Smart backboard interaction method and device, smart backboard, medium and program

Abstract

The invention relates to the technical field of smart backboards, in particular to a smart backboard interaction method and device, a smart backboard, a medium and a program.The method comprises the steps that a basketball video collected in a current working mode is obtained; whether goals exist in the smart backboard or not is recognized, if the goals exist in the smart backboard, the basketball video is input into the trained target neural network model, and the target neural network model outputs goals personnel, goals actions and shooting positions; the goal type is determined according to the goal action and the shooting position corresponding to the goaler, and the smart backboard is controlled to generate corresponding visual and auditory feedback according to the goal type. Therefore, the problem that in the related technology, a basketball stand can only achieve the goal purpose and cannot meet user requirements, so that the user experience feeling is poor is solved.

Description

Technical Field

[0001] This application relates to the technical field of intelligent backboards, and particularly to an intelligent backboard interaction method, device, intelligent backboard, medium and program. Background Art

[0002] Basketball, as a globally popular sport, is deeply loved by people of all ages. With the continuous development of technology, some limitations have gradually emerged in traditional basketball training and competition methods. For example: Traditional training methods rely on the subjective judgment of coaches and lack accurate recording and analysis of data such as shooting actions and hit rates; traditional backboards have a single function and cannot provide real-time feedback, resulting in a relatively monotonous user experience; when multiple people train or compete simultaneously, it is difficult to accurately distinguish the basketball videos of each user, affecting the fairness of training and competition; existing basketball equipment mostly has a single function and lacks the comprehensive application of technologies such as multimedia, AI, and data collection, and cannot meet the modern users' needs for intelligence and interaction. Summary of the Invention

[0003] This application provides an intelligent backboard interaction method, device, intelligent backboard, medium and program to solve problems such as that a basketball hoop can only achieve the purpose of scoring a goal and cannot meet user needs, resulting in a poor user experience.

[0004] In the first aspect of the embodiments of this application, an intelligent backboard interaction method is provided, including the following steps: obtaining a basketball video collected in the current working mode, identifying whether there is a goal in the intelligent backboard, and if there is a goal in the intelligent backboard, inputting the basketball video into a trained target neural network model, where the target neural network model outputs the goal scorer, the goal action, and the shooting position. Among them, the target neural network includes a convolutional neural network and a recurrent neural network. The convolutional neural network is used to extract the image features of the shooting actions and positions of each member, and the recurrent neural network is used to process time series data to capture the dynamic features of the shooting actions; determining the goal type according to the goal action corresponding to the goal scorer and the shooting position, and controlling the intelligent backboard to generate corresponding visual and auditory feedback according to the goal type.

[0005] Optionally, the identifying whether there is a goal in the intelligent backboard includes: obtaining whether the first target sensor and the second target sensor at the target position of the intelligent backboard basketball hoop receive reflected signals, where the first target sensor is located above the second target sensor; if both the first target sensor and the second target sensor receive the reflected signals, identifying the actual duration from the first target sensor receiving the reflected signal to the second target sensor receiving the reflected signal; if the actual duration is lower than the preset duration, determining it as a valid goal, otherwise, determining it as no goal.

[0006] Optionally, identifying whether there is a scored goal on the intelligent backboard further includes: inputting the basketball video into a target algorithm, and the target algorithm outputs a goal result, where the target algorithm identifies the movement trajectory of the basketball in the basketball video and the spatial relative position relationship between the basketball and the intelligent backboard, and generates a goal result according to the movement trajectory and the spatial relative position relationship of the intelligent backboard.

[0007] Optionally, the processing method of the target neural network model: extracting the time series image data of the basketball video; marking the shooting actions and shooting positions corresponding to each shooter in each frame of image data to generate label data; calling the image data in the label data aligned with the time stamp of the sensor trigger signal; capturing the image data to determine the corresponding goal scorer, goal action and shooting position.

[0008] Optionally, controlling the intelligent backboard to display corresponding visual and auditory feedback according to the goal type includes: querying a first preset table according to the goal type to determine the visual feedback of the intelligent backboard, and simultaneously querying a second preset table to determine the auditory feedback of the intelligent backboard, where the first preset table is the mapping relationship between the goal type and the visual feedback, and the second preset table is the mapping relationship between the goal type and the auditory feedback.

[0009] Optionally, before identifying whether there is a scored goal on the intelligent backboard, it includes: establishing a connection channel between the client and the intelligent backboard; determining the working mode of the intelligent backboard according to the control intention of the client, and adjusting the setting parameters of the intelligent backboard, where the working mode includes a game mode, a training mode, an idle mode, and an interaction mode.

[0010] Optionally, after controlling the intelligent backboard to generate corresponding visual and auditory feedback according to the goal type, it further includes: automatically generating a highlight video of the daily goal videos and uploading it to the client.

