input / output bus. Peripheral input devices can be any type of input device, such as a keyboard, mouse, trackball, touchpad, joystick, etc., while peripheral output devices can be any type of output device, such as a monitor, printer, etc. Peripheral input devices and peripheral output devices can also be the same device, such as a touchscreen. Communication interfaces can include wireless communication interfaces and / or wired communication interfaces. Wireless communication interfaces can include interfaces supporting wireless local area networks (such as Wi-Fi, Zigbee, etc.), Bluetooth, infrared, near-field communication (NFC), 3G / 4G / 5G mobile communication networks (cellular networks), or other wireless data transmission protocols. Wired communication interfaces can be Ethernet devices, DSL modems, cable modems, asynchronous transfer mode (ATM) devices, or fiber optic communication interfaces and / or components, etc. The data or signal receiving device may include a GPS receiver or a physiological signal receiver, the physiological signals received by the physiological signal receiver including, but not limited to, heartbeat, blood oxygenation, etc. The processor can periodically poll various peripheral devices and interfaces, enabling the computing device to input and output data through various peripheral devices and interfaces, and also to communicate with another computing device having the hardware components described above.

[0068] The following will first be based on " Figure 1 "The schematic diagram of the device for detecting gaps and providing offensive passing suggestions, as proposed in this invention, illustrates the device for implementing this invention. For example..." Figure 1 As shown, the device 100 of the present invention includes a memory 110, an input module 120, a communication interface 130, a storage medium 140, an output module 150, a processor 170, and a bus 190. The memory 110, the input module 120, the communication interface 130, the storage medium 140, the output module 150, and the processor 170 are interconnected via the bus 190.

[0069] The memory 110 can store one or more sets of computer instructions.

[0070] The input module 120 can provide input data through the peripheral output device 125 of the device 100. For example, the input module 120 can provide input gap rules, select or input the storage path of images such as match images (including historical match images and target match images) and / or player headshots through the peripheral output device 125 such as a keyboard and / or mouse, and input player personal data such as player limb parameters and / or jersey numbers. The limb parameters mentioned in this invention include, but are not limited to, the player's height, arm length, leg length, and stride distance.

[0071] The communication interface 130 can be connected to an external network storage device or server (not shown in the figure) and request and download data from the connected network device. For example, the communication interface 130 can connect to the network device to download data such as gap rules and match videos. The match videos can be transmitted to the device 100 by the network device in the form of files or streams. There are no particular limitations in this invention.

[0072] Storage medium 140 can store data received by communication interface 130, and can also store data required by processor 170 (i.e. provided to processor 170), such as gap rules, game images, etc.; storage medium 140 can also store data generated by processor 170, such as gap information, passing suggestions, etc.

[0073] The output module 150 can output the data generated by the processor 170 through the peripheral output device 155 of the device 100. For example, the output module 150 can output game images, open positions, passing suggestions, etc. through the peripheral output device 155 such as a display or touch screen, so that the device 100 can play game images, mark open positions, and display passing suggestions.

[0074] Processor 170 can be like " Figure 2 The schematic diagram of the processor module for detecting gaps and providing offensive passing suggestions proposed in this invention is shown. It includes modules such as a data acquisition module 210, an image analysis module 220, an image processing module 230, a model training module 250, a gap identification module 260, and a suggestion generation module 280. It may also include an attachable tactical selection module 270 and an information display module 290. In some embodiments, the processor 170 can execute computer instructions stored in the memory 110 and can generate a suggestion after executing the computer instructions. Figure 2 The modules in ""; in another part of the embodiment, " Figure 2 The modules in the "" can be generated from one or more circuits and / or complete or partial chips and other hardware components, that is, the processor 170 contains components that make up the " Figure 2 The hardware components of each module in the "processor 170" are, in other words, the modules included in the processor 170 can be software modules or hardware modules, and there are no particular limitations in this invention.

[0075] The data acquisition module 210 is responsible for acquiring gap rules. For example, the data acquisition module 210 can acquire gap rules input by the input module 120, gap rules received by the communication interface 130, or gap rules pre-stored in the storage medium 140. The gap rules acquired by the data acquisition module 210 may include conditions such as the distance between the target offensive player and other defensive players being greater than or equal to a predetermined value, and the presence of other offensive players between the target offensive player and a defensive player whose distance is less than a predetermined value. However, the present invention is not limited to these conditions.

