Intelligent table tennis scoring system and method based on wireless touch and mobile terminal
By collecting audio and vibration signals from the impact of ping-pong balls, calculating signal quality indicators, and performing weighted fusion, the problem of inaccurate scoring caused by environmental interference was solved, and a highly reliable scoring system was achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SILINGJIA SPORTS CULTURE IND (JIUJIANG) CO LTD
- Filing Date
- 2026-01-31
- Publication Date
- 2026-04-21
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In existing technologies, automated scoring systems for table tennis matches suffer from signal distortion due to environmental interference, making it difficult to accurately distinguish between table and ground impacts, thus affecting scoring accuracy and the fairness of the competition.
Audio and vibration signals are collected synchronously by mobile terminals, signal quality indicators are calculated, features are extracted and weighted fusion is performed, and highly reliable scoring results are generated by combining adaptive judgment intervals and iterative weight correction mechanisms.
It significantly improves the accuracy of scoring and the reliability of judgment, reduces the occurrence of missed or incorrect judgments, and ensures the accuracy of impact type identification and the reliability of scoring.
Smart Images
Figure CN121891765A_ABST
Abstract
Description
Technical Field
[0001] This application belongs to the field of intelligent scoring technology for sports events, and specifically relates to an intelligent table tennis scoring system and method based on wireless touch and mobile terminals. Background Technology
[0002] Some automated scoring systems in table tennis use audio or vibration signals collected by mobile terminals to distinguish the type of impact of the table tennis ball. By analyzing the sound waves or vibration waveforms generated after the table tennis ball is hit, it is determined whether it hits the table surface or the ground directly. When an impact on the table surface is detected, a scoring operation is performed, while when an impact on the ground is detected, no points are awarded.
[0003] However, factors such as people moving around, touching the table, and external sounds can affect the signals collected by the mobile terminal, making the signal characteristics prone to superposition or distortion. This makes it difficult to effectively distinguish the difference between the impact on the table and the ground, thus affecting the accuracy of impact type identification. Consequently, the accuracy of automated scoring is reduced, and situations such as omissions and misrecordings are likely to occur, interfering with the smooth progress of the game and affecting the fairness of the game results. Summary of the Invention
[0004] This application provides an intelligent table tennis scoring system and method based on wireless touch and mobile terminals, which effectively solves the problems of signal feature distortion caused by environmental interference, insufficient accuracy of impact type recognition, and easy impact on scoring accuracy in the prior art. It significantly improves scoring accuracy and decision reliability, and reduces missed and misjudged cases.
[0005] To achieve the above objectives, this application adopts the following technical solution:
[0006] In a first aspect, this application provides a smart table tennis scoring method based on wireless touch and a mobile terminal, comprising: acquiring audio signals and vibration signals, wherein the audio signals and vibration signals are generated by a mobile terminal attached to the side of the table during the process of the table tennis ball hitting the table surface or the ground; calculating the audio signal-to-noise ratio and vibration clarity based on the audio signals and vibration signals, and extracting audio features and vibration features; calculating a first comprehensive support of the audio features to the table impact and a second comprehensive support of the vibration features to the table impact, assigning weights to the first comprehensive support and the second comprehensive support and weighted summing to obtain a fusion confidence score; determining a judgment interval based on the audio signal-to-noise ratio and vibration clarity; comparing the fusion confidence score with the judgment interval, and if the fusion confidence score falls within the judgment interval, then correcting the weights based on the audio signal-to-noise ratio and vibration clarity, and returning to the step of calculating the fusion confidence score, until a preset maximum number of iterations is reached or the fusion confidence score falls outside the judgment interval; if the fusion confidence score falls outside the judgment interval, then generating a scoring result based on the fusion confidence score.
[0007] Furthermore, calculating the audio signal-to-noise ratio and vibration intelligibility based on the audio signal and vibration signal includes: calculating the ratio of the signal power of the audio signal to the background noise power to obtain the audio signal-to-noise ratio; and calculating the proportion of vibration energy matching the characteristic frequency band of the ping-pong ball impact to the total vibration energy to obtain the vibration intelligibility.
[0008] Furthermore, audio features and vibration features are extracted, including: audio features such as high-frequency energy ratio, pulse width, and time difference between the two ends; vibration features such as high-frequency vibration energy, attenuation constant, and waveform symmetry.