[0011] An embodiment of the second aspect of the present application provides an intelligent backboard interaction device, including: an acquisition module, configured to acquire a basketball video collected in the current working mode; an identification module, configured to identify whether there is a scored goal on the intelligent backboard. If there is a scored goal on the intelligent backboard, input the basketball video into a trained target neural network model, and the target neural network model outputs the goal scorer, goal action and shooting position, where the target neural network includes a convolutional neural network and a recurrent neural network, the convolutional neural network is used to extract the image features of the shooting actions and positions of each member, and the recurrent neural network is used to process time series data to capture the dynamic features of the shooting actions; a control module, configured to determine the goal type according to the goal action corresponding to the goal scorer and the shooting position, and control the intelligent backboard to generate corresponding visual and auditory feedback according to the goal type.

[0012] In a third aspect embodiment of the present application, a smart backboard is provided, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the program to perform the smart backboard interaction method as described in the above embodiments.

[0013] In a fourth aspect embodiment of the present application, a computer-readable storage medium is provided, on which a computer program is stored, and the program is executed by a processor to perform the smart backboard interaction method as described in the above embodiments.

[0014] Therefore, the present application has at least the following beneficial effects: Embodiments of the present application can detect in real time whether a basketball passes through the basket by integrating intelligent sensors, ensuring the accuracy and reliability of goal recognition, obtaining the basketball video collected in the current working mode, combining with the target neural network model, accurately distinguishing the shooting actions and positions of each user, meeting the needs of multiple people training and competing simultaneously, extracting the image features of shooting actions and positions through a convolutional neural network, combining with a recurrent neural network to process time series data, capturing the dynamic features of shooting actions, realizing efficient and accurate basketball video analysis, improving the accuracy of shooting action analysis and training effect through continuous training and optimization of the target neural network model, generating corresponding visual and auditory feedback according to the goal type, and enhancing the user interaction experience and sense of participation.

[0015] Additional aspects and advantages of the present application will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The above and / or additional aspects and advantages of the present application will become apparent and be readily understood from the following description of the embodiments in conjunction with the drawings, where: Figure 1 is a flowchart of a smart backboard interaction method according to an embodiment of the present application; Figure 2 is a functional schematic diagram of a smart backboard according to an embodiment of the present application; Figure 3 is a schematic diagram of the game mode of a smart backboard according to an embodiment of the present application; Figure 4 is a schematic diagram of the training mode of a smart backboard according to an embodiment of the present application; Figure 5 is a schematic diagram of the 100-shot shooting training of a smart backboard according to an embodiment of the present application; Figure 6 is an example diagram of a smart backboard interaction device according to an embodiment of the present application; Figure 7Schematic diagram of the structure of the intelligent backboard provided according to the embodiments of the present application. Detailed implementation manners

[0017] The embodiments of the present application will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present application, but should not be construed as a limitation of the present application.

[0018] The intelligent backboard interaction method, device, intelligent backboard, storage medium and program according to the embodiments of the present application will be described below with reference to the accompanying drawings.

[0019] Specifically, Figure 1 Flow chart of the intelligent backboard interaction method provided by the embodiments of the present application.

[0020] As Figure 1 shown, the intelligent backboard interaction method includes the following steps: In step S101, a basketball video collected in the current working mode is obtained.

[0021] It can be understood that in the embodiments of the present application, by obtaining a basketball video collected in the current working mode, subsequent operations can be facilitated.

[0022] In the embodiments of the present application, before identifying whether there is a goal in the intelligent backboard, it includes: establishing a connection channel between the client and the intelligent backboard; determining the working mode of the intelligent backboard according to the control intention of the client, and adjusting the setting parameters of the intelligent backboard, where the working mode includes a game mode, a training mode, an idle mode and an interaction mode.

[0023] It can be understood that in the embodiments of the present application, a connection channel between the client and the intelligent backboard can be established; the working mode of the intelligent backboard can be determined according to the control intention of the client, and the setting parameters of the intelligent backboard can be adjusted. Through a convenient connection method and real-time communication, users can remotely control the backboard functions and settings through the client without directly operating the backboard, reducing the user operation difficulty and improving the use experience; they can select a suitable working mode according to their own needs to obtain a personalized training and game experience, meet the diverse needs of users, and improve the functional utilization rate of the intelligent backboard.

[0024] It should be noted that users can quickly establish a connection channel by scanning the QR code on the intelligent backboard with the mobile APP or applet on the mobile phone to realize remote control and data interaction; real-time data transmission between the client and the intelligent backboard is achieved through the Wi-Fi / Bluetooth module to ensure the immediacy of operation instructions and feedback information; multiple clients are supported to connect simultaneously to meet the needs of team training and games.

[0025] In step S102, it is identified whether there is a scored goal on the intelligent backboard. If there is a scored goal on the intelligent backboard, the basketball video is input into the trained target neural network model, and the target neural network model outputs the goal scorer, the goal-scoring action, and the shooting position. Among them, the target neural network includes a convolutional neural network and a recurrent neural network. The convolutional neural network is used to extract the image features of the shooting actions and positions of each member, and the recurrent neural network is used to process time series data to capture the dynamic features of the shooting actions.