[0076] The data acquisition module 210 is also responsible for loading game footage (including historical game footage and target game footage). For example, the data acquisition module 210 can read game footage from storage medium 140 according to the storage path provided by input module 120, or it can download game footage by connecting to a network device through communication interface 130 according to the storage path provided by input module 120; the data acquisition module 210 can also directly read game footage from a specific directory of storage medium 140, or it can download game footage from a specific directory of network device through communication interface 130. Each game footage loaded by the data acquisition module 210 contains a basketball and players, and the players in the game footage include one or more offensive players and one or more defensive players. Generally speaking, different game footage loaded by the data acquisition module 210 may contain the same one or more players, or may contain the same team.

[0077] The image analysis module 220 is responsible for analyzing the match images loaded by the data acquisition module 210, thereby extracting multiple image features from each different match image. In this invention, the image features extracted by the image analysis module 220 from historical match images are called historical image features, and the target image features extracted from the target match image are called target image features.

[0078] The image analysis module 220 can extract features from each frame or a specific frame in the game image, thereby extracting image features from frames at different time points in each game image. The specific frames mentioned above can be frames in the game image at fixed intervals or at fixed time intervals (such as frames every five frames or frames every 0.2 seconds), or frames with a similarity to the previous frame that is less than a certain value, or frames where events such as shooting, passing, turnover, blocking, or interception occur. However, the image analysis module 220 is not limited to the above methods of selecting frames for feature extraction. For example, the image analysis module 220 can use pre-trained object detection models such as YOLO and Faster R-CNN to detect objects such as players and balls in the frame, thereby obtaining the bounding boxes of each object in the frame. It can also use pre-trained residual neural network (ResNet) and visual geometry group (VGG) models such as convolutional neural network (CNN) to extract spatial features from the frames in the game image, thereby extracting frame spatial feature data and object spatial feature data of each object in the frame. However, this invention is not limited to these methods. In some embodiments, the image analysis module 220 can also use multi-object tracking methods such as SORT (Simple Online and Realtime Tracking), Deep SORT (Simple Online and Realtime Tracking with a Deep AssociationMetric), and Kalman Filter. The Tracking over Time (MOT) algorithm tracks detected objects in subsequent frames to identify the same objects in later frames. It can also generate temporal feature data for each object in the game footage by arranging the spatial feature data of each object in each frame according to the chronological order of the frames. This temporal feature data represents the object's position, while the temporal feature data represents its trajectory and velocity.

[0079] The image processing module 230 can determine the behavior of each player based on the basketball and player's on-court position and movement trajectory or image features generated by the image analysis module 220, and can obtain the time when each player begins to perform the judged behavior. Generally speaking, the behavior of offensive players includes shooting, passing, breaking through, blocking, etc., and the behavior of defensive players includes guarding, intercepting, blocking, etc., but the present invention is not limited thereto.

[0080] The image processing module 230 can use a trained behavior recognition model to identify the behavior of the attacking player represented by the image features generated by the image analysis module 220. For example, the image processing module 230 can use a pose estimation model such as OpenPose to determine the position (e.g., coordinates) of key points (including but not limited to head, shoulders, elbows, wrists, hips, knees, ankles, etc.) of various parts of the player's body, and can use the position changes of each key point over time as input to identify the player's behavior using the behavior recognition model. However, the representation of key points and their positions proposed in this invention is not limited to the above.

[0081] The image processing module 230 can train a behavior recognition model based on a large number of player images with known actions. For example, the image processing module 230 provides a large number of player images and the actions of the players in each player image to the behavior recognition model, so that after the behavior recognition model determines the position of the key points of each part of the player's body in each player image, it can train based on the position of the key points of each part of the player's body in each player image and the known actions.

