Table tennis entertainment intelligent scoring system and method based on cloud collaborative data

Abstract

The invention relates to the technical field of sports intelligent scoring, in particular to a table tennis entertainment intelligent scoring system and method based on cloud collaborative data, and the system comprises the steps: accurately capturing a real-time three-axis acceleration data flow generated along with the vibration of a table tennis table top through an inertial sensor arranged in a first mobile terminal; the method comprises the following steps of: performing coordinate system decoupling and energy sequence extraction by using a preset signal preprocessing algorithm, and further inputting the sequence into a preset action recognition model for deep feature matching, thereby generating first candidate score event data containing a high-precision timestamp and vibration intensity confidence when a preset score slapping action threshold is met. The technical bottleneck that effective score slapping and global interference vibration are difficult to distinguish in a traditional single-point sensing scheme is effectively solved, the accuracy and the anti-interference capability of automatic scoring are greatly improved, and the misjudgment rate is remarkably reduced; and the preciseness and reliability of the whole scoring process are ensured.

Description

Technical Field

[0001] This invention belongs to the field of sports intelligent scoring technology, and relates to a table tennis entertainment-oriented intelligent scoring system and method based on cloud-based collaborative data. Background Technology

[0002] Table tennis, a globally popular indoor sport, combines competitive, recreational, and fitness benefits. Long-term, systematic participation can significantly improve reaction speed, coordination, and cardiopulmonary function, playing a vital role in promoting physical and mental health and enriching social leisure life. Meanwhile, smartphones, as highly prevalent mobile intelligent terminals in today's society, with their built-in high-precision inertial sensors (such as accelerometers and gyroscopes) and powerful edge computing capabilities, provide the hardware foundation for real-time perception of human movements and recognition of physical events, becoming an important platform for achieving lightweight, universal human-computer interaction and motion sensing.

[0003] However, existing intelligent scoring technologies for table tennis still have significant limitations. Current technologies are mostly limited to the automated recording of scores, with simplistic interaction methods and a lack of design that deeply integrates the scoring process with an entertainment experience. They fail to effectively utilize the audio, visual, and network capabilities of smartphones to enhance real-time feedback and immersion during matches, nor do they promote interaction among enthusiasts and event organization through lightweight social functions. This reflects that current technology remains at a rudimentary stage of "replacing manual scoring," rather than focusing on building an entertainment-oriented sports service ecosystem that integrates automatic sensing, engaging interaction, and social sharing. Furthermore, regarding sensing reliability, single devices are susceptible to environmental vibrations or non-scoring actions, and the lack of convenient manual intervention and error correction mechanisms affects the trust and flexibility of actual use. Overall, existing solutions have not achieved an effective balance between system cost, user experience, and functional completeness, hindering the widespread adoption of intelligent and entertainment-oriented development in table tennis. Summary of the Invention

[0004] In view of the problems existing in the prior art, the present invention provides a smart scoring system and method for table tennis entertainment based on cloud collaborative data. It utilizes the built-in inertial sensors of mobile terminals deployed on both sides of the table to collect the unique vibration signals generated by the player hitting the table surface after scoring, as the core scoring criterion. Automatic scoring is achieved through dual-end collaborative analysis, thereby solving the above-mentioned technical problems.

[0005] To achieve the above and other objectives, the technical solution adopted by the present invention is as follows:

[0006] The first aspect of this invention provides a smart scoring method for table tennis entertainment based on cloud-based collaborative data, the method comprising the following steps:

[0007] The real-time acceleration data stream is received from the inertial sensor built into the first mobile terminal, wherein the real-time acceleration data stream includes the triaxial acceleration components generated by the first mobile terminal as the ping-pong table vibrates.

[0008] The real-time acceleration data stream is denoised and the coordinate system is transformed using a preset signal preprocessing algorithm to extract the vertical vibration energy sequence perpendicular to the table surface.

[0009] The vertical vibration energy sequence is input into a preset motion recognition model for feature matching. When the matched vibration waveform features meet the preset scoring slapping action threshold, the first candidate scoring event data is generated. The first candidate scoring event data includes the local trigger time point and vibration intensity confidence.

[0010] The system sends a status verification request containing the first candidate score event data to the second mobile terminal on the other side through the established data exchange channel, and receives the time window status feedback data returned by the second mobile terminal.

[0011] Based on the time window status feedback data, collaborative conflict detection is performed on the first candidate score event data. When it is confirmed that the second mobile terminal has not generated a high confidence interference event within the preset time domain window corresponding to the local trigger time point, the final valid score instruction is generated.

[0012] Update the current game score data according to the final valid score instruction, and call the corresponding multimedia sound effect resource data for playback output according to the currently loaded entertainment mode configuration parameters.

[0013] A second aspect of this invention provides a cloud-based collaborative data-driven intelligent scoring system for table tennis entertainment, comprising:

[0014] An acceleration data acquisition module is used to receive real-time acceleration data streams from the inertial sensor built into the first mobile terminal, wherein the real-time acceleration data streams include triaxial acceleration components generated by the first mobile terminal as the ping-pong table vibrates.

[0015] The vibration signal preprocessing module is used to denoise and transform the real-time acceleration data stream using a preset signal preprocessing algorithm, and extract the vertical vibration energy sequence perpendicular to the table surface.

[0016] The slapping action recognition module is used to input the vertical vibration energy sequence into the preset action recognition model for feature matching. When the matched vibration waveform features meet the preset slapping action threshold, the first candidate score event data is generated. The first candidate score event data includes the local trigger time point and vibration intensity confidence.