[0026] Among them, the target duration can be set according to requirements. For example: 5s, without specific limitation.

[0027] It can be understood that in the embodiment of the present application, when there is a scored goal on the intelligent backboard, the goal scorer can be accurately identified according to the target neural network model, the shooting actions can be accurately classified, and the shooting position can be accurately judged, significantly improving the accuracy, stability, and user experience of the system.

[0028] It should be noted that the present application can determine whether there is a scored goal on the intelligent backboard in various ways to improve the accuracy of goal recognition.

[0029] As a possible implementation method, to identify whether there is a scored goal on the intelligent backboard, it includes: obtaining whether the first target sensor and the second target sensor at the target position of the intelligent backboard basket receive reflected signals; if both the first target sensor and the second target sensor receive reflected signals, identifying the actual duration from when the first target sensor receives the reflected signal to when the second target sensor receives the reflected signal; if the actual duration is lower than the preset duration, it is determined as a valid goal, otherwise, it is determined as not a goal.

[0030] Among them, the preset duration can be set according to the actual situation, without specific limitation.

[0031] It can be understood that the embodiment of the present application can ensure the accuracy and reliability of goal detection through multiple target sensors, avoid misjudgment caused by the basketball hitting the rim and not going in, and the collaborative work of multiple target sensors realizes redundant detection, further reducing the false detection rate; identifying the actual duration from when the first target sensor receives the reflected signal to when the second target sensor receives the reflected signal; determining whether it is a valid goal according to the actual duration, avoiding misjudgment caused by short-term optical path blockage, ensuring the accuracy of goal determination, meeting the real-time requirements of training and competitions, and improving the user experience.

[0032] It should be noted that multiple target sensors are installed under the basket to cover different detection areas. The target sensors of the present application are photoelectric sensors, and the first target sensor is located above the second target sensor.

[0033] It is determined whether there is an obstruction based on the reflection signal of the optoelectronic sensor. The optoelectronic sensor emits outward from the backboard. Under normal circumstances, infrared light is emitted. If there is an obstruction ahead, it will be reflected back, and the received reflection signal is determined to be an obstruction. At the same time, the distance where the obstruction occurs is also determined according to the intensity of the reflected signal.

[0034] Specifically, it is determined whether a goal is scored according to the order of receiving the reflection signals of two sensors. If only the signal of one of the sensors is received, it is not determined that a goal is scored; if the signal of the first (upper) sensor is received first, and then the signal of the second (lower) sensor is received within a preset time period, it is determined that a goal is scored; if the received order is reversed, it is not determined that a goal is scored either.

[0035] As another possible implementation method, it is determined whether a goal is scored on the intelligent backboard according to the basketball video, and it further includes: inputting the basketball video into a target algorithm, and the target algorithm outputs a goal result. Among them, the target algorithm identifies the movement trajectory of the basketball in the basketball video and the spatial relative position relationship between the basketball and the intelligent backboard, and generates a goal result according to the movement trajectory and the spatial relative position relationship of the intelligent backboard.

[0036] It can be understood that the embodiments of the present application can input the basketball video into the target algorithm, and the target algorithm outputs a goal result. Among them, the target algorithm identifies the movement trajectory of the basketball in the basketball video and the spatial relative position relationship between the basketball and the intelligent backboard, and generates a goal result according to the movement trajectory and the spatial relative position relationship, ensuring the accuracy of goal determination, meeting the real-time requirements of training and competitions, and improving the user experience.

[0037] It should be noted that based on the AI algorithm, the basketball detection frames in adjacent frames are matched to generate a continuous movement trajectory, and it is predicted whether the landing point is the center of the basket; the LSTM network is used to learn the historical trajectory pattern to correct the prediction deviation caused by occlusion or noise; the horizontal / vertical speed and shooting angle of the basketball are calculated in real time (for example, 45° is the optimal shooting angle) to determine whether it meets the effective goal conditions; if the basketball trajectory mutates (such as being slapped by a player), it is marked as "non-shooting action" to exclude interference.

[0038] Taking the center of the basket as the origin, a three-dimensional space coordinate system is defined. The real-time distance between the basketball and the basket is calculated through the binocular camera parallax. The optoelectronic sensor (detecting the basketball passing through the basket plane) and the IMU data (the rotation state of the basketball) are integrated to assist in positioning and calibration; the center of mass of the basketball needs to completely enter the virtual cylinder centered on the basket. If the basketball trajectory continuously passes through the cylinder and does not bounce out of the basket within 5 consecutive frames (≥83ms), it is determined that a goal is scored. If the basketball only touches the edge of the basket or the backboard, it is marked as "not scored".

[0039] In the embodiments of the present application, the structure of the target neural network model includes: a convolutional neural network, a recurrent neural network, feature fusion, and an output layer.