[0082] The image processing module 230 can also determine the offensive player's behavior based on the basketball, the offensive player's position and movement trajectory on the court, and the position of the basket. For example, when the basketball's movement trajectory moves from the offensive player with the ball to another offensive player, the image processing module 230 can determine that the offensive player with the ball's behavior is passing; when the basketball's movement trajectory moves from the offensive player with the ball to the basket, and there are no other offensive players within a certain distance around the basket, the image processing module 230 can determine that the offensive player with the ball's behavior is shooting; when the basketball's movement trajectory is the same as the offensive player with the ball, and there is a defensive player within a certain distance around the offensive player with the ball, the image processing module 230 can determine that the offensive player with the ball's behavior is driving to the basket; when the basketball's movement trajectory is the same as the offensive player with the ball, and there are other offensive players without the ball within a certain distance around the offensive player with the ball, and there is a defensive player on the side of the offensive player with the ball that is different from the offensive player with the ball, the image processing module 230 can determine that the offensive player without the ball's behavior is blocking. However, the way the image processing module 230 judges the behavior of the attacking player is not limited to the above.

[0083] The image processing module 230 is responsible for determining the distance and relative position between players based on the historical image features (object spatial feature data of each player) generated by the image analysis module 220, and can further determine whether the historical image features satisfy the gap rules obtained by the data acquisition module 210 based on the distance and relative position between players determined by the historical image features.

[0084] The image processing module 230 is also responsible for extracting frames whose historical image features satisfy the gap rules from the historical match images obtained by the data acquisition module 210 when the historical image features generated by the image analysis module 220 satisfy the gap rules obtained by the data acquisition module 210. In this invention, the frames extracted by the image processing module 230 from the historical match images are also called historical match frames, and the positions in the historical match frames that satisfy the gap rules are also called rule gap positions.

[0085] The image processing module 230 is also responsible for determining the passing direction of the basketball when a player passes the basketball to another player, based on the on-court positions of the basketball and each player represented by the target image features generated by the image analysis module 220. For example, the image processing module 230 can determine the relative positions between the basketball and each player based on the spatial feature data of the objects contained in the target image features, and then determine the player with the ball. The image processing module 230 can also obtain the trajectory of the basketball based on the temporal feature data of the objects contained in the target image features, and then determine the passing direction of the basketball.

[0086] The image processing module 230 is also responsible for acquiring the frame in the target game image where a player passes the basketball to another player when it is determined that a player has passed the basketball to another player. In this invention, the frame acquired by the image processing module 230 from the target game image is also referred to as the target game frame.

[0087] The image processing module 230 can also obtain the on-court positions of one or more defensive players in various historical game images based on the historical image features generated by the image analysis module 220, and can determine whether the on-court positions of these defensive players satisfy the open space rule. The image processing module 230 can also count the number of times the on-court positions of defensive players satisfy the open space rule to generate the open space distribution count. For example, the image processing module 230 can divide the basketball court into M×N blocks, and can count the number of times a specific defensive player satisfies the open space rule in each block to generate the open space distribution count, or it can count the number of times the combination of two or more defensive players in each block satisfies the open space rule to generate the open space distribution count. This invention has no particular limitations.

[0088] The image processing module 230 can determine whether a pass is successful when the player passes the basketball to a designated open space based on the historical image features generated by the image analysis module 220, and generate a pass success or failure status based on the determination result. The image processing module 230 can also generate a pass success or failure status indicating a successful pass when the basketball moves to an offensive player in an open space, based on object spatial feature data in the historical image features; conversely, it can generate a pass failure status indicating a failed pass when the basketball is not successfully passed to an offensive player in an open space, for example, when the basketball moves off the court or onto a defensive player.

[0089] The image processing module 230 can also identify defensive players at the edge of the rule gaps. For example, the image processing module 230 can compare facial features in historical game footage with pre-established facial features of each player to identify defensive players, or it can identify defensive players based on the bounding boxes determined when tracking objects in historical game footage, but the invention is not limited thereto.

[0090] The image processing module 230 can obtain the movement trajectory and speed of the basketball and players in the images of each historical game based on the object time feature data in the historical image features, and calculate the movement trajectory of the basketball and the on-court position of each player at different time points in the images of each historical game based on the obtained movement trajectory and speed of the basketball and players.

