Club recognition method, device and storage medium for golf simulator

Abstract

This application discloses a method, device, and storage medium for club recognition in a golf simulator, belonging to the field of image recognition technology. The method includes: acquiring user information; determining a first probability corresponding to each club type based on the user information and the user's current striking position on a virtual golf course; acquiring user striking visual information captured by a camera; identifying a second probability of the club type based on the visual features of the user's striking visual information; calculating a confidence level for the club type based on the first and second probabilities; and determining the user's striking club based on the confidence level.

Description

Technical Field

[0001] This application relates to the field of image recognition technology, and in particular to a method, device and storage medium for club recognition in a golf simulator. Background Technology

[0002] Indoor golf simulators can collect visual data from users during their swing by using visual devices. By analyzing user movements, club trajectory, and the forces acting on the golf ball, they can simulate the real-world trajectory of a golf ball, achieving the effect of using a golf ball indoors.

[0003] In related technologies, golf club identification typically relies on a camera system to capture images of the clubs, compare the image features with club templates in a database one by one, and select the club type with the highest matching degree as the identification result.

[0004] However, when the shot is taken at the boundary of the course type, or when the camera system partially obscures the club due to the user's stance, single-source identification methods are prone to producing unreliable results. For example, when relying solely on scene rules for recommendation, the system may fail to detect discrepancies between the club actually held by the user and the recommended one. Conversely, when relying solely on image comparison, partial occlusion or viewpoint distortion may lead to misclassification of the clubhead features as similar models. Both of these scenarios result in club type identification errors, causing subsequent physical model selection and trajectory calculations to be based on incorrect parameters, thus diminishing the realism of the shot simulation.

[0005] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0006] The main purpose of this application is to provide a golf simulator club recognition method, device and storage medium, which aims to solve the technical problem that the golf simulator's insufficient accuracy in recognizing golf clubs affects the realism of the shot simulation.

[0007] To achieve the above objectives, this application provides a club recognition method for a golf simulator, the method comprising the following steps: Obtain user information, and determine the first probability corresponding to each club type based on the user information and the user's current shot position on the virtual golf course; The system acquires the user's visual information about hitting the ball captured by the camera, and identifies the second probability of the club type based on the visual feature information of the user's visual information about hitting the ball. Based on the first probability and the second probability, the confidence level of the cue type is calculated, and the user's cue is determined based on the confidence level.

[0008] In one embodiment, after the steps of calculating the confidence level of the club type based on the first probability and the second probability, and determining the user's striking club based on the confidence level, the method further includes: Obtain visual information about the flight trajectory of the golf ball captured by the camera; The golf ball trajectory is generated based on the flight trajectory visual information, the hitting club, and the user's hitting visual information. The trajectory of the golf ball is rendered in the virtual golf course.

[0009] In one embodiment, the step of generating the golf ball trajectory based on the flight trajectory visual information, the hitting club, and the user's hitting visual information includes: From the preset club physical parameter library, retrieve the clubface elasticity coefficient, clubhead center of gravity offset, and shaft torque curve corresponding to the hitting club; Using a physics engine, the initial velocity vector and rotation vector of the golf ball are calculated based on the clubface elasticity coefficient, the clubhead center of gravity offset, and the shaft torque curve. The golf ball trajectory is generated based on the initial velocity vector and rotation vector.

[0010] In one embodiment, the step of obtaining user information and determining the first probability corresponding to each club type based on the user information and the user's current shot position on the virtual golf course includes: The real-time environmental parameters of the virtual golf course are obtained, including wind direction, wind speed, and site humidity. The user's physical condition assessment value before the current shot is obtained, and the physical condition assessment value is synchronously uploaded by the vital signs sensor worn by the user; The current hitting position, the real-time environmental parameters, and the physical condition assessment value are input into a pre-trained scenario decision model. The scenario decision model, combined with the user's historical hitting preferences, outputs the first probability corresponding to each club type.

[0011] In one embodiment, the step of acquiring the user's visual information about hitting the ball captured by the camera, and identifying the second probability of the club type based on the visual feature information of the user's visual information about hitting the ball, includes: From the user's visual information about hitting the ball, extract the static appearance features of the club, including clubhead profile curvature, loft scale and / or shaft length ratio; Extract the user's dynamic posture features from the user's visual information of hitting the ball. The dynamic posture features include the distance between the grip hands and / or the width of the stance. By using a pre-trained multimodal recognition network, based on the static appearance features and the dynamic posture features, the second probability of the club type is generated according to the appearance matching degree and posture matching degree.

[0012] In one embodiment, the step of calculating the confidence level of the club type based on the first probability and the second probability, and determining the user's striking club based on the confidence level, includes: Determine the course type corresponding to the hitting position on the golf course, and the club occlusion information in the user's hitting visual information; Based on the field type and the club obstruction information, the reliability information of the first probability and the second probability in the current hitting scenario is evaluated to obtain the first weight and the second weight. Based on the first weight and the second weight, the first probability and the second probability are weighted and summed to obtain the confidence level of the cue type. Based on the confidence level, the club is selected from the preset club types.

[0013] In one embodiment, before the step of acquiring the user's visual information about hitting the ball captured by the camera, and identifying the second probability of the club type based on the visual feature information of the user's visual information about hitting the ball, the method further includes: The video stream captured by the camera is subjected to ball-hitting phase detection to identify the user's ready posture sequence; When the prepared posture sequence matches the preset club display posture and the duration exceeds the preset threshold, the video segment corresponding to the prepared posture sequence is marked as the target segment. The user's visual information about hitting the ball is selected based on the clarity of the club area in the target segment.