[0040] The convolutional neural network is used to extract image features of the shooting action and position, and capture static information. The structure of the convolutional neural network: Input layer: Receives preprocessed time-series image data (such as video frames in the 3 seconds before scoring). Convolutional layer: Uses multiple convolutional kernels (such as 3x3, 5x5) to extract image features, and each convolutional layer is followed by a ReLU activation function. Pooling layer: Uses a max pooling layer to reduce the dimension of the feature map and reduce the computational amount. Fully connected layer: Flattens the feature vectors extracted by the convolutional layer and inputs them into the fully connected layer. Output layer: Outputs the feature vectors of the shooting action and position.

[0041] The recurrent neural network is used to process time-series data and capture the dynamic features of the shooting action. The structure of the recurrent neural network: Input layer: Receives the time-series feature vectors extracted by the CNN. LSTM unit: Uses a long short-term memory network (LSTM) to process time-series data and capture the dynamic changes of the shooting action. Bidirectional RNN: Uses a bidirectional RNN (Bi-RNN) to capture the forward and backward dependencies of time-series data. Output layer: Outputs the dynamic feature vectors of the shooting action.

[0042] The feature fusion and output layer is used to splice the static features extracted by the CNN and the dynamic features extracted by the RNN to generate a comprehensive feature vector, and maps the comprehensive feature vector to the output category through a fully connected layer.

[0043] Output: Scorer: Classification result based on user identity information (such as facial recognition or wearable device data); Shooting action: Probability distribution of action categories such as jump shot, dunk, three-pointer, etc.; Shooting position: Probability distribution of position categories such as under the basket, mid-range, beyond the three-point line, etc.

[0044] In the embodiments of the present application, the processing method of the target neural network model: Extracts the time-series image data of the basketball video; Marks the shooting action and shooting position corresponding to each shooter in each frame of image data to generate label data; Invokes the image data in the label data aligned with the time stamp of the sensor trigger signal; Captures the image data to determine the corresponding scorer, scoring action, and shooting position.

[0045] It can be understood that the embodiments of the present application can extract the time-series image data of the basketball video, generate label data, invoke the aligned image data, and capture the image data. This technical solution realizes accurate scorer recognition, shooting action classification, and shooting position judgment, and significantly improves the accuracy, stability, and user experience of the system.

[0046] It should be noted that the training process of the target neural network model: (1) Dataset construction Data sources: Video data of multiple people shooting hoops captured by intelligent cameras; the goal time points triggered by transmissive optoelectronic sensors; user identity information (such as RFID bracelet or facial recognition data).

[0047] Data annotation: Perform frame-by-frame annotation on the video data, marking the shooting personnel, shooting actions, and shooting positions.

[0048] Align the time stamps of the sensor trigger signals to ensure data consistency.

[0049] (2) Model training Training method: Use the annotated dataset for supervised learning, adopt the cross-entropy loss function to optimize the model parameters; adopt transfer learning technology and fine-tune based on pre-trained vision models (such as ResNet, EfficientNet); improve the model generalization ability through data augmentation techniques (such as rotation, scaling, brightness adjustment).

[0050] Optimizer: Use the Adam optimizer to update the parameters, with an initial learning rate of 0.001; adopt a learning rate decay strategy, and halve the learning rate every 10 epochs.

[0051] Training strategy: Use early stopping to prevent overfitting, and stop training when the validation set loss does not decrease for 5 consecutive times.

[0052] (3) Model evaluation and optimization Evaluation metrics: Accuracy: The classification accuracy of the model for the goal scorer, shooting action, and shooting position; Recall: The detection ability of the model for various actions and positions; F1 score: Comprehensively measure the precision and recall of the model.

[0053] Optimization method: Continuously optimize the model parameters through user feedback data (such as misclassification records). Adopt online learning technology to update the model in real time to adapt to new users and new actions.

[0054] Process the data through the target neural network model after the above training. The specific process is as follows: (1) Data input Extract frames and perform time series annotation on the video segment within the target duration before the goal captured by the intelligent camera to generate a basketball video. Input the basketball video into the target neural network model for feature extraction and classification.

[0055] (2) Feature extraction Use CNN to extract the image features of the shooting action and position. Use RNN / LSTM to process time series data and capture the dynamic features of the shooting action.

[0056] (3) Feature fusion and classification Concatenate the features extracted by CNN and RNN to generate a comprehensive feature vector. Map the comprehensive feature vector to the output categories through a fully connected layer, and output the probability distributions of the scoring player, shooting action, and shooting position.

[0057] (4) Post-processing: Perform threshold judgment on the probability distribution output by the model to determine the final classification result. Combine the sensor trigger signal to verify the reliability of the model output.

[0058] In step S103, determine the goal type according to the goal action and shooting position corresponding to the scoring player, and control the smart backboard to generate corresponding visual and auditory feedback according to the goal type.

[0059] It can be understood that the embodiments of the present application can determine the goal type according to the goal action and shooting position corresponding to the scoring player, and control the smart backboard to generate corresponding visual and auditory feedback according to the goal type, thereby enhancing the user interaction experience and sense of participation.