[0091] The model training module 250 is responsible for training a gap analysis model using a first machine learning algorithm based on the historical image features generated by the image analysis module 220 and the regular gap positions generated by the image processing module 230. The first machine learning algorithm can be a reinforcement learning (RL) model, but this invention is not limited thereto.

[0092] In some embodiments, in order to improve the accuracy of the gap analysis model in determining the gap location, the model training module 250, in addition to using historical image features and regular gap locations, can also use the pass success or failure status generated by the image processing module 230 and the limb parameters of the defending player at the edge of the regular gap location to train the gap analysis model.

[0093] The model training module 250 can also train the action prediction model based on the object time feature data of the basketball and each player in the historical image features generated by the image analysis module 220 trained by the model training module 250. That is, the action prediction model is trained based on the movement trajectory of the basketball and each player at different times in each historical game image and the on-court position of each player at different times in the same historical game image. However, the data used to train the action prediction model in this invention is not limited to the above.

[0094] The gap recognition module 260 is responsible for using the gap analysis model trained by the model training module 250 to analyze the target image features (i.e., the target image features generated by the image analysis module 220 analyzing the target match image) in the target match frame obtained by the image processing module 230, thereby obtaining the gap positions in the target match frame. It should be noted that the gap recognition module 260 can analyze one or more gap positions from a single target match frame.

[0095] The gap recognition module 260 can also use the action prediction model trained by the model training module 250 to predict the position of each player at a specific time in the future (in this invention, the position of the player at a specific time in the future is also referred to as the predicted position) based on the on-court position of the basketball and each player in the target game image, and use the gap analysis model to analyze the gap prediction position at a specific time based on the predicted position of each player.

[0096] The tactical selection module 270 can select an offensive tactic based on the target image features generated by the image analysis module 220, which represent the on-court positions of one or more defensive players, and the number of open looks generated by the image processing module 230. For example, the tactical selection module 270 can select an offensive tactic that attacks the passing open space position corresponding to the highest number of open looks corresponding to the on-court positions of the defensive players. In some embodiments, when multiple offensive tactics exist, the tactical selection module 270 can select the offensive tactic with the highest scoring rate.

[0097] The suggestion generation module 280 is responsible for generating a passing suggestion when the passing direction determined by the image processing module 230 does not match the open position generated by the open position recognition module 260. The passing suggestion generated by the suggestion generation module 280 may include the passing time and passing route. For example, the passing suggestion may be to pass the basketball to an open position at the same time as the ball handler passes the ball, or it may be to pass the basketball to an open position before or after the time of the ball handler's pass. For example, the suggestion generation module 280 can select the largest open position to generate a passing suggestion, but the present invention is not limited to the above.

[0098] The suggestion generation module 280 can also generate passing suggestions based on the offensive tactics generated by the tactical selection module 270. For example, the suggestion generation module 280 can suggest passing the basketball to the highest open space, ensuring the basketball is passed in a position that aligns with the offensive tactics.

[0099] The information display module 290 can provide the image signal of the target match image loaded by the data acquisition module 210 to the output module 150, so that the output module 150 can play the target match image on the display of the device 100.

[0100] The information display module 290 can also add passing suggestions generated by the suggestion generation module 280 to the image signal of the target match image provided to the output module 150 when the output module 150 plays the target match image, and add gap recognition module 260 to the image signal of the target match image to mark the gap position generated by the gap analysis model of the gap recognition module 260.

[0101] In some embodiments, the information display module 290 may also add a display area of ​​the tactical board to the image signal of the target game image played by the output module 150, and mark the player's on-field position determined by the image processing module 230, the gap position determined by the gap recognition module 260, and the passing suggestion generated by the suggestion generation module 280 (such as the full content of the passing suggestion or part of the passing route in the passing suggestion).

[0102] The information display module 290 can also provide the option to select one of the viewpoints to play the match video corresponding to the selected viewpoint when the target match video contains multiple synchronized match videos from different perspectives, through the input module 120.

[0103] The operating system and method of the present invention will then be explained using an embodiment, and please refer to " Figure 3A "A flowchart of the method for identifying gaps to provide offensive passing suggestions proposed in this invention. In this embodiment, it is assumed that device 100 is a computer or server containing multiple top-level processing cores."