[0017] The multi-terminal collaborative verification module is used to send a status verification request containing first candidate score event data to the second mobile terminal on the other side through the established data exchange channel, and to receive time window status feedback data returned by the second mobile terminal.

[0018] The scoring event arbitration module is used to perform collaborative conflict detection on the first candidate scoring event data based on the time window status feedback data. When it is confirmed that the second mobile terminal has not generated a high confidence interference event within the preset time domain window corresponding to the local trigger time point, the final valid scoring instruction is generated.

[0019] The game score update and response module is used to update the current game score data according to the final valid score instruction, and to call the corresponding multimedia sound effect resource data for playback output according to the currently loaded entertainment mode configuration parameters.

[0020] A third aspect of the present invention provides a smart scoring device for table tennis entertainment based on cloud-based collaborative data, including a processor, a memory, and a communication bus;

[0021] The memory stores a computer-readable program that can be executed by the processor;

[0022] The communication bus enables communication between the processor and the memory;

[0023] When the processor executes the computer-readable program, it performs steps in the intelligent scoring method for table tennis entertainment based on cloud-based collaborative data as described in any one of the present invention.

[0024] As described above, the intelligent scoring system and method for table tennis entertainment based on cloud-based collaborative data provided by the present invention has at least the following beneficial effects:

[0025] This invention provides a cloud-based collaborative data-driven intelligent scoring system and method for table tennis entertainment. It accurately captures real-time three-axis acceleration data streams generated by the vibration of the table tennis table using an inertial sensor built into a first mobile terminal. A preset signal preprocessing algorithm is then used for coordinate system decoupling and energy sequence extraction. This sequence is input into a preset action recognition model for deep feature matching. When a preset scoring action threshold is met, a first candidate scoring event data containing a high-precision timestamp and vibration intensity confidence level is generated. Based on this, a status verification request is sent to a second mobile terminal via a data exchange channel. The returned time-window status feedback data is compared and analyzed, effectively solving the technical bottleneck of traditional single-point sensing schemes that struggle to distinguish between valid scoring shots and global interference vibrations (such as ball impacts and environmental noise). On one hand, data verification from the novel perspective of the differences in vibration distribution in physical space greatly improves the accuracy and anti-interference capability of automated scoring, significantly reducing the misjudgment rate. On the other hand, generating a final valid scoring command to drive score updates ensures the rigor and reliability of the entire scoring process, providing a solid data foundation for subsequent entertainment-oriented interactions. Attached Figure Description

[0026] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 This is a schematic diagram showing the connections between the steps of the method of the present invention.

[0028] Figure 2 This is a schematic diagram of the logical connection for deploying the self-test step in this invention.

[0029] Figure 3 This is a schematic diagram showing the connections of the various modules in the system of the present invention.

[0030] Figure 4 This is a schematic diagram of the device structure connection of the present invention. Detailed Implementation

[0031] The following description, in conjunction with the implementation of this invention, is merely an example and illustration of the concept of this invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the inventive concept or exceed the scope defined in these claims, all of which should fall within the protection scope of this invention.

[0032] In existing technologies, automated scoring solutions for table tennis matches largely rely on dedicated sensor arrays deployed on the table or net, independent central processing units, and display devices. While these solutions can automatically record scores, their core hardware is highly customized, resulting in high overall system costs and complex installation and debugging, making them difficult to widely adopt among amateur enthusiasts, for home entertainment, or in temporary venues. Furthermore, the functional design of existing solutions focuses primarily on the basic need to "replace manual scoring," with interaction limited to numerical changes, resulting in a monotonous and repetitive process. They lack mechanisms to enhance the real-time enjoyment of the match through multimedia feedback and fail to effectively build social interaction scenarios connecting online and offline activities, thus failing to meet users' growing demands for interactivity and social engagement in sports and entertainment activities.

[0033] To address these issues, research revealed that the action of an athlete striking their own side of the table after scoring generates a vibration signal with specific time and frequency domain characteristics, distinct from regular shots, player movement, or other environmental noise. When a smartphone is fixed to the side of the table in a specific manner, its built-in inertial sensor can effectively capture this signal. However, the detection of a single device is highly susceptible to interference from other simultaneous strong vibrations (such as powerful shots or external impacts), leading to false triggers. This reveals that distinguishing between "scoring shots" and "game interference" through multi-node collaborative sensing is crucial for achieving reliable automatic scoring. Furthermore, the inventors recognized that linking accurate scoring results with personalized, gamified multimedia feedback (such as fighting game sound effects) in real time can significantly enhance participants' immersion and entertainment experience, while a convenient human error correction mechanism is essential for balancing automation and user experience reliability.

[0034] Based on the above findings, this solution constructs an intelligent scoring system that collaboratively operates between mobile sensing terminals and a cloud service platform. Its core lies in the following: smartphones deployed on both sides of the table first continuously collect and process vibration signals perpendicular to the table surface using built-in high-precision accelerometers. A pre-set motion recognition model is used to initially determine scoring impact events, generating candidate events with timestamps and confidence levels. Subsequently, the system immediately performs collaborative verification with the opposite phone via a local wireless channel: if one end detects a high-confidence impact feature, and the other end does not report a similar intensity of interference event within a very short time window, the collaborative decision is made as a valid score. This dual-node collaborative decision-making mechanism fundamentally eliminates misjudgments caused by simultaneous actions from both sides or common-mode noise in the environment. After the score is determined, the system will trigger corresponding voice broadcasts or gamified sound effects feedback based on the user's selected "professional mode" or "entertainment mode," and will synchronize score updates and key match milestones (such as match points) to the cloud in real time. The cloud server not only records scores and updates friend leaderboards but also pushes the match progress to authorized viewing mini-programs with low latency, achieving remote "live streaming." In addition, the system provides a one-click undo mechanism based on screen touch, allowing users to quickly revoke the most recent score when a misjudgment occurs, ensuring the system's usability and user control.