[0014] In one embodiment, after the steps of calculating the confidence level of the club type based on the first probability and the second probability, and determining the user's striking club based on the confidence level, the method further includes: The camera captures a sequence of images of the moment the user hits the ball, and the actual initial launch angle and actual initial linear velocity of the golf ball are calculated. The actual initial launch angle and the actual initial linear velocity are compared with the expected launch angle and expected linear velocity estimated by the scenario decision model when identifying the current shot club, and a deviation evaluation value is generated. The deviation evaluation value is used as a training signal to update the weight parameters corresponding to the historical shot preference in the scenario decision model.

[0015] In addition, to achieve the above objectives, this application also provides a club recognition device for a golf simulator, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the club recognition method for a golf simulator as described above.

[0016] In addition, to achieve the above objectives, this application also provides a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the golf simulator club recognition method as described above.

[0017] One or more technical solutions proposed in this application have at least the following technical effects: This application obtains user information and combines it with the user's current hitting position on a virtual golf course to determine the first probability corresponding to each club type. At the same time, it obtains the user's hitting visual information captured by the camera and identifies the second probability of each club type based on the visual feature information therein. The first probability and the second probability are fused to calculate the confidence level of each club type, and the user's hitting club is determined based on the confidence level. This makes the club determination stable in complex usage environments, thereby ensuring that the physical parameter model called for subsequent trajectory calculation matches the club type actually used by the user, thus improving the realism of the virtual hitting simulation. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the first embodiment of the golf simulator club recognition method of this application. Figure 2 This is a flowchart illustrating the second embodiment of the golf simulator club recognition method of this application. Figure 3 This is a flowchart illustrating the third embodiment of the golf simulator club recognition method of this application. Figure 4 This is a flowchart illustrating the fourth embodiment of the golf simulator club recognition method of this application. Figure 5This is a schematic diagram of the structure of a golf simulator club recognition device in the hardware operating environment of the embodiment of this application.

[0021] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0022] It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0023] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0024] The main solution of this application embodiment is: to obtain user information, and based on the user information and the user's current hitting position on the virtual golf course, to determine the first probability corresponding to each club type; to obtain the user's hitting visual information captured by the camera, and based on the visual feature information of the user's hitting visual information, to identify the second probability of the club type; to calculate the confidence level of the club type based on the first probability and the second probability, and to determine the user's hitting club based on the confidence level.

[0025] In current technologies, golf club identification typically relies on camera systems to capture club images. These images are then compared one by one with club templates in a database, and the club type with the highest matching degree is selected as the identification result. However, when the shot is taken at the boundary of a course type, or when the user's position causes partial obstruction of the club, this single-source identification method is prone to producing unreliable results. For example, relying solely on scene rules for recommendations may fail to detect discrepancies between the club actually held by the user and the recommended one. Similarly, relying solely on image comparison may lead to partial obstruction or perspective distortion causing the clubhead features to be misidentified as a similar model. Both of these scenarios result in club type identification errors, which in turn cause subsequent physical model selection and trajectory calculations to be based on incorrect parameters, diminishing the realism of the shot simulation.

[0026] This application obtains user information and combines it with the user's current hitting position on a virtual golf course to determine the first probability corresponding to each club type. At the same time, it obtains the user's hitting visual information captured by the camera and identifies the second probability of each club type based on the visual feature information therein. The first probability and the second probability are fused to calculate the confidence level of each club type, and the user's hitting club is determined based on the confidence level. This makes the club determination stable in complex usage environments, thereby ensuring that the physical parameter model called for subsequent trajectory calculation matches the club type actually used by the user, thus improving the realism of the virtual hitting simulation.

[0027] To better understand the above technical solutions, exemplary embodiments of this application will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of this application are shown in the drawings, it should be understood that this application can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of this application and to fully convey the scope of this application to those skilled in the art.

[0028] It should be noted that the executing entity in this embodiment can be a golf simulator, or a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device capable of performing the above functions, such as a golf simulator club recognition device. This embodiment does not specifically limit it in this way. The following uses a golf simulator as an example to describe this embodiment and the following embodiments.

[0029] Based on this, embodiments of this application provide a club recognition method for a golf simulator, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the golf simulator club recognition method of this application.

[0030] In this embodiment, the golf simulator club recognition method includes steps S10~S30: Step S10: Obtain user information, and determine the first probability corresponding to each club type based on the user information and the user's current hitting position on the virtual golf course; In this embodiment, the processor obtains the user information of the current golfer through identity recognition and reads the current location of the golf ball from the virtual golf course engine. Using both user information and location as the basis for querying, the processor statistically analyzes the probability of using each club type in the current scenario from the user's historical golfing behavior records, thereby generating a first probability distribution. Here, the current hitting position of the virtual golf course refers to the coordinates of the golf ball's current location in the 3D golf course scene presented by the simulator and its corresponding course type label. The course type label indicates whether the location belongs to the tee box, fairway, rough, bunker, or green. The first probability refers to the prior probability estimate of each club type the user might use, based solely on user identity, hitting position, and historical experience data, without relying on real-time visual observation.