[0104] When basketball coaches, assistants, or players want to improve the passing skills of players (especially point guards) by reviewing game content, they can use this invention to first train the gap analysis model.

[0105] First, team members can connect to device 100 using a personal computer and input the storage path and file name of historical match images through the user interface provided by the data acquisition module 210 of device 100. The input storage path and file name of the historical match images are then transmitted back to device 100, allowing device 100 to load the historical match images (step 301). Assuming that in this embodiment, the historical match images are stored on an external file server, after executing the computer instructions stored in the memory 110 of device 100, the processor 170 of device 100 can generate commands such as " Figure 2 "The executable modules are such that the data acquisition module 210 of the device 100 can connect to an external file server through the communication interface 130 of the device 100, and download the historical match images from the file server according to the storage path and file name of the historical match images received by the communication interface 130."

[0106] While the data acquisition module 210 of device 100 loads historical match footage (step 301), it can also read gap rules (step 305). In this embodiment, it is assumed that the data acquisition module 210 can read pre-established gap rules from the storage medium 140 of device 100. It should be noted that in this invention, there is no sequential relationship between the data acquisition module 210 loading historical match footage (step 301) and reading gap rules (step 305). That is, the data acquisition module 210 can also read gap rules first (step 305) and then load historical match footage (step 301).

[0107] After the data acquisition module 210 of device 100 loads historical match images (step 301), the image analysis module 220 of device 100 can analyze each loaded historical match image to extract historical image features at different time points from each historical match image (step 310).

[0108] Subsequently, the image processing module 230 of device 100 can determine whether the distance and relative position between each player represented by each historical image feature generated by the image analysis module 220 satisfies the gap rule. The model training module 250 of device 100 can train a gap analysis model using a first machine learning algorithm based on the historical image features that satisfy the gap rule as determined by the image processing module 230 (step 330). In this embodiment, it is assumed that the image processing module 230 can obtain the distance and relative position between each player represented by each historical image feature, and determine whether the distance between each attacking player and all defending players is greater than or equal to a predetermined value. If so, it indicates that there is a rule gap position corresponding to the attacking player. If not, it continues to determine whether there are other attacking players between the attacking player and all defending players whose distance is less than the predetermined value. If so, it indicates that there is a rule gap position corresponding to the attacking player. If not, it indicates that there is no rule gap position corresponding to the attacking player. When any historical image feature indicates that the distance and relative position between players meets the gap rule, the image processing module 230 can record the historical image feature that meets the gap rule and continue to judge whether other historical image features meet the gap rule. After all historical image features have been judged, the model training module 250 can train the gap analysis model based on the historical image features recorded by the image processing module 230, thereby completing the training of the gap analysis model.

[0109] In addition, to increase the accuracy of the neutral position analysis module, it is also possible to use methods such as " Figure 3B As shown in the process, the image processing module 230 of device 100 can determine the success or failure status of a pass in a regular open space that meets the open space rule based on the historical image features generated by the image analysis module 220 of device 100 (step 335), and can identify the defending player at the edge of the determined regular open space based on the historical image features (step 337). The model training module 250 of device 100 can train the open space analysis model based on the historical image features generated by the image analysis module 220, the success or failure status of a pass in a regular open space determined by the image processing module 230, and the limb parameters of the defending player at the edge of the regular open space obtained by the data acquisition module 110 of device 100 (step 339).

[0110] Continue back to " Figure 3AFollowing the process of "After the model training module 250 of device 100 completes the training of the gap analysis model, device 100 can provide team members with viewing the target match video to review the match content. Thus, the data acquisition module 210 of device 100 can first load the target match video (step 340). In this embodiment, assuming a similar process to the data acquisition module 210 of device 100 loading historical match video (step 301), team members can use a personal computer to connect to device 100 and input the storage path and file name of the target match video on the user interface provided by the data acquisition module 210 of device 100 through the personal computer. The input storage path and file name of the target match video are then transmitted back to device 100, allowing the data acquisition module 210 of device 100 to connect to an external file server through the communication interface 130 of device 100, and download the target match video from the file server according to the storage path and file name of the target match video received through the communication interface 130.