[0035] Compared to existing technologies, traditional solutions suffer from high barriers to entry and rigid functionality due to their reliance on dedicated hardware, and lack effective anti-interference and engaging interactive designs. This solution innovatively utilizes ubiquitous smartphones as distributed sensing nodes, effectively filtering out interference from complex environments through a dual-end collaborative decision-making algorithm, ensuring the reliability of the core scoring function. Simultaneously, the system creatively integrates automated scoring with customizable entertaining sound effects and lightweight social event services. Through a ready-to-use mobile application and cloud collaboration, it achieves a full-link experience upgrade from accurate sensing and engaging interaction to social sharing. Dynamic collaborative verification, one-click rollback and error correction, and real-time cloud synchronization collectively constitute a highly reliable, engaging, easily accessible, and socially-oriented intelligent scoring and entertainment solution for table tennis.

[0036] After introducing the basic concept of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0037] Example 1:

[0038] Please see Figure 1 As shown, a smart scoring method for table tennis entertainment based on cloud-based collaborative data is applied to a first mobile terminal with a table tennis scoring application installed. The first mobile terminal is rigidly coupled to one side of the table tennis table via a physical support. The method includes the following steps:

[0039] The first mobile terminal receives a real-time acceleration data stream from its built-in inertial sensor. The real-time acceleration data stream includes triaxial acceleration components generated by the first mobile terminal as it vibrates with the ping-pong table surface.

[0040] The real-time acceleration data stream is denoised and its coordinate system is transformed using a preset signal preprocessing algorithm to extract the vertical vibration energy sequence perpendicular to the table surface.

[0041] The vertical vibration energy sequence is input into a preset action recognition model for feature matching. When the matched vibration waveform features meet the preset scoring slapping action threshold, first candidate scoring event data is generated. The first candidate scoring event data includes the local trigger time point and vibration intensity confidence.

[0042] The system sends a status verification request containing the first candidate score event data to the second mobile terminal on the other side through the established data exchange channel, and receives the time window status feedback data returned by the second mobile terminal.

[0043] Based on the time window status feedback data, collaborative conflict detection is performed on the first candidate score event data. When it is confirmed that the second mobile terminal has not generated a high confidence interference event within the preset time domain window corresponding to the local trigger time point, a final valid score instruction is generated.

[0044] The current match score data is updated according to the final valid score instruction, and the corresponding multimedia sound effect resource data is called for playback output according to the currently loaded entertainment mode configuration parameters.

[0045] Preferably, the step of using a preset signal preprocessing algorithm to denoise and transform the real-time acceleration data stream includes:

[0046] Obtain the gravitational acceleration component data of the first mobile terminal in a stationary state, and calculate the tilt matrix of the device coordinate system relative to the physical coordinate system of the platform;

[0047] The real-time acceleration data stream is mapped to the physical coordinate system of the platform using the tilt matrix, and the vertical component data of the Z-axis is separated.

[0048] The Z-axis vertical component data is subjected to bandpass filtering to filter out signal components that are lower than the preset ambient noise frequency and higher than the preset natural frequency of the platform, thereby generating the vertical vibration energy sequence for use by the motion recognition model.

[0049] Specifically, the process first involves adaptive attitude calibration and coordinate system decoupling based on static gravity field characteristics. During the static initialization phase after application startup, this process continuously collects raw triaxial acceleration data from the accelerometer built into the first mobile terminal at a preset sampling frequency (preferably 100Hz-200Hz, to satisfy the Nyquist sampling theorem and take power consumption into account). The moving average filter is applied to eliminate high-frequency white noise interference, and a gravity reference vector that characterizes the current tilt attitude of the device in physical space is calculated. ,in Divided into the average acceleration values ​​of the x, y, and z axes, with units of . ;

[0050] Subsequently, a coordinate transformation matrix is ​​constructed and vertical component separation along the Z-axis is implemented. Specifically, a projection operator is constructed by calculating the direction cosine of the gravity reference vector to separate the real-time non-stationary acceleration data stream. Dynamically map to a physical coordinate system with the normal to the ping-pong table surface as the Z-axis, the calculation formula is as follows: ,in The net vibration acceleration component perpendicular to the platform is separated, with units of . ,· represents the vector dot product operation. The normalized unit vector of the direction of gravity. The standard gravitational acceleration constant (taken as 9.8) The purpose of this calculation formula is to eliminate the inter-axis coupling error caused by different bracket installation angles (such as horizontal, vertical or oblique placement of mobile phones) through geometric projection, and only retain the vertical vibration component that is effective for scoring.

[0051] Next, the separated A digital bandpass filter based on a second-order Butterworth model is performed to remove environmental background noise and interference in non-hitting frequency bands. The time-domain difference equation of the filter is: ,in The current time is the input. The value, y[n] is the filtered output value. , The filter coefficients are divided into dimensionless filter coefficients calculated based on a preset cutoff frequency, where p represents the index number, which can take values ​​from 0 to n; the preset cutoff frequency is configured as a passband range. The lower cutoff frequency It is recommended to set it to 10Hz to filter out low-frequency vibration signals caused by people walking or slightly leaning against the table. The upper limit cutoff frequency is [not specified]. It is recommended to set it to 300Hz to filter out the thermal noise of the sensor itself and high-frequency spikes that are much higher than the characteristic frequency of ping-pong ball impact, so as to obtain a pure vertical vibration signal sequence.