[0031] Optionally, a golf simulator refers to an electronic device installed on an indoor hitting table to collect user's hitting behavior data and simulate the trajectory of a golf ball. It includes the hitting table, a camera, sensor components, a processor, a memory, and an output device. The hitting table is a physical platform for the user to stand on and perform the swing. At least one camera is installed directly in front of or slightly above the side of the hitting table, with its field of view covering the user's standing area and the space where the club swings. This camera collects video streams during the user's preparation phase and the hitting process. The video streams include the user's visual information about the shot, visual information about the golf ball's flight trajectory, and a sequence of images at the moment of impact. The sensor components include vital signs sensors and / or environmental sensors. The vital signs sensors are worn by the user and collect an assessment of the user's physical condition before hitting the ball, which is then synchronously uploaded to the processor. The environmental sensors, including wind speed and direction sensors and / or humidity sensors, are located outside the hitting table and collect real-time environmental parameters, including wind direction, wind speed, and field humidity. The memory stores a computer program, a preset list of golf club types, a golf club physical parameter database, a user preference database, and a golf club type database. The golf club physical parameter database stores physical parameters such as the clubface elasticity coefficient, clubhead center of gravity offset, and shaft torque curve for each club type. The user preference database stores each user's historical shot records to support the determination of the first probability. The golf club type database stores the standard name, model number, and corresponding physical parameters for each club type. When the computer program is executed by the processor, it implements the golf club recognition method of the various embodiments of this application. The processor is connected to a camera, sensor components, memory, and an output device to execute the golf club recognition method, realizing golf club type recognition, golf club recommendation, and the generation and rendering of the golf ball's trajectory. The processor also runs a virtual golf course engine and a physics engine. The virtual golf course engine is used to construct and render a virtual golf course scene, and the physics engine is used to calculate the golf ball's trajectory based on the golf club's physical parameters. The output device includes at least one of a display device, a head-mounted display device, and a voice broadcasting device, used to display the virtual golf course scene, the golf ball's trajectory, and recommendation information to the user.

[0032] Specifically, the processor's process of determining the first probability includes four sub-actions: identity acquisition, scene parsing, history matching, and probability calculation. In the identity acquisition sub-action, the processor obtains the current user's identity identifier by reading the biometric module or user interaction terminal set on the golf simulator. In the scene parsing sub-action, the processor establishes data communication with the virtual golf course engine, reads the three-dimensional coordinates of the golf ball's location in the current simulated scene from the engine's memory state, spatially maps these coordinates to a grid representing the course area, determines the course type label to which the coordinates belong, and calculates the Euclidean distance between the hole coordinates as the endpoint and the current ball coordinates as the starting point, using this distance as the target distance. In the history matching sub-action, the processor uses the identity identifier as an index to retrieve the user's historical shot record table from the user preference database. Each row of the historical shot record table stores the course type, target distance, and club type used for a past shot. The processor uses the currently parsed course type label and target distance value as filtering conditions, searching the historical shot record table for all historical shot entries with the same course type and target distance deviation within a preset tolerance range, forming a candidate historical entry set. In the probability calculation sub-action, the processor traverses the preset list of cue types. For each cue type, it counts the number of times it appears in the candidate history entry set. The number of times it appears is divided by the total number of entries in the candidate history entry set. The resulting ratio is the first probability corresponding to that cue type.

[0033] As an optional implementation, when a user is newly registered and there are not enough entries in the historical shot record table, the processor does not execute the historical matching sub-action, but instead calls a preset general club recommendation rule base. The general club recommendation rule base stores the mapping relationship between course type, target distance, and recommended club types. The processor compares the current course type and target distance with the conditional parts of each rule in the rule base one by one. The recommended club type corresponding to the successfully matched rule receives a preset base probability value, and the remaining club types receive an equally distributed residual probability value. This method allows the system to provide a usable first probability estimate based on general experience even in the cold start phase when there is a lack of individual user historical data.

[0034] As an alternative implementation, in the historical matching sub-actions, the processor not only uses venue type and target distance as filtering conditions, but also introduces real-time environmental parameters at the time of the shot as an additional filtering dimension. These real-time environmental parameters are collected by wind speed and direction sensors and humidity sensors located outside the simulator and uploaded to the processor. The processor divides the environmental parameters into several levels and simultaneously matches entries in the historical shot record table that are similar in venue type, target distance, and environmental level. By adding constraints to the environmental parameters, the filtered candidate entries are closer to the actual current shot conditions, thereby improving the accuracy of the first probability in reflecting the user's true club-playing intentions.

[0035] Optionally, the processor can also use a camera to capture an image sequence of the moment the ball is released, calculate the actual initial launch angle and the actual initial linear velocity of the ball, compare the actual initial launch angle and the actual initial linear velocity with the expected launch angle and expected linear velocity estimated by the scene decision model when identifying the current shot, generate a deviation evaluation value, and use the deviation evaluation value as a training signal to update the weight parameters corresponding to historical shot preferences in the scene decision model.

[0036] Step S20: Obtain the user's visual information of hitting the ball captured by the camera, and identify the second probability of the club type based on the visual feature information of the user's visual information of hitting the ball; In this embodiment, the processor utilizes a high-speed camera equipped in the golf simulator to capture real-time image data containing the user's body and club as the user's visual information for hitting the ball, during the stage when the user completes their stance and grips the club in the ready-to-hit posture. Subsequently, a pre-trained deep learning model extracts visual feature information related to the club type from this image data. The extracted visual feature information is then matched and compared with template features for each preset club type, outputting a second probability corresponding to each club type. The user's visual information for hitting the ball refers to the image data obtained by the camera configured in the golf simulator during the user's ready-to-hit action, capturing the user's body and club. This image data clearly shows the user's grip posture and the club's appearance. Visual feature information refers to the quantitative descriptive parameters extracted from the user's visual information for hitting the ball, used by computer vision algorithms to distinguish different club types. These include the clubhead's contour shape, the angle between the clubface and shaft, and the length ratio of the shaft relative to a reference scale. The second probability refers to the visual matching probability estimate given for each type of golf club, without relying on prior information such as user identity and shot position, but based entirely on visual evidence such as the appearance of the club and the user's posture captured by the camera.