[0060] In the embodiments of the present application, controlling the smart backboard to display corresponding visual and auditory feedback according to the goal type includes: querying the first preset table according to the goal type to determine the visual feedback of the smart backboard, and at the same time querying the second preset table to determine the auditory feedback of the smart backboard, where the first preset table is the mapping relationship between the goal type and the visual feedback, and the second preset table is the mapping relationship between the goal type and the auditory feedback.

[0061] Among them, both the first preset table and the second preset table can be set according to actual needs, and no specific limitations are made.

[0062] It can be understood that the embodiments of the present application can control the smart backboard to display corresponding visual and auditory feedback according to the goal type, and this technical solution realizes real-time, accurate, and personalized multi-sensory feedback, significantly enhancing the user experience, interactivity, and training effect.

[0063] It should be noted that the visual feedback can include display screen animations: playing different dynamic animations according to the goal type (such as displaying a fireworks animation when a three-pointer is made), LED light strip effects triggering different light colors and flashing frequencies according to the goal type (such as the lights flashing red when dunking), data display: real-time display of basketball videos (such as shooting percentage, score) on the display screen, etc.

[0064] Auditory feedback can include sound effects playback: playing different sound effects according to the type of goal (such as playing cheers when a three-pointer is made); voice prompts: playing voice prompts according to the type of goal (such as "Three-pointer made").

[0065] For example, assuming that the types of goals include layups, mid-range shots, three-pointers, dunks, free throws, etc., and the visual feedback includes animation types, light colors, blinking frequencies, data display layouts, etc., the first preset table can be as shown in Table 1 below, where Table 1 is a mapping relationship table between goal types and visual feedback.

[0066]

[0067] Assuming that the types of goals include layups, mid-range shots, three-pointers, dunks, free throws, etc., and the auditory feedback includes sound effect types, voice prompt contents, etc., the second preset table can be as shown in Table 2 below, where Table 2 is a mapping relationship table between goal types and auditory feedback.

[0068]

[0069] In the embodiment of the present application, after controlling the intelligent backboard to generate corresponding visual and auditory feedback according to the type of goal, it further includes: automatically generating a highlight video of the daily goal videos and uploading it to the client.

[0070] It can be understood that the embodiment of the present application can automatically generate a highlight video of the daily goal videos and upload it to the client. This technical solution realizes automated video generation, real-time uploading, user interaction, and data-driven optimization, significantly improving the user experience, interactivity, and commercial value of the device.

[0071] According to the intelligent backboard interaction method proposed in the embodiment of the present application, by integrating intelligent sensors to detect in real time whether the basketball passes through the basket, ensuring the accuracy and reliability of goal recognition, obtaining the on-court multi-person basketball videos within the target duration before the goal, combining with the target neural network model, accurately distinguishing the shooting actions and positions of each user, meeting the needs of multi-person simultaneous training and competitions, extracting the image features of the shooting actions and positions through a convolutional neural network, combining with a recurrent neural network to process time series data, capturing the dynamic features of the shooting actions, realizing efficient and accurate basketball video analysis, through continuous training and optimization of the target neural network model, improving the accuracy of shooting action analysis and training effects, generating corresponding visual and auditory feedback according to the type of goal, and enhancing the user interaction experience and sense of participation.

[0072] The following will be combined with Figures 2 - 3 The intelligent backboard interaction method of the present application will be elaborated in detail as follows: Since the intelligent backboard includes but is not limited to functions such as game mode, training mode, video highlights, idle mode, system settings, etc., therefore, it will be elaborated in detail in combination with Figure 2 each function as follows: (1) Game mode (as shown in Figure 3 ) Suppose the game mode is set to a three-minute challenge, then the corresponding parameter function description is: shooting competition mode, timed for three minutes, and the number of shots made by both sides is counted to distinguish the winner.

[0073] There is a teaching video before the game starts to guide users on matters to note when participating in the game.

[0074] During the game, the intelligent backboard displays information such as frame lines, scores, and timekeeping, and automatically counts the shooting results of the participants.

[0075] When a participant makes a wonderful goal, give a flashing light strip, voice prompt and animation effect.

[0076] After the game ends, give the score and data analysis according to the results of the participants. The system will record the entire game process and automatically capture the wonderful videos during the training and present them to both sides of the game.

[0077] (2) Training mode (as shown in Figure 4 ) Suppose the training goal of the training mode: 5-minute timed shooting, function description: conduct shooting training with 5 minutes as a timing unit.

[0078] Before the training officially starts, provide a shooting teaching video to introduce the correct shooting action.

[0079] During the training, the screen of the intelligent backboard displays information such as frame lines, scores, and timekeeping, and automatically counts the shooting results of the shooter within 5 minutes.

[0080] When the shooter makes a wonderful goal, give a flashing light strip, voice prompt and animation effect.

[0081] After the training is completed, the system gives data analysis and dynamic ranking according to the shooter's results.

[0082] The system records the entire training process and can capture the wonderful videos during the training to assist the shooter in replaying and observing.