[0052] Finally, a short-time energy integration is performed on the filtered signal to generate a vertical vibration energy sequence for use by the motion recognition model. Vertical vibration energy sequence It is a set of multiple signal energy values The time series data is composed of the following formulas: ,in The signal energy value of the k-th frame, in units of W is the width of the sliding window (preset to the number of sampling points corresponding to 50ms, covering the decay period of one ball impact vibration), and S is the sliding step size. It transforms the waveform data of high-frequency oscillation into a smooth energy envelope that can characterize the impact force, thus completing the data dimensionality reduction and feature enhancement from the original physical signal to the feature energy sequence.

[0053] Preferably, the step of inputting the vertical vibration energy sequence into a preset motion recognition model for feature matching includes:

[0054] A waveform segment exceeding a preset trigger level is extracted from the vertical vibration energy sequence;

[0055] Extract the time-domain feature data and frequency-domain feature data of the waveform segment. The time-domain feature data includes at least the peak amplitude, energy attenuation rate and waveform duration. The frequency-domain feature data includes at least the main frequency energy percentage.

[0056] Calculate the Euclidean distance between the time-domain feature data and the frequency-domain feature data and the pre-stored standard slapping score action template;

[0057] When the Euclidean distance is less than the preset judgment distance threshold, the action is judged to be matched successfully, and the calculated matching similarity is converted into the vibration intensity confidence score, which is used to generate the first candidate score event data.

[0058] Specifically, for vertical vibration energy sequences Endpoint detection based on adaptive dual thresholds is performed, and the trigger level is dynamically set using a long short-term memory algorithm. When the short-term energy mean exceeds the background noise baseline... Double the time, lock the start time point. and intercept from to Data within the time window is used as the waveform segment to be analyzed. ,in For the preset capture duration, it is recommended to set it to 200ms to 400ms to fully cover the free decay oscillation period generated by the tap. The preset trigger coefficient is recommended to be set to 3.0-5.0 to distinguish between accidental collisions and intentional slaps.

[0059] Next, the extracted waveform segments Perform deep feature extraction and construct a four-dimensional feature vector. , where T denotes the transpose of the matrix. The peak amplitude, i.e. The unit is ; The energy decay rate reflects the damping characteristics of the table vibration, and is calculated using the following formula: The unit is , here At peak time, Let ln(E(·)) be the moment when the energy decays to 10% of its peak value, and let ln(E(·)) represent the logarithmic function with the natural logarithm e as the base. The duration of the effective waveform, in seconds; The proportion of main frequency energy, through the... The corresponding raw acceleration data is subjected to a Fast Fourier Transform to obtain the spectrum X(f), which is then calculated using the following formula: , is a dimensionless ratio, where [ , [This is a preset frequency band for the impact characteristics. For the wooden structure of the ping-pong table, this frequency band is preferably set to 80Hz-150Hz to distinguish between the impact (mid frequency) and the ping-pong ball collision (high-frequency crisp sound) or the human body collision (low-frequency dull sound).]

[0060] The constructed feature vector will then be Compared with the pre-stored standard slapping scoring action template vector Perform distance measurement. These baseline parameters were obtained by collecting a large number of standard tapping motions in a laboratory environment and averaging the results. To eliminate the influence of inconsistent dimensions among the various features, a weighted Euclidean distance algorithm was used for matching calculations, as shown in the formula below. ,in This is a distance metric. and These are the i-th feature components of the input vector and the template vector, respectively; This is the standard deviation of the feature component in the training set, used for normalization. The preset feature weight coefficients, and satisfy the following conditions: Considering that the "intensity" and "dullness" of the striking action are the core distinguishing points, it is recommended to set the weights w_1 and w_4 of the corresponding peak amplitude and the proportion of the main frequency energy to a higher value to enhance the robustness of recognition.

[0061] Finally, based on the calculated distance Determine the validity of the action and generate a confidence score. If Less than the preset judgment distance threshold If the match is successful, the distance is mapped to a normalized vibration intensity confidence level using a Gaussian kernel function. The conversion formula is: ,in To control the bandwidth parameter of the confidence decay rate, this That is, the data of the first candidate scoring event.

[0062] Preferably, the step of performing collaborative conflict detection on the first candidate score event data based on the time window state feedback data includes:

[0063] The time window status feedback data is analyzed, and the maximum vibration amplitude data detected by the second mobile terminal within a preset collaborative verification period before and after the local trigger time point is extracted.

[0064] Compare the maximum vibration amplitude data with a preset interference judgment threshold;

[0065] If the maximum vibration amplitude data is lower than the interference determination threshold, the first candidate score event data is determined to be a valid unilateral tap, and the final valid score instruction for confirming the score is generated.

[0066] If the maximum vibration amplitude data is higher than the interference determination threshold, it is determined that there is simultaneous interference from both sides or non-scoring ball impact at the current moment, an event discard command is generated and the subsequent scoring process is terminated.

[0067] Specifically, the first mobile terminal generates a data containing a local trigger time point. After obtaining the first candidate score event data, a collaborative conflict detection process based on cross-terminal timing consistency is initiated. First, the time window status feedback data packets transmitted back by the second mobile terminal through the data exchange channel are parsed, and the corresponding events are extracted from them. Vertical acceleration amplitude sequence within a preset co-verification time ΔT, symmetrically distributed before and after. The value of △T is based on the weighted sum of network synchronization error and the propagation delay of physical vibration energy in the ball table medium. It is recommended to set it to a dynamic range of 80ms to 200ms. To ensure that the complete interference signal envelope is captured and to avoid cross-frame aliasing, it is recommended to select 100ms as the verification time.