[0037] Optionally, the processor's execution of the second probabilistic recognition process includes three sub-actions: image acquisition, feature extraction, and probability mapping. In the image acquisition sub-action, the processor controls a camera to continuously capture images of the hitting area at a preset frame rate. The camera is mounted directly in front of or slightly above the side of the hitting table in the golf simulator, its field of view covering the user's standing area and the space where the club swings. The processor analyzes the video stream output by the camera in real time to detect whether the user has entered the ready-to-hit posture. The detection criteria for the ready-to-hit posture include whether the user's foot stance is stable, whether the line connecting their shoulders points in the target direction, and whether the clubhead is positioned behind the ball. When the processor detects that the user has maintained the ready-to-hit posture for a duration exceeding a preset frame count threshold, it selects the single frame with the highest image clarity within the corresponding time period as the user's visual information about the hit. The clarity is judged based on the Laplace gradient value of the club area in the image.

[0038] Furthermore, in the feature extraction sub-action, the processor inputs the user's visual information about hitting the ball into a pre-trained convolutional neural network model. The convolutional neural network model progressively extracts low-level geometric features such as the clubhead edge contour and the shaft's straight-line direction through multiple stacked convolutional layers. These low-level geometric features are then combined into high-level semantic features such as clubhead shape type, loft angle, and shaft length ratio. Finally, a fully connected layer unfolds these high-level semantic features into a one-dimensional feature vector as visual feature information. In the probability mapping sub-action, an output layer follows the fully connected layer. The number of neurons in the output layer is the same as the number of preset club types. Each neuron corresponds to one club type and outputs an initial matching score. The processor applies a normalized exponential function to the initial matching scores of the output layer, transforming each score into a probability distribution that sums to 1. This probability distribution represents the second probability corresponding to each club type.

[0039] As an optional implementation, the training process of the convolutional neural network model employs a transfer learning strategy. Transfer learning consists of two stages. The first stage uses a publicly available large-scale object image dataset to pre-train the model's basic convolutional layers, enabling them to possess general edge and texture extraction capabilities. This stage does not involve club-specific labeled data. The second stage uses a labeled dataset actually collected on a golf simulator to fine-tune the model's fully connected layers and output layer. Each image in the labeled dataset is labeled by professionals with the actual type of golf club used. During training, a cross-entropy loss function and a stochastic gradient descent optimizer are used, and the model parameters are updated round by round through backpropagation until the model's recognition accuracy on the validation set reaches a preset standard. The advantage of transfer learning is that the general feature extraction capabilities learned by the basic convolutional layers from large-scale data can be transferred to the golf club recognition task, allowing the model to achieve high recognition accuracy even with relatively limited labeled golf club samples.

[0040] Optionally, a sequence of consecutive frame images of the user from the moment of takeoff to the moment of impact is acquired by the camera. For each frame in the consecutive frame image sequence, the clubface angle is extracted to generate a clubface angle change curve. The clubface angle change curve is matched temporally with the standard swing curve template corresponding to each club type, and the second probability corresponding to each club type is determined based on the curve similarity.

[0041] As an alternative implementation, the processor is not limited to a single frame image in the feature extraction sub-action. Instead, it selects frames from multiple viewpoints or time points in the image sequence during the ball preparation phase, extracts the visual feature information of each frame, performs average pooling on the multi-frame features, and uses the fused features after pooling as the final visual feature information for probability mapping. This method utilizes the complementary information between multiple frames. When the clubhead in one frame is locally overexposed due to light reflection or the user's slight movement causes slight motion blur, the clear features of the same area in other frames can suppress noise through average pooling, thereby reducing the impact of random factors in a single frame image on the recognition result.

[0042] Step S30: Calculate the confidence level of the cue type based on the first probability and the second probability, and determine the user's cue based on the confidence level.

[0043] In this embodiment, the processor fuses information from the first probability distribution and the second probability distribution. For each candidate club type, the processor combines its first probability value in the first probability distribution and its second probability value in the second probability distribution according to a preset fusion rule to calculate the comprehensive confidence level of that club type. Then, the type with the highest confidence level is selected from all candidate club types as the final identification result, and this result is output to the physics engine of the virtual golf course for subsequent shot trajectory simulation. This confidence level reflects the statistical regularity of the user's historical preferences and the physical evidence of current visual observation, thus using the golf club as the final club identification result based on this shot simulation.

[0044] Specifically, the process of the processor performing confidence calculation and cue selection includes three sub-actions: weight acquisition, weighted fusion, and result selection. In the weight acquisition sub-action, the processor reads preset first and second weight coefficients. The first weight coefficient corresponds to the importance of the first probability in the fusion calculation, and the second weight coefficient corresponds to the importance of the second probability in the fusion calculation. Both the first and second weight coefficients are non-negative real numbers, and their sum is 1. The weight coefficients are set dynamically and adaptively; the processor evaluates the reliability of the first and second probabilities in the current scenario and adjusts the weight coefficients in real time based on the evaluation results. In the weighted fusion sub-action, the processor iterates through each candidate cue type in the preset cue type list, reads the second probability value obtained for that type, multiplies the first probability value by the first weight coefficient to obtain the first weighted term, multiplies the second probability value by the second weight coefficient to obtain the second weighted term, and adds the first weighted term and the second weighted term to obtain the confidence level for that cue type. In the result selection sub-action, the processor sorts the confidence scores of all candidate club types from high to low, selects the club type with the highest confidence score as the recognition result for this shot, retrieves the corresponding physical parameter model from the club physical parameter database based on the recognized club type, and sends the physical parameter model to the physics engine module of the virtual golf course so that the physics engine can use mechanical parameters that match the actual club when calculating the trajectory of the golf ball.