[0111] After the data acquisition module 210 of device 100 loads the target game image, the image analysis module 220 of device 100 can analyze the loaded target game image to extract target image features at different time points from the target game image. Simultaneously, during the extraction of target image features, it identifies and tracks the basketball and the on-court positions of each player contained in the target game image (step 350). In this embodiment, it is assumed that the image analysis module 220 can use a multi-target tracking algorithm to identify and track the basketball and the on-court positions of each player contained in the target game image at different time points (different frames), and can extract target image features at different time points in the target game image based on the changes in the on-court positions of the basketball and the players in the target game image.

[0112] After the image analysis module 220 of device 100 extracts the target image features of the target game image and identifies and tracks the basketball and the on-court positions of each player contained in the target game image (step 350), the image processing module 230 of device 100 can determine when the player with the ball passes the basketball to another player based on the on-court positions of the basketball and each player obtained by the image analysis module 220, obtain the target game scene of the player with the ball passing the basketball in the target game image, and determine the passing direction of the player with the ball passing the basketball (step 360).

[0113] Similarly, after the image analysis module 220 of device 100 extracts the target image features of the target game image and identifies and tracks the basketball and the on-court positions of each player contained in the target game image (step 350), the gap recognition module 260 of device 100 can use the gap analysis model trained by the model training module 250 of device 100 to analyze the gap positions in the target game image (step 370). In this embodiment, it is assumed that the image analysis module 220 of device 100 can acquire 5 frames per second of the target game image for image analysis, which means that the image analysis module 220 can generate a set of target image features every 0.2 seconds in the target game image. The gap recognition module 260 can use the gap analysis model to analyze each set of target image features to identify a gap position every 0.2 seconds in the target game image, including the passing gap position when the player with the ball passes the basketball (411-413), such as " Figure 4 As shown in the figure.

[0114] After the gap recognition module 260 of device 100 analyzes the gap position in the target game image using the gap analysis model (step 370), the suggestion generation module 280 of device 100 can determine that the passing direction generated by the image processing module 230 of device 100 does not match the gap position analyzed by the gap recognition module 260 and generate a passing suggestion (step 380). In this embodiment, assuming that the passing direction of the ball-handling player does not match the gap position, the suggestion generation module 280 can choose to generate a passing suggestion at the gap position at the moment the ball-handling player makes the pass, or it can choose to generate a passing suggestion 420 at the optimal gap position before the ball-handling player makes the pass, such as " Figure 4 As shown in the figure.

[0115] Thus, through this invention, it is possible to determine whether the player with the ball in a basketball chooses the best time to pass the ball to the best position, so that the player with the ball can receive feedback after the actual game.

[0116] In the above embodiments, if the processor 170 of the device 100 can generate the information display module 290, then it can be as follows: Figure 3CAs shown in the process, after the suggestion generation module 280 of device 100 generates a passing suggestion (step 380), when the information display module 290 generates the image signal of the target match image loaded by the data acquisition module 210 of device 100 and provides it to the output module 150 of device 100 for playback on the display of device 100 (or when the information display module 290 provides the image signal to the communication interface 130 of device 100 so that the communication interface 130 transmits the image signal to an external display device for playback), the information display module 290 can add a tactical board block to the generated image signal, and display the passing suggestion and the gap position generated by the gap identification module 260 of device 100 on the image signal of the target match image and the tactical board, so that the output module 150 transmits the image signal through the display (or the communication interface 130) to an external display device for playback. In addition to displaying the match screen 410, the target match video played on the external display device can also display the tactical board 420. The match screen 410 can display open positions (411-413), and the tactical board 420 can display open positions 421-423 corresponding to open positions 411-413, and can also display passing suggestions 450 (step 395). When the target match video loaded by the data acquisition module 210 contains match videos from different synchronized perspectives, the input module 120 of the device 100 can allow team members to select one of the perspectives. The information display module 290 can generate an image signal of the target match video corresponding to the selected perspective, thereby enabling the output module 120 to play the target match video from the selected perspective on the display (step 397).