[0068] Next, the extreme value retrieval algorithm is used in the sequence Calculate the maximum vibration amplitude data The calculation formula is: ,in The value of the vertical component acceleration of the second mobile terminal at time t after coordinate transformation is given in units of . , This characterizes the maximum vibration disturbance intensity generated in the opposite table area at the moment the first mobile terminal triggers a suspected scoring action.

[0069] Then, the extracted A quantitative comparison is performed with a preset interference judgment threshold. This threshold is a physical limit pre-calibrated based on the response amplitude of different table tennis table materials to non-contact impacts (such as a ping-pong ball directly hitting the table surface, environmental footsteps, etc.), and its value is recommended to be set within the range of 1.5. Up to 4.5 In this embodiment, the value is preferably set to 2.5. To achieve a balance between sensitivity and robustness;

[0070] If the judgment logic relationship is satisfied If the score is less than the preset interference threshold, the first candidate score event is determined to have spatial locality characteristics, excluding the possibility of simultaneous mechanical impact on both sides. It is then marked as a single-sided effective tap and a final effective score instruction containing the current scorer ID, timestamp, and confirmation mark is generated to drive the local score logic register to accumulate.

[0071] Conversely, if If the value is greater than or equal to the preset interference judgment threshold, it is determined that the table has experienced global vibration at the current moment (such as bilateral resonance caused by the ball bouncing after landing or a large-scale impact on the table by a person). The system logic automatically generates an event discard instruction, masks the current first candidate scoring event, and terminates the scoring process, thereby eliminating the false triggering deviation through the difference in vibration distribution in physical space.

[0072] Preferably, the step of calling the corresponding multimedia sound effect resource data according to the currently loaded entertainment mode configuration parameters includes:

[0073] Read the current match score data and analyze the key node characteristics of the match process. The key node characteristics include match point status, score reversal status, and winning streak status.

[0074] Based on the sound effect theme package index determined by the entertainment mode configuration parameters, and combined with the key node features, the target sound effect file address is retrieved from the preset audio database.

[0075] The audio data stream corresponding to the target sound effect file address is loaded into the audio output buffer and played synchronously with the final valid score instruction to achieve differentiated auditory feedback.

[0076] Specifically, after generating the final valid score instruction, the system enters a multimedia dynamic feedback stage based on the game context awareness. First, it reads the current game score data by accessing the local score logic register. and historical score time series, among which This is the current total score for the scoring team. To assess the opponent's total score, a game tension evaluation model is constructed to quantify the criticality of the current stage. This model calculates a comprehensive tension index. To characterize, the calculation formula is as follows ,in For local point identification operators, when or When the target score is reduced by 1 (e.g., 10 points in an 11-point system), the score is reduced by 1. The value is 1 if it is set to 1, and 0 otherwise. The score reversal identification operator takes a value of 1 if the score causes the leader to switch sides, and 0 otherwise. This represents the current scoring team's winning streak by a maximum of [number] goals. , , The preset emotional weighting coefficients are all dimensionless constants. To highlight the importance of the key moments and ensure the progression of the entertainment effect, it is recommended to... (Location weight) is set to 10.0 for highest priority triggering. (Overtaking weight) is set to 5.0. The (winning streak weight) is set to 2.0. This weighted calculation transforms the abstract game process into a quantifiable tension value.

[0077] Next, the system configures the entertainment mode based on the user's pre-selected settings (such as the theme index corresponding to "Fighting Mode" or "Variety Show Mode"). ), combined with the comprehensive tension index generated above The target sound effect file address is retrieved from a pre-defined discrete audio database using a hash mapping function. The retrieval logic satisfies ,in For quantification, the continuous tension index is divided into three level intervals: "normal," "intense," and "decisive." The recommended value ranges are [0,5), [5,10), and [10,+]. Subsequently, the system calls the audio decoding kernel to... The compressed audio data stream is decompressed and loaded into a loop audio output buffer. The buffer size is recommended to be set to 1024 to 4096 sampling points according to the preset system latency tolerance. To ensure the perceptual consistency between sound effect broadcasts and table-tapping actions, a preferred size of 2048 sampling points is recommended to achieve near real-time feedback. Finally, the audio data stream in the buffer is sent to the hardware speaker for pulse width modulation output at the output driver layer, thereby transforming a single scoring event into differentiated auditory feedback with a game immersive feel, completing a deep transformation from game logic data to interactive entertainment experience.

[0078] Preferably, the method further includes a misjudgment rollback processing step:

[0079] Within a preset reversible time after the final valid score instruction is generated, the touch signal of the first mobile terminal touch screen input interface is continuously monitored.

[0080] When a touch signal matching the preset reversal gesture characteristics is detected, score rollback request data is generated;

[0081] Based on the score rollback request data, the current match score data is restored to the numerical state before the final valid score instruction was generated, and a rollback synchronization data packet is generated.

[0082] The rollback synchronization data packet is sent to the second mobile terminal through the data exchange channel to drive the second mobile terminal to synchronously execute the rollback operation of the score display.

[0083] Specifically, after generating the final valid score instruction, the system immediately starts a time-domain monitoring timer with the highest thread priority, and performs a pre-set reversible operation for a set duration. The system continuously and asynchronously polls the raw touch event stream of the first mobile terminal's touchscreen input interface. It is recommended to set it to an adjustable range of 3.0 seconds to 10.0 seconds. To ensure the immediacy of error correction response and prevent the score chain from becoming chaotic due to long time spans, 5.0 seconds is recommended as the preferred value.