[0045] As an optional implementation, the processor employs a dynamic weight adjustment strategy based on occlusion detection and court boundary determination in the weight acquisition sub-action. When the processor detects that the area of ​​the cue stick obscured by a body or other object in the camera image exceeds a preset proportion, the reliability of the second probability determination decreases, and the first weight coefficient is automatically increased to 0.7 and the second weight coefficient is decreased to 0.3, making the fusion result more dependent on historical statistical data. When the processor detects that the court type label of the current hitting position is in the transition area between two court types, i.e., the distance between the current position coordinates and the boundary line between the two court types is less than a preset threshold, the court type-based scene rule determination may result in a mismatch, the reliability of the first probability decreases, and the first weight coefficient is automatically decreased to 0.3 and the second weight coefficient is increased to 0.7, making the fusion result more dependent on visual observation. When neither the occlusion area nor the court boundary distance triggers the above adjustment conditions, the first and second weight coefficients remain at their default values ​​of 0.5, indicating that scene priors and visual perception are equally important in the fusion process.

[0046] As an alternative implementation, the processor introduces a confidence level discrimination mechanism in the result selection sub-action. When the difference between the highest confidence level and the second highest confidence level is greater than a preset discrimination threshold, the processor directly selects the club type corresponding to the highest confidence level as the recognition result. When the difference between the highest confidence level and the second highest confidence level is less than or equal to the preset discrimination threshold, the processor determines that the current recognition is ambiguous. In this case, the processor does not automatically select the club with the highest confidence level, but instead controls the golf simulator's display screen to pop up a confirmation interface for the user. The confirmation interface displays the names and images of the two club types with the highest and second highest confidence levels for the user to manually select. The processor receives the selection operation input by the user through the touch screen or voice command and takes the club type selected by the user as the final recognition result. This mechanism introduces the user as the final arbiter in critical situations where neither scene rules nor visual features can provide a clear judgment, avoiding the impact on the realism of subsequent shot simulations due to the system forcibly determining an incorrect result.

[0047] This application embodiment obtains user information and combines it with the user's current hitting position on a virtual golf course to determine the first probability corresponding to each club type. At the same time, it obtains the user's hitting visual information captured by the camera and identifies the second probability of each club type based on the visual feature information therein. The first probability and the second probability are fused to obtain the confidence level of each club type, and the user's hitting club is determined based on the confidence level. This makes the club determination stable in complex usage environments, thereby ensuring that the physical parameter model called for subsequent trajectory calculation matches the club type actually used by the user, thus improving the realism of the virtual hitting simulation.

[0048] Based on the same inventive concept, this application also provides a second embodiment, referring to... Figure 2 , Figure 2 This is a flowchart illustrating the second embodiment of the golf simulator club recognition method of this application.

[0049] In this embodiment, the golf simulator club recognition method further includes steps S41-S43: Step S41: Obtain visual information of the flight trajectory of the golf ball captured by the camera; Step S42: Generate the golf ball trajectory based on the flight trajectory visual information, the hitting club, and the user's hitting visual information; Step S43: Render the trajectory of the golf ball in the virtual golf course.

[0050] In this embodiment, after determining the user's club strike, the processor further utilizes a camera to continuously track the golf ball's flight path after it is hit, acquiring visual information about the flight trajectory. This information is then combined with the physical parameters of the club and the swing motion features extracted from the user's visual information to generate the golf ball's trajectory in three-dimensional space. Finally, the trajectory is rendered and overlaid onto the virtual golf course in a visual manner. The flight trajectory visual information refers to a sequence of multiple frames captured by the camera at a high frame rate after the golf ball is hit. Each frame records the golf ball's position coordinates on the imaging plane. By performing triangulation or depth estimation on the ball's position between consecutive frames, the ball's path in real three-dimensional space can be reconstructed. The golf ball's trajectory is a spatial curve formed by connecting all the positions the ball's center of mass experiences in three-dimensional space in chronological order from the moment it is hit by the club until it lands on the ground. This curve is calculated frame-by-frame by the physics engine based on the initial hitting conditions and aerodynamic equations.

[0051] Specifically, the processor performs motion trajectory generation and rendering as follows: During the flight trajectory visual information acquisition stage, the processor immediately switches the camera to high-speed tracking mode after recognizing the striking club. The camera's frame rate is increased to hundreds of frames per second or more, and the exposure time is shortened to freeze the rapidly moving golf ball image. The processor performs background subtraction and circular target detection on each frame to locate the golf ball's subpixel-level center coordinates within the image. When the golf ball flies out of the camera's field of view, the processor terminates image acquisition, using the entire sequence of acquired ball coordinates as the flight trajectory visual information.

[0052] Optionally, from a preset library of club physics parameters, the clubface elasticity coefficient, clubhead center of gravity offset, and shaft torque curve corresponding to the striking club are retrieved. Using a physics engine, the initial velocity vector and rotation vector of the golf ball are calculated based on the clubface elasticity coefficient, clubhead center of gravity offset, and shaft torque curve. Based on the initial velocity vector and rotation vector, the trajectory of the golf ball is generated.