[0009] Further, calculating the first comprehensive support of audio features for table impact and the second comprehensive support of vibration features for table impact includes: matching audio features with a preset table impact audio feature template to obtain a first matching degree; matching audio features with a preset ground impact audio feature template to obtain a second matching degree; and using the difference between the first and second matching degrees as the first comprehensive support; matching vibration features with a preset table impact vibration feature template to obtain a third matching degree; matching vibration features with a preset ground impact vibration feature template to obtain a fourth matching degree; and using the difference between the third and fourth matching degrees as the second comprehensive support.
[0010] Further, weights are assigned to the first comprehensive support and the second comprehensive support, and the weighted sum is obtained to obtain the fusion confidence score, including: assigning a first weight to the first comprehensive support based on the audio signal-to-noise ratio, and assigning a second weight to the second comprehensive support based on the vibration clarity. The first weight is positively correlated with the audio signal-to-noise ratio, and the second weight is positively correlated with the vibration clarity. The product of the first comprehensive support and the first weight is added to the product of the second comprehensive support and the second weight to obtain the fusion confidence score.
[0011] Furthermore, the determination interval is determined based on the audio signal-to-noise ratio and vibration clarity, including: calculating the comprehensive signal quality coefficient based on the audio signal-to-noise ratio and vibration clarity; and calculating the upper and lower boundary values of the determination interval based on the preset basic threshold and the comprehensive signal quality coefficient.
[0012] Further, based on the preset base threshold and the comprehensive signal quality coefficient, the upper and lower boundary values of the judgment interval are calculated, including: calculating the bandwidth coefficient based on the first support and the second support; when the product of the first comprehensive support and the second comprehensive support is greater than zero, the bandwidth coefficient is the preset first coefficient; otherwise, the bandwidth coefficient is the preset second coefficient, and the second coefficient is greater than the first coefficient; the upper and lower boundary values of the judgment interval are calculated based on the preset base threshold, the bandwidth coefficient, and the comprehensive signal quality coefficient.
[0013] Furthermore, the weights are adjusted based on the audio signal-to-noise ratio (SNR) and vibration intelligibility, including: normalizing the audio SNR and vibration intelligibility and calculating the difference rate; multiplying the difference rate by a preset adjustment coefficient to obtain the weight adjustment range; comparing the normalized audio SNR and vibration intelligibility: if the audio SNR is greater than the vibration intelligibility, the first weight is increased by the weight adjustment range, and the second weight is decreased by the weight adjustment range; if the vibration intelligibility is greater than the audio SNR, the second weight is increased by the weight adjustment range, and the first weight is decreased by the weight adjustment range.
[0014] Furthermore, a scoring result is generated based on the fusion confidence level, including: if the fusion confidence level is greater than the upper boundary of the judgment interval, it is judged as a table impact, and the scoring result is a valid ball inside the table and a score is awarded; if the fusion confidence level is less than the lower boundary of the judgment interval, it is judged as a ground impact, and the scoring result is a ball out of bounds; if the fusion confidence level still falls within the judgment interval when the preset maximum number of iterations is reached, it is judged as insufficient evidence.
[0015] Secondly, this application provides an intelligent table tennis scoring system based on wireless touch control and a mobile terminal, comprising:
[0016] Data acquisition module: used to acquire audio and vibration signals, which are generated by a mobile terminal attached to the side of the table tennis table during the impact of the ping-pong ball on the table or ground.
[0017] Feature extraction module: used to calculate audio signal-to-noise ratio and vibration clarity based on audio and vibration signals, and to extract audio and vibration features.
[0018] Confidence generation module: used to calculate the first comprehensive support of audio features to table impact and the second comprehensive support of vibration features to table impact, assign weights to the first comprehensive support and the second comprehensive support and sum them up to obtain the fused confidence; determine the judgment interval based on audio signal-to-noise ratio and vibration clarity.
[0019] Iterative arbitration module: It is used to compare the fusion confidence score with the judgment interval. If the fusion confidence score falls into the judgment interval, the weights are adjusted according to the audio signal-to-noise ratio and vibration clarity, and the calculation of the fusion confidence score is returned until the preset maximum number of iterations is reached or the fusion confidence score falls outside the judgment interval. If the fusion confidence score falls outside the judgment interval, a scoring result is generated based on the fusion confidence score.
[0020] Thirdly, this application provides a readable storage medium storing computer program instructions, which are read and executed by a processor to perform the steps of an intelligent ping-pong scoring method based on wireless touch and a mobile terminal.