[0083] (3)100-shot shooting training (as shown in Figure 5 ) Goal: Complete 100-shot shooting training.

[0084] Function description: Conduct training with the goal of 100-shot shooting.

[0085] Before the formal training starts, a shooting teaching video is provided to introduce the correct shooting action.

[0086] During the training process, the screen display frame lines, scores, timing and other information of the intelligent backboard are shown, and the time taken for the shooter to make 100 shots is automatically counted.

[0087] When the shooter makes a wonderful goal, there will be flashing light strips, voice prompts and animation effects.

[0088] After the training is completed, the system gives data analysis and dynamic ranking according to the shooter's performance.

[0089] The system records the entire training process and can capture wonderful videos during the training to assist the shooter in replaying and observing.

[0090] (4)Idle mode In the idle mode, preset advertising videos and pictures can be played according to the idle time of the venue.

[0091] (5)Video highlights The system automatically intercepts the wonderful videos of shooting in the game mode and training mode, stores them locally, and provides a playlist for playing at any time.

[0092] (6)System settings The main functions of the system settings are: theme settings and backboard settings. Among them, the theme settings are for personalized settings of the entire intelligent backboard system. The system provides 2-3 sets of theme schemes for the management personnel to choose to meet the theme requirements of different venues and different atmospheres. The backboard settings are for personalized settings of the color and frame lines of the intelligent backboard. The system provides 2-3 sets of setting schemes for the management personnel to choose to meet the theme requirements of different venues and different atmospheres.

[0093] In summary, through the intelligent, interactive and personalized design, this application significantly improves the user experience, interactivity and commercial value of the device, and has broad application prospects and market potential.

[0094] Secondly, the intelligent backboard interaction device according to the embodiment of the present application is described with reference to the accompanying drawings.

[0095] Figure 6 It is a block diagram of the intelligent backboard interaction device according to the embodiment of the present application.

[0096] As Figure 6 shown, the intelligent backboard interaction device 10 includes: an identification module 100, an acquisition module 200, a processing module 300 and a control module 400.

[0097] Among them, the recognition module 100 is used to recognize whether there is a goal in the intelligent backboard; the acquisition module 200 is used to obtain the on-court multi-person basketball video of the target duration before the goal if there is a goal in the intelligent backboard; the processing module 300 is used to input the basketball video into the trained target neural network model, and the target neural network model outputs the goal scorer, the goal action, and the shooting position. Among them, the target neural network includes a convolutional neural network and a recurrent neural network. The convolutional neural network is used to extract the image features of the shooting actions and positions of each member, and the recurrent neural network is used to process time series data to capture the dynamic features of the shooting actions; the control module 400 is used to determine the goal type according to the goal action and shooting position corresponding to the goal scorer, and control the intelligent backboard to generate corresponding visual and auditory feedback according to the goal type.

[0098] Specifically, the hardware configuration of the intelligent backboard of the present application is as follows: 1. Display screen Model and specifications: BOE 49-inch 2K display screen. Features: Shockproof: High-strength materials are used to ensure that it is not easily damaged under the impact of a basketball or other external forces. High brightness: Ensure clear display of content in indoor and outdoor environments and adapt to different lighting conditions. Functions: Used for video playback, advertisement display, interactive animation display, and data display, providing a user interaction interface.

[0099] 2. Motherboard Model and specifications: RK3588 motherboard, running the Android system. Configuration: Memory: 16G memory to ensure the smoothness of multitasking. Storage: 256G hard disk, supporting a large amount of data storage and fast read and write. Functions: Run the operating system and application programs of the intelligent backboard, support AI model inference and data processing. Provide hardware interfaces to connect other hardware components (such as cameras, sensors, LED light strips, etc.).

[0100] 3. Storage hard disk Model and specifications: 480G SSD. Features: High-speed read and write: Support fast data storage and reading to ensure real-time system response. Large capacity: Meet the storage needs of a large amount of videos, data, and application programs.

[0101] Function: Used to store video clips, user data, application programs, and system files.

[0102] 4. Network module Model and specifications: Support Wi-Fi, 5G, and Ethernet connections. Features: High-speed transmission: Support high-speed data transmission to ensure the real-time transmission of video streams and data. Multi-mode connection: Support Wi-Fi, 5G, and wired networks to adapt to different network environments. Functions: Realize real-time data transmission, cloud synchronization, and remote control.

[0103] 5. Camera Model and Specifications: 4K resolution, 180-degree viewing angle, with anti-shake function. Features: High resolution: 4K resolution ensures clear video shooting and supports precise motion analysis. Wide viewing angle: 180-degree viewing angle covers the entire basket area, ensuring no blind spots for capturing shooting actions. Anti-shake function: Ensures the stability of video shooting and avoids blurred images caused by vibration or movement.

[0104] Function: Shoot the shooting action in real time, record video data for motion analysis and sharing.