[0053] Optionally, the flight trajectory of the golf ball is calculated based on the club's striking action and the user's swing. The flight trajectory is then corrected based on visual information. The corrected flight trajectory is matched against the visual information, and the portion of the trajectory to be rendered is determined based on the matching result.

[0054] Optionally, during the trajectory generation stage, the processor retrieves the clubface elasticity coefficient, clubhead center of gravity offset, and shaft torque curve corresponding to the striking club from the club physics parameter library. Simultaneously, it extracts the swing trajectory arc and the clubhead velocity at impact from the user's visual information, inputting these parameters into the physics engine. The physics engine starts with the momentum and angular momentum conservation equations for the clubface-ball collision, calculating the initial velocity and initial rotation vectors at the moment of impact. Then, using these initial velocity and rotation vectors as boundary conditions, it substitutes the air resistance equation, Magnus effect equation, and gravity equation frame by frame to solve for the spatial position coordinates and velocity direction of the golf ball at each time step. The position coordinates of all time steps are then concatenated to form the golf ball's trajectory. During the rendering stage, the processor sends the frame-by-frame sequence of position coordinates of the trajectory to the rendering module of the virtual golf course. The rendering module generates a trajectory line with color or transparency gradients along the trajectory coordinates in the 3D golf course scene, simultaneously generating an animation of a 3D model of the golf ball moving synchronously along the trajectory line. This animation is then output to the display screen or projection screen of the golf simulator.

[0055] This embodiment, based on club recognition, seamlessly integrates the recognition results into the generation and rendering of the motion trajectory, ensuring that the hitting parameter model called by the physics engine strictly corresponds to the type of club actually used by the user. This guarantees the physical authenticity of the trajectory simulation from the source, while transforming the abstract calculation results into a visual flight trajectory image that the user can see, allowing the user to intuitively observe the hitting effect in the virtual golf course.

[0056] Since the system described in Embodiment 2 of this application is a system used to implement the method of Embodiment 1 of this application, those skilled in the art can understand the specific structure and variations of the system based on the method described in Embodiment 1 of this application, and therefore will not be described again here. All systems used in the method of Embodiment 1 of this application fall within the scope of protection of this application.

[0057] Based on the same inventive concept, this application also provides a third embodiment, referring to... Figure 3 , Figure 3 This is a flowchart illustrating the third embodiment of the golf simulator club recognition method of this application.

[0058] In this embodiment, the golf simulator club recognition method further includes steps S51-S53: Step S51: Capture the image sequence of the moment the user hits the ball using the camera, and calculate the actual initial launch angle and actual initial linear velocity of the golf ball; Step S52: Compare the actual initial launch angle and the actual initial linear velocity with the expected launch angle and expected linear velocity estimated by the scene decision model when identifying the current shot club, and generate a deviation evaluation value; Step S53: Use the deviation evaluation value as a training signal to update the weight parameters corresponding to the historical shot preference in the scenario decision model.

[0059] In this embodiment, after completing club recognition and shot simulation, the processor captures images of the actual ball's release moment using a camera and calculates the actual flight parameters. These actual parameters are compared with the expected flight parameters predicted by the scene decision model during the recognition process. The resulting deviation information is used to update the scene decision model, forming a closed loop from recognition to evaluation to optimization. The release moment image sequence refers to a series of frames captured by the camera at a high frame rate within a very short time after the golf ball is hit by the club. These images record the ball's motion state the instant it leaves the clubface. The displacement of the ball's position between frames can be used to deduce the ball's initial motion parameters. The actual initial launch angle is the angle between the velocity vector direction calculated from the difference in ball position between two adjacent frames in the release moment image sequence and the horizontal plane. This angle reflects the ball's tilt in the vertical direction during takeoff. The actual initial linear velocity is the velocity of the ball's center of mass calculated from the difference in ball position between two adjacent frames and the time interval between frames in the release moment image sequence. The deviation assessment value is a quantitative difference index generated by comparing the actual measured parameters with the expected parameters. This index is used to measure the degree of agreement between the identification result and the actual hitting effect.

[0060] Specifically, the processor performs feedback optimization as follows: In the ball-launch parameter acquisition phase, after detecting a shot, the processor immediately extracts several frames before and after the shot from the camera's high-speed image buffer to form a shot-launch image sequence. For adjacent frames in the image sequence, the processor locates the ball region using inter-frame difference and extracts the pixel coordinates of the ball's centroid. Combining this with the camera's intrinsic parameter matrix and calibration parameters, the pixel coordinates are converted into three-dimensional spatial coordinates in the camera coordinate system. Based on the difference in three-dimensional spatial coordinates between adjacent frames and the inter-frame time interval, the processor calculates the ball's velocity vector in the initial stage after the shot. It decomposes the velocity vector into horizontal and vertical components, calculates the ratio of the vertical to horizontal components using the arctangent function to obtain the actual initial launch angle, and calculates the magnitude of the velocity vector to obtain the actual initial linear velocity. In the deviation comparison phase, the processor reads the expected launch angle and expected linear velocity generated synchronously by the scene decision model when recognizing the shot club from the model's inference log. The expected launch angle and expected linear velocity are reference values ​​estimated by the scene decision model based on the recognized club type, the user's historical shot data, and current environmental parameters. The processor divides the absolute value of the difference between the actual initial launch angle and the expected launch angle by the expected launch angle to obtain the launch angle deviation rate. It also divides the absolute value of the difference between the actual initial linear velocity and the expected linear velocity by the expected linear velocity to obtain the linear velocity deviation rate. The processor then weights and sums the launch angle deviation rate and the linear velocity deviation rate to obtain the deviation evaluation value. During the model update phase, the processor uses the deviation evaluation value as a training signal and employs an online gradient descent method to update the user's historical shot preference weight parameters stored in the scene decision model. The update direction is to make the predicted launch angle and expected linear velocity for the same club type in similar scenarios closer to the actual measured values. The magnitude of the update is positively correlated with the magnitude of the deviation evaluation value; the larger the deviation, the larger the parameter adjustment.