[0021] The beneficial effects of this application are:
[0022] This application synchronously collects audio and vibration signals via a mobile terminal, calculates signal quality indicators, extracts time-frequency domain features, and combines adaptive judgment intervals and iterative weight correction mechanisms to finally generate highly reliable scoring results. It effectively solves the problems of signal feature distortion caused by environmental interference, insufficient accuracy of impact type identification, and susceptibility to scoring accuracy in existing technologies. It realizes multi-source information complementarity enhancement, dynamic weight optimization allocation, and intelligent iterative decision-making, significantly improving signal anti-interference capability and feature discrimination, ensuring the accuracy of impact type identification and the reliability of scoring decisions, and effectively reducing misjudgments and omissions.
[0023] Other features and advantages of this application will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures pointed out in the description and the accompanying drawings. Attached Figure Description
[0024] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0025] Figure 1 A flowchart illustrating the intelligent table tennis scoring method based on wireless touch and mobile terminal of this application is shown.
[0026] Figure 2 A flowchart illustrating the calculation of the first and second comprehensive support scores is shown.
[0027] Figure 3 A flowchart illustrating the calculation of the upper and lower boundary values of the decision interval is shown.
[0028] Figure 4 A flowchart illustrating the process of adjusting the weights is shown. Detailed Implementation
[0029] To address the problems raised in the background technology, this application synchronously collects audio and vibration signals through a mobile terminal, calculates multidimensional support and performs weighted fusion, and combines adaptive judgment intervals and iterative weight correction mechanisms to finally generate highly reliable scoring results, which significantly improves the scoring accuracy and judgment reliability and reduces the occurrence of missed judgments and misjudgments.
[0030] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0031] In some embodiments, such as Figure 1 As shown, this application provides a smart table tennis scoring method based on wireless touch control and a mobile terminal, including:
[0032] S1. Acquire audio and vibration signals, which are generated by a mobile terminal attached to the side of the table during the impact of the ping-pong ball on the table or ground.
[0033] S2. Calculate the audio signal-to-noise ratio (SNR) and vibration clarity (C) based on the audio and vibration signals. The SNR reflects the quality of the audio signal, and the C reflects the quality of the vibration signal. Extract audio and vibration features from the audio and vibration signals, and use these features as evidence to analyze and determine whether the ping-pong ball hit the table surface or the ground.
[0034] S3. Calculate the first comprehensive support of audio features for table impact. The second comprehensive support of vibration characteristics for table impact First overall support The degree to which audio features support the conclusion that the ball hit the table was quantified; the second comprehensive support score... The degree to which the vibration characteristics supported the judgment that the ball struck the table surface was quantified.
[0035] First overall support Second overall support By assigning weights and performing a weighted summation, the fusion confidence score is obtained. fusion confidence The evidence combines both audio and vibrational data, and the magnitude and sign of the values represent the current overall judgment tendency.
[0036] The decision interval is determined based on the audio signal-to-noise ratio and vibration clarity. This decision interval is used to define the ambiguity zone for subsequent decisions. If confidence levels are fused... The value falls within the judgment range, indicating that the current fused evidence is still ambiguous and has not formed a clear judgment.
[0037] S4. Combine confidence levels Compared with the decision interval:
[0038] If the fusion confidence level If the value falls within the judgment range, the weights are adjusted based on the audio signal-to-noise ratio (SNR) and vibration clarity (C), and the calculation of the fusion confidence score is returned. The process continues until the preset maximum number of iterations or fusion confidence level is reached. It falls outside the judgment range.
[0039] If the fusion confidence level If it falls outside the judgment interval, then it is determined according to the fusion confidence level. The scoring results are generated by determining whether the ping-pong ball hits the table or lands directly. Excluding cases where the ball touches the net, if the first sound produced after the ping-pong ball is hit is determined to be an impact on the table, a score is awarded; otherwise, no score is awarded.
[0040] For example, by collecting audio signals through the terminal's microphone and vibration signals through the accelerometer, the calculated audio signal-to-noise ratio (SNR) is 6.0 dB and the vibration intelligibility (C) is 0.7. After extracting the audio and vibration features, the first overall support is calculated. The second overall support score is 0.35. The fusion confidence level is 0.34. The confidence level is 0.343. If the judgment interval is [-0.45, 0.45], the fusion confidence level is... If the value falls outside the decision interval, after multiple iterations, the fused confidence level is determined. The score is 0.46, which is outside the judgment range [-0.45, 0.45], so it is judged as a table collision and a score is calculated.