[0105] 6. LED Light Strip Model and Specifications: RGB light strip, surrounding the basketball hoop. Features: Multi-color display: Supports multiple color combinations to enhance the visual effect. Programmable control: Supports users to customize the light color and flashing frequency to meet personalized needs. Function: Trigger cool lighting effects according to the type of goal or user actions, enhancing the interactive experience.

[0106] 7. Plastic Handboard Function: Used to fix the backboard and camera, ensuring the stability and safety of the device. Features: High-strength material: Ensures that it is not easily damaged under the impact of a basketball or other external forces. Modular design: Facilitates installation and maintenance, reducing maintenance costs.

[0107] 8. Multimedia System Model and Specifications: In-vehicle power amplifier and audio system. Features: High sound quality: Ensures clear sound effects and moderate volume, enhancing the auditory feedback. Multi-functional support: Supports sound effect playback and voice prompts to meet different scenario requirements. Function: Plays sound effects and voice prompts, enhancing the user's auditory interactive experience.

[0108] 9. Cable Accessories Composition: Power cord, CAT5 network cable, peripheral interface, cooling fan, internal cable, safety switch, power adapter board, cable label, chassis fan, etc. Function: Power management: Ensures stable power supply for the device, supporting 12V and 24V power inputs. Heat dissipation design: Ensures the stability of the device during long-term operation through the layout of the cooling fan and internal cables. Safety protection: Prevents device damage caused by overload or short circuit through the safety switch and power adapter board.

[0109] 10. Hardware Components Function: Used for the structural support and installation of the entire backboard, ensuring the stability and durability of the device. Features: High-strength material: Ensures the stability of the device under the impact of a basketball or other external forces. Modular design: Facilitates installation and maintenance, reducing maintenance costs.

[0110] Sensor Model and Specification: Photoelectric Sensor. Function: Detect whether a basketball passes through the basket through the principle of optical path interruption and trigger a goal event. Characteristics: High sensitivity: Ensure accurate detection of the moment when the basketball passes through the basket. Environmental adaptability: Adopt infrared light source and filter technology to effectively avoid environmental light interference.

[0111] 12. Power Switch Model and Specification: 12V Power Supply - 350W + 24V Power Supply 120W. Function: Dual power input: Support 12V and 24V power input to ensure the stable operation of the device under different voltage conditions. High-efficiency power supply: Ensure stable power supply of the device under high load conditions to meet the long-term operation requirements.

[0112] 13. Remote Control Model and Specification: Infrared / Bluetooth Remote Control. Function: Remote control: Support users to remotely control the functions and settings of the intelligent backboard through the remote control; Multi-mode connection: Support infrared and Bluetooth connections to adapt to different usage scenarios.

[0113] It should be noted that the foregoing explanations of the embodiments of the intelligent backboard interaction method also apply to the intelligent backboard interaction device of this embodiment, and will not be repeated here.

[0114] According to the intelligent backboard interaction device proposed in the embodiments of the present application, by integrating intelligent sensors, it can detect in real time whether a basketball passes through the basket, ensure the accuracy and reliability of goal recognition, obtain the on-court multi-person basketball video within the target duration before the goal, combine with the target neural network model, accurately distinguish the shooting actions and positions of each user, meet the needs of multi-person simultaneous training and competitions, extract the image features of shooting actions and positions through convolutional neural networks, combine with recurrent neural networks to process time series data, capture the dynamic features of shooting actions, achieve efficient and accurate basketball video analysis, through continuous training and optimization of the target neural network model, improve the accuracy of shooting action analysis and training effect, generate corresponding visual and auditory feedback according to the goal type, and enhance the user interaction experience and sense of participation.

[0115] FIG. 7 is a schematic structural diagram of the intelligent backboard provided by the embodiments of the present application. The intelligent backboard may include: A memory 701, a processor 702, and a computer program stored on the memory 701 and executable on the processor 702.

[0116] When the processor 702 executes the program, it implements the intelligent backboard interaction method provided in the above embodiments.

[0117] Further, the intelligent backboard further includes: A communication interface 703 for communication between the memory 701 and the processor 702.

[0118] A memory 701 for storing a computer program that can run on a processor 702.

[0119] The memory 701 may include a high-speed RAM memory and may also include non-volatile memory, such as at least one disk memory.

[0120] If the memory 701, the processor 702, and the communication interface 703 are implemented independently, the communication interface 703, the memory 701, and the processor 702 can be interconnected through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of representation, Figure 7 only a thick line is shown in the figure, but it does not mean that there is only one bus or one type of bus.

[0121] Optionally, in a specific implementation, if the memory 701, the processor 702, and the communication interface 703 are integrated on a chip, the memory 701, the processor 702, and the communication interface 703 can communicate with each other through an internal interface.

[0122] The processor 702 may be a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0123] The embodiments of the present application also provide a computer-readable storage medium, on which a computer program or instruction is stored. When the computer program or instruction is executed by a processor, the intelligent backboard interaction method as described above is implemented.

[0124] The embodiments of the present application also provide a computer program product, including a computer program or instruction. When the computer program or instruction is executed, the intelligent backboard interaction method as described above is implemented.