[0117] If, in the above embodiment, the processor 170 of the device 100 can generate the tactical selection module 270, then it can be as follows: Figure 3D As shown in the process, when the suggestion generation module 380 determines that the passing direction does not match the open position (step 381), the tactical selection module 270 can count the number of times the open position of the defending player on the field is distributed in the historical image features when the open position rule is satisfied (step 383), and select an offensive tactic based on the on-field position of the defending player and the number of open position distributions represented by the target image features (step 387), so that the suggestion generation module 380 can generate a passing suggestion based on the offensive tactic selected by the tactical selection module 270 (step 389).

[0118] Furthermore, in the above embodiments, it is even possible to use methods such as " Figure 3EAs shown in the process, the image analysis module 220 of device 100 can identify and track the movement trajectory of the basketball and the on-court position of each player in each historical game image during the process of analyzing historical game images to generate historical image features (steps 310, 315). The model training module 250 of device 100 can train the action prediction model based on the movement trajectory of the basketball and the on-court position of each player analyzed by the image analysis module 220 (step 320). After that, the gap recognition module 260 of device 100 can use the action prediction model to predict the predicted position of each player at a specific time in the future based on the on-court position of the basketball and each player in the target game image (step 375), and can use the gap analysis model to analyze the gap prediction position at a specific time based on the predicted position of each player (step 377).

[0119] In summary, the difference between this invention and existing technologies lies in its technical means of training a gap analysis model using historical match footage, and then using the trained gap analysis model to determine the gap position in the target match footage. When the passing direction of the ball-handling player in the target match footage does not match the gap position, a passing suggestion is generated. This technical means can solve the problem that existing technologies cannot provide players with sufficient opportunities to understand and apply tactics in actual matches, thereby achieving the technical effect of providing a teaching model that can provide experience of the match process and feedback.

[0120] Furthermore, the method of discovering gaps to provide offensive passing suggestions according to the present invention can be implemented in hardware, software, or a combination of hardware and software, or it can be implemented in a centralized manner in a computer system or in a distributed manner with different components distributed among several interconnected computer systems.

[0121] While the embodiments disclosed in this invention are as described above, the content is not intended to directly limit the scope of patent protection for this invention. Any modifications or refinements made by those skilled in the art to which this invention pertains, in terms of form and detail, without departing from the spirit and scope disclosed herein, shall fall within the scope of patent protection for this invention. The scope of patent protection for this invention shall still be determined by the scope defined in the appended claims.

Claims

1. A method for identifying gaps to provide offensive passing suggestions, applied to a device, the method comprising at least the following steps: Load multiple historical game footage, each of which contains a basketball and multiple players, including at least one offensive player and at least one defensive player; Read the empty space rules; Analyze the multiple historical match images to extract multiple historical image features at different time points from each of the historical match images; A gap analysis model is trained using a first machine learning algorithm based on the multiple historical image features that indicate that the distance and relative position between the multiple players satisfy the gap rule. Load the target match video, which contains at least one of the plurality of players; The target game video is analyzed to extract multiple target image features at different time points from the target game video, and the basketball and the on-court positions of each player in the target game video are identified and tracked. Based on the basketball and the on-court positions of the multiple players, when one of the multiple players passes the basketball to another player among the multiple players, the target game scene in the target game image where the basketball is passed is obtained, and the passing direction of the basketball is determined; The gap analysis model is used to analyze at least one passing gap position in the target match footage; and When it is determined that the passing direction does not match the position of at least one passing gap, a passing suggestion is generated.

2. The method for identifying gaps to provide offensive passing suggestions as described in claim 1, wherein, After determining that the passing direction does not match the passing gap position, the method further includes the steps of statistically analyzing the number of gap distributions of the at least one defensive player in each field position represented by the multiple historical image features when the gap rule is satisfied, selecting an offensive tactic based on the field positions of the at least one defensive player represented by the target image features and the number of gap distributions, and generating the passing suggestion with reference to the offensive tactic.

3. The method for identifying gaps to provide offensive passing suggestions as described in claim 1, wherein, The step of training the gap analysis model using the first machine learning algorithm based on the multiple historical image features indicating that the distance and relative position between the multiple players satisfy the gap rule, further includes determining the success or failure status of a pass using at least one rule gap position and identifying at least one defensive player at the edge of the rule gap position based on the multiple historical image features, the multiple pass success or failure status, and the limb parameters of at least one defensive player at the edge of the rule gap position, and training the gap analysis model using the first machine learning algorithm based on the multiple historical image features, the multiple pass success or failure status, and the limb parameters of at least one of the defensive player at the edge of the rule gap position, wherein the limb parameters include at least one of height, arm length, leg length, and stride distance.