[0084] This process captures pressure values ​​in real time by calling the operating system's underlying touch sensing operators. Touch coordinate point sequence and duration of contact The raw touch data is input into a pre-set gesture recognition dynamics model to calculate the confidence level of the withdrawal action. The calculation formula is: ,in The duration of physical dwell time of the finger in the preset retraction function area, in milliseconds; The pixel displacement deviation generated during touch is expressed in pixels and is calculated as the Euclidean distance between the starting and ending points. For time-incentive coefficient, These are spatial penalty coefficients, all dimensionless constants. Their underlying logic is that a longer dwell time and smaller displacement indicate a clearer subjective intent to correct errors. In this embodiment, to avoid accidental touches, it is recommended to... Set to 0.005. Set to 0.1. and These represent the maximum set physical dwell time and the maximum pixel displacement deviation, respectively.

[0085] when If the confidence threshold is exceeded, a touch signal matching the preset withdrawal gesture characteristics is detected. The system then retrieves the previous moment's match metadata vector from the local state loop stack. Based on this, score rollback request data is generated. This rollback logic follows the inverse state transition operation, expressed by the formula as follows: ,in This is the corrected score after the rollback. This is the most recent score increment vector (e.g., [1,0] or [0,1]). Represents the XOR update of the logical state bit (involving the inversion of the serve change logic). This represents the current game state vector, including information such as the current score and the serving side;

[0086] Subsequently, based on the score rollback request data, the system forcibly restores the current match score data in local memory to the historical snapshot state, and simultaneously encapsulates a rollback synchronization data packet containing a global sequence number, checksum, and rollback flag, which is sent to the second mobile terminal via the data exchange channel using the UDP / TCP protocol. After receiving the data packet, the second mobile terminal parses and verifies the continuity of the sequence number. If the verification is successful, it forcibly resets its local display buffer and logic registers according to the correction value in the data packet, thereby driving the second mobile terminal to synchronously execute the rollback operation of the score display. This ultimately eliminates the scoring data deviation caused by system hardware perception errors or athlete human error, ensuring absolute consistency of the scoring results between the two ends under the distributed perception system.

[0087] Preferred, such as Figure 2 As shown, before receiving the real-time acceleration data stream from the inertial sensor built into the first mobile terminal, a self-test step is also included:

[0088] Detect the static posture data of the first mobile terminal and determine whether the static posture data is within a preset reasonable installation angle range;

[0089] If the installation angle is within a reasonable range, a calibration command will be output to prompt the user to tap the tabletop.

[0090] The system receives calibration vibration signals generated by the user tapping the table surface and dynamically adjusts the scoring tapping action threshold based on the transmission characteristic parameters of the calibration vibration signals to adapt to the physical attenuation characteristics of table tennis tables made of different materials.

[0091] Specifically, before the system enters the formal sensing phase, the installation attitude assessment process based on gravity vector distribution is first initiated. This process obtains the raw triaxial acceleration components in a stationary state by calling the built-in accelerometer of the first mobile terminal. The installation angle between the main plane of the equipment and the horizontal plane is calculated using inverse trigonometric functions. and compare it with the preset reasonable installation angle range. A comparison was conducted, among which It is recommended to set it to 30.0°. It is recommended to set the angle to 90.0°. The logic behind this setting is as follows: if the included angle is too small (less than 30°), the vertical component of the phone's center of gravity will cause non-rigid slippage between it and the contact surface with the bracket, weakening the fidelity of high-frequency vibration transmission; if the comparison result is within this range, the UI will output a calibration command prompting the user to tap the table, guiding the user to preset the distance from the phone's deployment position. A standard-intensity pulse strike is performed at a distance of 200mm to 500mm.

[0092] Subsequently, the system enters the physical conduction characteristic modeling stage, which calculates the energy conduction coefficient of the table surface material by real-time acquisition of the calibration vibration signal sequence generated by the impact and performing time-domain feature extraction. The calculation formula is: ,in To calibrate the maximum vibration acceleration amplitude of the signal in the vertical direction, the unit is... ; The preset standard impact energy reference value is calibrated based on a standard wooden ping-pong table in the laboratory. This is the material damping correction factor; The decay time of a signal from its peak to the background noise level, expressed in seconds, reflects the absorption characteristics of the ball table material (such as wood, composite materials, or metal) for mechanical waves.

[0093] Finally, based on the calculations... The nonlinear mapping function is used to determine the threshold for the scoring tapping action in the core scoring algorithm. Dynamic compensation is performed, and the adjustment formula is as follows: ,in The adjusted vertical vibration energy sequence threshold, in units of ; The system's preset baseline trigger threshold (preferably ranging from 15.0 to 25.0) is used. ); To prevent logarithmic overflow, a smoothing constant (with a value of 1.0) is used. This compensation logic ensures that the threshold is appropriately increased on hard, highly conductive surfaces to filter ball bounce interference, while the threshold is decreased on soft or porous surfaces to compensate for energy loss, thereby outputting adaptive deployment status data that can adapt to different physical environments.

[0094] Preferably, the method further includes a real-time live data distribution step:

[0095] The match event log data, which includes the final valid score instruction, updated match score data, and user identification of both sides, is encapsulated into an encrypted data packet.

[0096] The encrypted data packets are uploaded to the cloud server via the wireless network communication module.