[0061] This application embodiment introduces feedback on the deviation between the actual shot effect and the expected effect after each shot, enabling the scenario decision model to continuously learn from the user's real shot data and gradually correct the model's internal parameters to adapt to the user's unique swing habits and force characteristics. Over time, the model's accuracy in predicting the user's intention to use the club gradually improves, thereby driving the continuous optimization of the recognition accuracy of the entire club recognition system in long-term use.

[0062] Since the system described in Embodiment 3 of this application is a system used to implement the method of Embodiment 1 of this application, those skilled in the art can understand the specific structure and variations of the system based on the method described in Embodiment 1 of this application, and therefore will not be described again here. All systems used in the method of Embodiment 1 of this application fall within the scope of protection of this application.

[0063] Based on the same inventive concept, this application also provides a fourth embodiment, referring to... Figure 4 , Figure 4 This is a flowchart illustrating the fourth embodiment of the golf simulator club recognition method of this application.

[0064] In this embodiment, the golf simulator club recognition method further includes steps S11 to S13: Step S11: Based on the first probability corresponding to each club type, select the club type with the highest first probability as the recommended club; Step S12: Generate recommendation information based on the recommended cue stick; Step S13: Output the recommendation information to the user to prompt the user to use the recommended cue to hit the ball.

[0065] In this embodiment, the golf simulator recommends the best golf club to the user before the user begins the shot. The processor can select the best club based on the current hitting position, or it can recommend clubs by combining the current hitting position with user preferences from the user information.

[0066] Specifically, the processor sorts each club type in descending order according to its first probability value, selects the club type with the highest first probability value at the top of the list as the recommended club, reads the type identifier of the recommended club, generates recommendation information containing the type identifier and the corresponding first probability value, and sends the recommendation information to the display device or user interaction terminal configured on the golf simulator for display, so as to prompt the user to give priority to using the recommended club in this shot.

[0067] Optionally, since the recommended club is a priori result inferred only from the user's historical hitting preferences and the current hitting scenario, when the user adopts the recommended club to hit the ball, the user's hitting visual information captured by the camera points to the same club type as the recommended club. The second probability obtained is mutually corroborated with the first probability, thereby improving the reliability of the confidence obtained by fusion.

[0068] Optionally, the processor iterates through a preset list of cue types. For each cue type in the list, it reads the first probability value corresponding to that cue type from a generated first probability distribution. Using the first probability value as the sorting criterion, the cue types are arranged in descending order of their first probability values, resulting in an ordered sequence of cue types. The cue type appearing earlier in the sequence has a higher prior probability of being used by the user. The processor selects the cue type that appears first in the ordered sequence, i.e., the cue type with the highest first probability value, as the recommended cue for this shot and writes it to a temporary cache for use in subsequent steps.

[0069] For example, when the first probability difference between the first and second ranked cue types in an ordered sequence is less than a preset difference threshold, the processor determines that the two are not statistically distinguishable. In this case, the processor does not select the first ranked cue type alone, but selects both the first and second ranked cue types as recommended cues. This ensures that the subsequently generated recommendation information includes two candidate cue types and their respective first probability values, allowing the user to choose from the two candidates to avoid missing similar cue types that the user might use in a single recommendation result.

[0070] This application provides a golf simulator club recognition device, the device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the golf simulator club recognition method of the above embodiment 1.

[0071] The following is for reference. Figure 5 The diagram illustrates a structural schematic of a club recognition device suitable for implementing the golf simulator embodiments of this application. The club recognition device for the golf simulator in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 5 The golf simulator club recognition device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0072] like Figure 5As shown, the club recognition device of a golf simulator may include a processing unit 1001 (e.g., a core processor, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 1002 or a program loaded from storage device 1003 into random access memory (RAM) 1004. The random access memory 1004 also stores various programs and data required for the operation of the golf simulator's club recognition device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the golf simulator's club recognition device to communicate wirelessly or wiredly with other devices to exchange data. Although golf simulator club recognition devices with various systems are shown in the figures, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems may be implemented alternatively.

[0073] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0074] The golf simulator club recognition device provided in this application, employing the golf simulator club recognition method described in the above embodiments, can solve the technical problem of insufficient accuracy in golf simulator club recognition, which affects the realism of the shot simulation. Compared with the prior art, the beneficial effects of the golf simulator club recognition device provided in this application are the same as those of the golf simulator club recognition method provided in the above embodiments, and other technical features in this golf simulator club recognition device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0075] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0076] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0077] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the golf simulator club recognition method in the above embodiments.

[0078] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, radio frequency (RF), etc., or any suitable combination thereof.

[0079] The aforementioned computer-readable storage medium may be included in the club recognition device of the golf simulator; or it may exist independently and not be assembled into the club recognition device of the golf simulator.

[0080] The aforementioned computer-readable storage medium carries one or more programs that, when executed by the club recognition device of the golf simulator, cause the club recognition device of the golf simulator to: acquire user information, and determine a first probability corresponding to each club type based on the user information and the user's current hitting position on the virtual golf course; acquire user hitting visual information captured by a camera, and identify a second probability of the club type based on the visual feature information of the user hitting visual information; calculate the confidence level of the club type based on the first and second probabilities, and determine the user's hitting club based on the confidence level.