[0041] In some embodiments, calculating the audio signal-to-noise ratio (SNR) and vibration clarity C based on the audio signal and vibration signal includes:
[0042] The signal-to-noise ratio (SNR) is obtained by calculating the ratio of the signal power of the audio signal to the background noise power; the vibration clarity (C) is obtained by calculating the proportion of vibration energy that matches the characteristic frequency band of the ping-pong ball impact to the total vibration energy.
[0043] Specifically, for an audio signal containing a suspected impact event, short-time energy detection or zero-crossing rate analysis is used to pinpoint the short-time window in which the suspected impact occurred. In this short window Inside, calculate the average power of the audio signal. Select a background noise window during the silent period before or after the suspected collision event. Calculate the average power of the background noise. Then, the audio signal-to-noise ratio (SNR) is calculated based on the power ratio definition. The higher the SNR value, the clearer the audio signal and the less it is affected by environmental noise.
[0044] The acquired vibration signals were also located within a short time window aligned with the audio event time. The vibration signal within the window is subjected to a Fast Fourier Transform to obtain its spectrum. The characteristic frequency band where the vibration energy is mainly concentrated when the ping-pong ball hits the table or the ground is determined in advance. Calculate the characteristic frequency Vibrational energy within Calculate the total vibration energy of the entire analysis frequency band. Vibrational energy in characteristic frequency bands Total vibrational energy The ratio is the vibration clarity C.
[0045] For example, the audio signal power is 0.4. The background noise power is 0.1. The calculated audio signal-to-noise ratio (SNR) is approximately 6.0 dB, the vibration energy in the characteristic frequency band is 0.7 μJ, and the total vibration energy is 1.0 μJ. Therefore, the calculated vibration clarity C is 0.7.
[0046] In some embodiments, audio features and vibration features are extracted, including: audio features including high-frequency energy ratio, pulse width, and time difference between the two ends; vibration features including high-frequency vibration energy, attenuation constant, and waveform symmetry.
[0047] The high-frequency energy ratio is used to distinguish the material hardness of the impacting object; the pulse width is used to characterize the duration of the impact event; the time difference between the two ends is used to estimate the spatial location of the sound source and help determine whether the impact point is within the table tennis area. Specifically, when using two mobile terminals placed on opposite sides of the table tennis table, the time difference between the arrival of the same impact sound at the two terminal microphones is calculated. According to the time difference The distance difference d between the sound source and the two microphones can be calculated. v represents the speed of sound. Based on the distance difference d, we can determine that the sound source is located on a hyperbola with the two mobile terminal microphones as the focal points. Combined with the geometric dimensions of the table, we can determine whether the impact point is within the area of the table.
[0048] High-frequency vibration energy is used to assess the severity of the impact and its excitation intensity on the high-frequency modes of the table tennis structure; the attenuation constant is used to describe the rate at which vibration energy is dissipated in the table tennis structure, reflecting the energy transfer characteristics of the impact; waveform symmetry is used to analyze the regularity of the vibration waveform and to determine whether the impact event is direct and clean.
[0049] For example, calculating the energy proportion above 3000Hz from the audio spectrum, the high-frequency energy proportion is 0.75, the pulse width is 0.02s, and using both terminal A and terminal B, with the audio signal from terminal B only introduced when acquiring the time difference between the two ends, the time difference between the two ends is obtained as 0.001s. Extracting energy above 2000Hz from the vibration signal, the high-frequency vibration energy is 0.8, the attenuation constant is 0.04, and the waveform symmetry is 0.85.
[0050] In some embodiments, such as Figure 2 As shown, the first comprehensive support of audio features for table impact is calculated. The second comprehensive support of vibration characteristics for table impact ,include:
[0051] By collecting a large number of samples of ping-pong balls hitting the table surface and the ground, audio features including the proportion of high-frequency energy, pulse width, and time difference between the two ends, and vibration features including high-frequency vibration energy, attenuation constant, and waveform symmetry were extracted. Statistical averages were calculated for the features of each type of sample to form an audio feature template for table impact. Ground impact audio feature template Tabletop impact vibration characteristic template Ground impact vibration characteristic template .
[0052] The audio features are compared with a preset table impact audio feature template. Perform matching to obtain the first matching degree. The audio features are compared with a preset ground impact audio feature template. Perform matching to obtain the second matching degree. , the first matching degree Matching degree with the second The difference is used as the first overall support. .
[0053] The vibration characteristics are compared with a pre-set table impact vibration characteristic template. Perform matching to obtain the third matching degree. The vibration characteristics are compared with a preset ground impact vibration characteristic template. Perform matching to obtain the fourth matching degree. , the third matching degree Matching degree with the fourth The difference is used as the second overall support. .