[0125] In the description of this specification, the descriptions with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of this application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0126] In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In the description of this application, the meaning of "N" is at least two, such as two, three, etc., unless otherwise specifically defined.

[0127] Any process or method description shown in a flowchart or described in other ways herein can be understood to represent a module, segment, or portion of code including one or N executable instructions for implementing a customized logical function or process, and the scope of the preferred embodiments of this application includes additional implementations, where the functions can be executed in a substantially simultaneous manner or in a reverse order according to the involved functions, rather than in the order shown or discussed, which should be understood by those skilled in the art to which the embodiments of this application belong.

[0128] It should be understood that each part of this application can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any combination of one or more of the following techniques well known in the art: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0129] Those of ordinary skill in the art of this technology can understand that all or part of the steps carried by the method of implementing the above embodiments can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

Claims

1. A smart backboard interaction method, characterized in that, Including the following steps: Obtain the basketball video collected in the current working mode; Identify whether there is a goal on the intelligent backboard. If there is a goal on the intelligent backboard, input the basketball video into the trained target neural network model. The target neural network model outputs the goal scorer, the goal action, and the shooting position. Among them, the target neural network includes a convolutional neural network and a recurrent neural network. The convolutional neural network is used to extract the image features of the shooting actions and positions of each member, and the recurrent neural network is used to process time series data to capture the dynamic features of the shooting actions; Determine the goal type according to the goal action corresponding to the goal scorer and the shooting position, and control the intelligent backboard to generate corresponding visual and auditory feedback according to the goal type.

2. The intelligent backboard interaction method according to claim 1, wherein, The identification of whether there is a goal on the intelligent backboard includes: Obtain whether the first target sensor and the second target sensor at the target position of the intelligent backboard basket receive reflected signals, where the first target sensor is located above the second target sensor; If both the first target sensor and the second target sensor receive the reflected signal, identify the actual duration from the first target sensor receiving the reflected signal to the second target sensor receiving the reflected signal; If the actual duration is lower than the preset duration, it is determined as a valid goal, otherwise, it is determined as no goal.

3. The intelligent backboard interaction method according to claim 1, wherein The identification of whether there is a goal on the intelligent backboard further includes: Input the basketball video into the target algorithm, and the target algorithm outputs the goal result. The target algorithm identifies the movement trajectory of the basketball in the basketball video and the spatial relative position relationship between the basketball and the intelligent backboard, and generates the goal result according to the movement trajectory and the spatial relative position relationship of the intelligent backboard.

4. The intelligent backboard interaction method according to claim 1, wherein The processing method of the target neural network model: Extract the time series image data of the basketball video; Mark the shooting actions and shooting positions corresponding to each shooting person in each frame of image data to generate label data; Call the image data in the label data aligned with the time stamp of the sensor trigger signal; Capture the image data to determine the corresponding goal scorer, goal action, and shooting position.

5. The intelligent backboard interaction method according to claim 1, wherein, Controlling the intelligent backboard to display corresponding visual and auditory feedback according to the goal type includes: Query the first preset table according to the goal type to determine the visual feedback of the intelligent backboard, and at the same time query the second preset table to determine the auditory feedback of the intelligent backboard. Among them, the first preset table is the mapping relationship between the goal type and the visual feedback, and the second preset table is the mapping relationship between the goal type and the auditory feedback.

6. The intelligent backboard interaction method according to claim 1, wherein, Before identifying whether there is a goal on the intelligent backboard, it includes: Establish a connection channel between the client and the intelligent backboard; Determine the working mode of the intelligent backboard according to the control intention of the client and adjust the setting parameters of the intelligent backboard. The working modes include the game mode, the training mode, the idle mode, and the interaction mode.

7. The intelligent backboard interaction method according to claim 1, characterized in that After controlling the intelligent backboard to generate corresponding visual and auditory feedback according to the goal type, it further includes: Automatically generate a highlight video of the daily goal videos and upload it to the client.

8. An intelligent backboard interaction device, characterized in that, Including: An acquisition module for obtaining the basketball video collected in the current working mode; An identification module, configured to identify whether there is a scored goal on the intelligent backboard. If there is a scored goal on the intelligent backboard, the basketball video is input into the trained target neural network model, and the target neural network model outputs the goal scorer, the goal-scoring action, and the shooting position. Among them, the target neural network includes a convolutional neural network and a recurrent neural network. The convolutional neural network is used to extract the image features of the shooting actions and positions of each member, and the recurrent neural network is used to process time series data to capture the dynamic features of the shooting actions; A control module, configured to determine the goal type according to the goal-scoring action corresponding to the goal scorer and the shooting position, and control the intelligent backboard to generate corresponding visual and auditory feedback according to the goal type.

9. A smart backboard, characterized in that, Comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor. The processor executes the program to implement the intelligent backboard interaction method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to be used for implementing the intelligent backboard interaction method according to any one of claims 1-7.

Images