[0097] Receive the real-time win rate ranking data and friend leaderboard update data generated by the cloud server based on historical match data and match event log data, and render the real-time win rate ranking data to the local display interface;

[0098] After receiving the encrypted data packet, the cloud server parses and extracts the score change event;

[0099] The score change event will be pushed to the mini-program interface of a third-party mobile terminal that has been authorized to watch the game;

[0100] The third-party mobile terminal refreshes the score numbers on the remote viewing interface in real time based on the score change event, enabling low-latency live streaming of the match.

[0101] Example 2:

[0102] See Figure 3 As shown, a cloud-based collaborative data-driven intelligent scoring system for table tennis entertainment includes:

[0103] An acceleration data acquisition module is used to receive real-time acceleration data streams from the inertial sensor built into the first mobile terminal, wherein the real-time acceleration data streams include triaxial acceleration components generated by the first mobile terminal as the ping-pong table vibrates.

[0104] The vibration signal preprocessing module is used to denoise and transform the real-time acceleration data stream using a preset signal preprocessing algorithm, and extract the vertical vibration energy sequence perpendicular to the table surface.

[0105] The slapping action recognition module is used to input the vertical vibration energy sequence into the preset action recognition model for feature matching. When the matched vibration waveform features meet the preset slapping action threshold, the first candidate score event data is generated. The first candidate score event data includes the local trigger time point and vibration intensity confidence.

[0106] The multi-terminal collaborative verification module is used to send a status verification request containing first candidate score event data to the second mobile terminal on the other side through the established data exchange channel, and to receive time window status feedback data returned by the second mobile terminal.

[0107] The scoring event arbitration module is used to perform collaborative conflict detection on the first candidate scoring event data based on the time window status feedback data. When it is confirmed that the second mobile terminal has not generated a high confidence interference event within the preset time domain window corresponding to the local trigger time point, the final valid scoring instruction is generated.

[0108] The game score update and response module is used to update the current game score data according to the final valid score instruction, and to call the corresponding multimedia sound effect resource data for playback output according to the currently loaded entertainment mode configuration parameters.

[0109] Example 3:

[0110] like Figure 4 As shown, the intelligent scoring device for table tennis entertainment based on cloud-based collaborative data includes a processor, memory, and communication bus;

[0111] The memory stores a computer-readable program that can be executed by the processor;

[0112] The communication bus enables communication between the processor and the memory;

[0113] When the processor executes the computer-readable program, it performs steps in the intelligent scoring method for table tennis entertainment based on cloud-based collaborative data as described in any one of the present invention.

[0114] It should be noted that the interval and threshold sizes are set for ease of comparison. The size of the threshold depends on the amount of sample data and the base number set by those skilled in the art for each set of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. Furthermore, the above formulas are all dimensionless calculations, and the formulas are derived from software simulations using a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0115] It should be understood that, in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0116] It should be understood that determining B based on A does not mean determining B solely based on A; it also means determining B based on A and / or other information.

[0117] 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.

[0118] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart scoring method for table tennis entertainment based on cloud-based collaborative data, characterized in that, The method includes: The real-time acceleration data stream is received from the inertial sensor built into the first mobile terminal, wherein the real-time acceleration data stream includes the triaxial acceleration components generated by the first mobile terminal as the ping-pong table vibrates. The real-time acceleration data stream is denoised and the coordinate system is transformed using a preset signal preprocessing algorithm to extract the vertical vibration energy sequence perpendicular to the table surface. The vertical vibration energy sequence is input into a preset motion recognition model for feature matching. When the matched vibration waveform features meet the preset scoring slapping action threshold, the first candidate scoring event data is generated. The first candidate scoring event data includes the local trigger time point and vibration intensity confidence. The system sends a status verification request containing the first candidate score event data to the second mobile terminal on the other side through the established data exchange channel, and receives the time window status feedback data returned by the second mobile terminal. Based on the time window status feedback data, collaborative conflict detection is performed on the first candidate score event data. When it is confirmed that the second mobile terminal has not generated a high confidence interference event within the preset time domain window corresponding to the local trigger time point, the final valid score instruction is generated. Update the current game score data according to the final valid score instruction, and call the corresponding multimedia sound effect resource data for playback output according to the currently loaded entertainment mode configuration parameters.

2. The intelligent scoring method for table tennis entertainment based on cloud-based collaborative data as described in claim 1, characterized in that, The real-time acceleration data stream is denoised and its coordinate system transformed using a pre-defined signal preprocessing algorithm, including: Obtain the gravitational acceleration component data of the first mobile terminal in a stationary state, and calculate the tilt matrix of the device coordinate system relative to the physical coordinate system of the platform; The real-time acceleration data stream is mapped to the physical coordinate system of the platform using the tilt matrix, and the vertical component data of the Z-axis is separated. Bandpass filtering is performed on the Z-axis vertical component data to filter out signal components that are lower than the preset ambient noise frequency and higher than the preset natural frequency of the platform, thereby generating a vertical vibration energy sequence.

3. The intelligent scoring method for table tennis entertainment based on cloud-based collaborative data as described in claim 1, characterized in that, The vertical vibration energy sequence is input into a pre-set motion recognition model for feature matching, including: Extract waveform segments exceeding a preset trigger level from the vertical vibration energy sequence; Extract time-domain and frequency-domain feature data of waveform segments. The time-domain feature data includes at least peak amplitude, energy attenuation rate and waveform duration. The frequency-domain feature data includes at least the main frequency energy percentage. Calculate the Euclidean distance between the time-domain feature data and the frequency-domain feature data and the pre-stored standard slapping score action template; When the Euclidean distance is less than the preset judgment distance threshold, the action is judged to be successfully matched, and the calculated matching similarity is converted into vibration intensity confidence, which is used to generate the first candidate score event data.