[0081] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0082] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0083] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0084] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the club recognition method of the golf simulator described above. This solves the technical problem of insufficient accuracy in golf club recognition by the golf simulator, which affects the realism of the shot simulation. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the club recognition method of the golf simulator provided in the above embodiments, and will not be repeated here.

[0085] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for club recognition in a golf simulator, characterized in that, An application is made in a golf simulator, which includes a camera and a processor. The processor determines the trajectory of a golf ball based on a video stream captured by the camera and maps the trajectory of the golf ball onto a virtual golf course. The method includes the following steps: Obtain user information, and determine the first probability corresponding to each club type based on the user information and the user's current shot position on the virtual golf course; The system acquires the user's visual information about hitting the ball captured by the camera, and identifies the second probability of the club type based on the visual feature information of the user's visual information about hitting the ball. Based on the first probability and the second probability, the confidence level of the cue type is calculated, and the user's cue is determined based on the confidence level.

2. The golf simulator club recognition method as described in claim 1, characterized in that, After the steps of calculating the confidence level of the club type based on the first probability and the second probability, and determining the user's striking club based on the confidence level, the method further includes: Obtain visual information about the flight trajectory of the golf ball captured by the camera; From the preset club physical parameter library, retrieve the clubface elasticity coefficient, clubhead center of gravity offset, and shaft torque curve corresponding to the hitting club; Using a physics engine, the initial velocity vector and rotation vector of the golf ball are calculated based on the clubface elasticity coefficient, the clubhead center of gravity offset, and the shaft torque curve. Based on the initial velocity vector and rotation vector, a golf ball trajectory is generated and rendered in the virtual golf course.

3. The golf simulator club recognition method as described in claim 1, characterized in that, The step of obtaining user information and determining the first probability corresponding to each club type based on the user information and the user's current shot position on the virtual golf course includes: The real-time environmental parameters of the virtual golf course are obtained, including wind direction, wind speed, and site humidity. The user's physical condition assessment value before the current shot is obtained, and the physical condition assessment value is synchronously uploaded by the vital signs sensor worn by the user; The current hitting position, the real-time environmental parameters, and the physical condition assessment value are input into a pre-trained scenario decision model. The scenario decision model, combined with the user's historical hitting preferences, outputs the first probability corresponding to each club type.

4. The golf simulator club recognition method as described in claim 1, characterized in that, After the steps of obtaining user information and determining the first probability corresponding to each club type based on the user information and the user's current shot position on the virtual golf course, the method further includes: Based on the first probability corresponding to each club type, the club type with the highest first probability is selected as the recommended club. Based on the recommended club, generate recommendation information; The recommended information is output to the user to prompt the user to use the recommended cue to hit the ball.

5. The golf simulator club recognition method as described in claim 1, characterized in that, The step of acquiring the user's visual information about hitting the ball captured by the camera, and identifying the second probability of the club type based on the visual feature information of the user's visual information about hitting the ball, includes: From the user's visual information about hitting the ball, extract the static appearance features of the club, including clubhead profile curvature, loft scale and / or shaft length ratio; Extract the user's dynamic posture features from the user's visual information of hitting the ball. The dynamic posture features include the distance between the grip hands and / or the width of the stance. By using a pre-trained multimodal recognition network, based on the static appearance features and the dynamic posture features, the second probability of the club type is generated according to the appearance matching degree and posture matching degree.

6. The golf simulator club recognition method as described in claim 1, characterized in that, The step of calculating the confidence level of the club type based on the first probability and the second probability, and determining the user's striking club based on the confidence level, includes: Determine the course type corresponding to the hitting position on the golf course, and the club occlusion information in the user's hitting visual information; Based on the field type and the club obstruction information, the reliability information of the first probability and the second probability in the current hitting scenario is evaluated to obtain the first weight and the second weight. Based on the first weight and the second weight, the first probability and the second probability are weighted and summed to obtain the confidence level of the cue type. Based on the confidence level, the club is selected from the preset club types.

7. The golf simulator club recognition method as described in claim 1, characterized in that, Before the step of acquiring the user's visual information about hitting the ball captured by the camera, and identifying the second probability of the club type based on the visual feature information of the user's visual information about hitting the ball, the method further includes: The video stream captured by the camera is subjected to ball-hitting phase detection to identify the user's ready posture sequence; When the prepared posture sequence matches the preset club display posture and the duration exceeds the preset threshold, the video segment corresponding to the prepared posture sequence is marked as the target segment. The user's visual information about hitting the ball is selected based on the clarity of the club area in the target segment.

8. The golf simulator club recognition method as described in claim 1, characterized in that, After the steps of calculating the confidence level of the club type based on the first probability and the second probability, and determining the user's striking club based on the confidence level, the method further includes: The camera captures a sequence of images of the moment the user hits the ball, and the actual initial launch angle and actual initial linear velocity of the golf ball are calculated. The actual initial launch angle and the actual initial linear velocity are compared with the expected launch angle and expected linear velocity estimated by the scenario decision model when identifying the current shot club, and a deviation evaluation value is generated. The deviation evaluation value is used as a training signal to update the weight parameters corresponding to the historical shot preference in the scenario decision model.

9. A club recognition device for a golf simulator, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the club recognition method for a golf simulator as described in any one of claims 1 to 8.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the golf simulator club recognition method as described in any one of claims 1 to 8.

Images