4. The method for identifying gaps to provide offensive passing suggestions as described in claim 1, wherein, After generating the passing suggestion, the method further includes, when playing the target match video, displaying the passing suggestion and marking the gap positions generated by analyzing the target match video using the gap analysis model, and when the target match video contains multiple synchronized match videos from different perspectives, providing a step of selecting one of the multiple perspectives to play the match video.

5. The method for identifying gaps to provide offensive passing suggestions as described in claim 1, wherein, After extracting the historical image features at different time points from each of the historical game images, the method further includes a step of training an action prediction model based on the movement trajectories of the basketball and the on-court positions of the players in each of the historical game images. Furthermore, when using the gap analysis model to analyze the at least one passing gap position in the target game image, the method further includes using the action prediction model to predict the predicted position of each player at a specific future time based on the basketball and the on-court positions of the players in the target game image, and using the gap analysis model to analyze the predicted gap position at that specific time based on the predicted positions of the players.

6. A system for identifying gaps to provide offensive passing suggestions, applied to an apparatus, said system comprising at least: Memory for storing at least one computer instruction; and A processor, connected to the memory, is configured to execute the at least one computer instruction to cause the system to perform the following steps: Read the gap rules and load multiple historical game images, each of which contains a basketball and multiple players, including at least one offensive player and at least one defensive player; Analyze the multiple historical match images to extract multiple historical image features at different time points from each of the historical match images; A gap analysis model is trained using a first machine learning algorithm based on the multiple historical image features that indicate that the distance and relative position between the multiple players satisfy the gap rule. Load the target match video, which contains at least one of the plurality of players; The target game image is analyzed to extract multiple target image features at different time points from the target game image, and the basketball and the on-court positions of each player are identified and tracked in the target game image based on the multiple target image features. Based on the basketball and the on-court positions of the multiple players, when the player with the ball among the multiple players passes the basketball to another player among the multiple players, the target game scene in the target game image where the basketball is passed is obtained, and the passing direction of the basketball is determined; The gap analysis model is used to analyze at least one passing gap position in the target match footage; and When it is determined that the passing direction does not match the position of at least one passing gap, a passing suggestion is generated.

7. The system for identifying gaps to provide offensive passing suggestions as described in claim 6, wherein, The system is also used to count the number of times the at least one defensive player, represented by the multiple historical image features, satisfies the gap rule in each position on the field to generate the number of gap distributions, select an offensive tactic based on the position of the at least one defensive player, represented by the target image features, in each position on the field and the number of gap distributions, and generate the passing suggestion based on the offensive tactic.

8. The system for identifying gaps to provide offensive passing suggestions as described in claim 6, wherein, The system is also used to determine the success or failure status of a pass at least one regular open space based on the multiple historical image features and to identify at least one defensive player at the edge of the regular open space. The system trains the open space analysis model using the first machine learning algorithm based on the multiple historical image features, the multiple pass success or failure status, and the limb parameters of at least one defensive player at the edge of the regular open space. The limb parameters include at least one of height, arm length, leg length, and stride distance.

9. The system for identifying gaps to provide offensive passing suggestions as described in claim 6, wherein, The system is also used to display the passing suggestion and mark the gap position generated by analyzing the target match video using the gap analysis model when playing the target match video, and to provide the option to select one of the multiple viewpoints to play the match video when the target match video contains multiple synchronized match videos from different perspectives.

10. The system for identifying gaps to provide offensive passing suggestions as described in claim 6, wherein, The system is also used to identify and track the movement trajectories of the basketball and the on-court positions of the players, train an action prediction model based on the movement trajectories of the basketball and the on-court positions of the players, use the action prediction model to predict the predicted positions of the players at a specific time in the future based on the on-court positions of the basketball and the players in the target game image, and use the gap analysis model to analyze the gap prediction positions at the specific time based on the predicted positions of the players.

Images