4. The intelligent scoring method for table tennis entertainment based on cloud-based collaborative data as described in claim 1, characterized in that, Collaborative conflict detection is performed on the first candidate score event data based on time window state feedback data, including: Analyze the time window status feedback data and extract the maximum vibration amplitude data detected by the second mobile terminal within the preset collaborative verification time before and after the local trigger time point; Compare the maximum vibration amplitude data with the preset interference judgment threshold; If the maximum vibration amplitude data is lower than the interference judgment threshold, the first candidate scoring event data is determined to be a valid unilateral tap, and a final valid score instruction is generated to confirm the score. If the maximum vibration amplitude data is higher than the interference judgment threshold, it is determined that there is a non-scoring ball impact at the current moment, an event discard instruction is generated and the subsequent scoring process is terminated.

5. The intelligent scoring method for table tennis entertainment based on cloud-based collaborative data according to claim 1, characterized in that, Based on the currently loaded entertainment mode configuration parameters, the corresponding multimedia sound effect resource data is retrieved, including: Read the current match score data and analyze the key node characteristics of the match process. The key node characteristics include match point status, score reversal status, and winning streak status. Based on the sound effect theme package index determined by the entertainment mode configuration parameters, and combined with key node features, the target sound effect file address is retrieved from the preset audio database. The audio data stream corresponding to the target sound effect file address is loaded into the audio output buffer and played synchronously with the final valid score instruction.

6. The intelligent scoring method for table tennis entertainment based on cloud-based collaborative data as described in claim 1, characterized in that, The method also includes steps for handling misjudgments and rollbacks: Within a preset reversible time after the final valid score instruction is generated, the touch signal of the touch input interface of the first mobile terminal is continuously monitored. When a touch signal matching the preset reversal gesture characteristics is detected, score rollback request data is generated; Based on the score rollback request data, restore the current game score data to the value state before the final valid score instruction was generated, and generate a rollback synchronization data packet; The rollback synchronization data packet is sent to the second mobile terminal through the data exchange channel to drive the second mobile terminal to synchronously execute the rollback operation of the score display.

7. The intelligent scoring method for table tennis entertainment based on cloud-based collaborative data as described in claim 1, characterized in that, Before receiving the real-time acceleration data stream from the inertial sensor built into the first mobile terminal, a deployment self-test step is also included: Detect the stationary posture data of the first mobile terminal and determine whether the stationary posture data is within the preset reasonable installation angle range; If the installation angle is within a reasonable range, a calibration command will be output to prompt the user to tap the tabletop. It receives calibration vibration signals generated by users tapping the tabletop and dynamically adjusts the scoring tapping action threshold based on the transmission characteristics parameters of the calibration vibration signals.

8. The intelligent scoring method for table tennis entertainment based on cloud-based collaborative data as described in claim 1, characterized in that, The method also includes a real-time live data distribution step: The match event log data, which includes the final valid score instruction, updated match score data, and user identification of both sides, is encapsulated into an encrypted data packet. The encrypted data packets are uploaded to the cloud server via the wireless network communication module. Receive real-time win rate ranking data and friend leaderboard update data generated by the cloud server based on historical match data and match event log data, and render the real-time win rate ranking data to the local display interface. After receiving the encrypted data packet, the cloud server parses and extracts the score change event; Push score change events to the mini-program interface of third-party mobile terminals that have been authorized to watch the game; Third-party mobile terminals refresh the score numbers on the remote viewing interface in real time based on the score change events, enabling low-latency live streaming of the match.

9. A cloud-based collaborative data-driven intelligent scoring system for table tennis entertainment, characterized in that: It is implemented based on the cloud-based collaborative data-driven intelligent scoring method for table tennis entertainment as described in any one of claims 1-8. The system includes: An acceleration data acquisition module is used to receive real-time acceleration data streams from the inertial sensor built into the first mobile terminal, wherein the real-time acceleration data streams include triaxial acceleration components generated by the first mobile terminal as the ping-pong table vibrates. The vibration signal preprocessing module is used to denoise and transform the real-time acceleration data stream using a preset signal preprocessing algorithm, and extract the vertical vibration energy sequence perpendicular to the table surface. The slapping action recognition module is used to input the vertical vibration energy sequence into the preset action recognition model for feature matching. When the matched vibration waveform features meet the preset slapping action threshold, the first candidate score event data is generated. The first candidate score event data includes the local trigger time point and vibration intensity confidence. The multi-terminal collaborative verification module is used to send a status verification request containing first candidate score event data to the second mobile terminal on the other side through the established data exchange channel, and to receive time window status feedback data returned by the second mobile terminal. The scoring event arbitration module is used to perform collaborative conflict detection on the first candidate scoring event data based on the time window status feedback data. When it is confirmed that the second mobile terminal has not generated a high confidence interference event within the preset time domain window corresponding to the local trigger time point, the final valid scoring instruction is generated. The game score update and response module is used to update the current game score data according to the final valid score instruction, and to call the corresponding multimedia sound effect resource data for playback output according to the currently loaded entertainment mode configuration parameters.

10. A smart scoring device for table tennis entertainment based on cloud-based collaborative data, characterized in that, Includes processor, memory, and communication bus; The memory stores a computer-readable program that can be executed by the processor; The communication bus enables communication between the processor and the memory; When the processor executes the computer-readable program, it performs the steps of the table tennis entertainment-oriented intelligent scoring method based on cloud collaborative data as described in any one of claims 1 to 8.

Images