[0054] Specifically, the high-frequency energy ratio, pulse width, and time difference between the two ends are combined into a three-dimensional vector. Calculate three-dimensional vectors Audio feature template of impact with tabletop The cosine similarity is used as the first matching degree. Calculate three-dimensional vectors Audio feature template of impact with tabletop Cosine similarity as the second matching degree .
[0055] If the first overall support A value greater than 0 indicates that the current audio features are closer to a table impact; if the first overall support score is greater than 0... A value less than 0 indicates that the current audio features are closer to a ground impact, representing the first overall support score. The size reflects the strength of the tendency.
[0056] Similarly, high-frequency vibration energy, attenuation constant, and waveform symmetry can be combined into a three-dimensional vector. Calculate the three-dimensional vectors respectively Vibration characteristic template of impact with table surface Cosine similarity as the third matching degree and calculating three-dimensional vectors Ground impact vibration characteristic template Cosine similarity as the fourth matching degree .
[0057] If the second overall support A value greater than 0 indicates that the current vibration characteristics are closer to those of a table impact; if the second comprehensive support... A value less than 0 indicates that the current vibration characteristics are closer to a ground impact, representing the second overall support level. The size reflects the strength of the tendency.
[0058] For example, the first matching degree The second matching degree is 0.8. If the score is 0.45, then the first overall support score is... The third matching degree is 0.35. The fourth matching degree is 0.82. The second overall support is 0.48. It is 0.34.
[0059] In some embodiments, the first overall support level Second overall support By assigning weights and performing a weighted summation, the fusion confidence score is obtained. ,include:
[0060] S3.1. Based on the audio signal-to-noise ratio (SNR) as the primary comprehensive support... Assign first weight Based on vibration clarity C as the second comprehensive support... Assigning a second weight First weight The second weight is positively correlated with the audio signal-to-noise ratio (SNR). It is positively correlated with vibration clarity C.
[0061] Specifically, the audio signal-to-noise ratio (SNR) and vibration clarity (C) are linearly normalized and mapped to the [0,1] interval to obtain the normalized audio signal-to-noise ratio. and normalized vibration clarity First weight Possible forms: Second weight Possible forms: .
[0062] S3.2. The first overall support With the first weight The product plus the second overall support With the second weight The product of these factors yields the fusion confidence level. , A value greater than 0 indicates that the evidence after fusing audio and vibration signals leans towards a table impact. A value less than 0 indicates that the evidence after combining audio and vibration signals tends to be related to ground impact.
[0063] For example, the normalized audio signal-to-noise ratio The normalized vibration clarity is 0.3. If the value is 0.7, then the first weight is calculated. The second weight is 0.3. The fusion confidence level is 0.7. It is 0.343.
[0064] In some embodiments, determining the judgment interval based on the audio signal-to-noise ratio (SNR) and vibration clarity (C) includes:
[0065] S3.3. Calculate the overall signal quality coefficient based on the audio signal-to-noise ratio and vibration intelligibility. Specifically, for the normalized audio signal-to-noise ratio... and normalized vibration clarity The arithmetic mean is calculated to obtain the overall signal quality coefficient Q. The closer the overall signal quality coefficient Q is to 1, the better the overall signal quality of the audio and vibration signals.
[0066] S3.4. Based on the preset basic threshold Combined with the overall signal quality coefficient Q, the upper boundary value of the decision interval is calculated. and lower boundary value .
[0067] In some embodiments, such as Figure 3 As shown, based on the preset basic threshold Combined with the overall signal quality coefficient Q, the upper boundary value of the decision interval is calculated. and lower boundary value ,include:
[0068] S3.4.1. Based on the first support Second support Calculate the broadband coefficient When the first overall support With the second overall support When the product is greater than zero, it represents the first level of support. Second support If both values are positive or both are negative, indicating that the audio and vibration evidence both suggest the same type of impact, the evidence is consistent, and a narrower judgment interval should be used to facilitate a clear ruling. Therefore, the width coefficient... The preset first coefficient Otherwise, it represents a conflict of evidence, and the width factor... The second preset coefficient Second coefficient Greater than the first coefficient For example, the first coefficient Take 0.5, the second coefficient Take 1.0.
[0069] S3.4.2. Based on the preset basic threshold Broadband coefficient The upper boundary value of the decision interval is calculated using the combined signal quality coefficient Q. and lower boundary value Reference formula: , , ;in, This represents the range adjustment amount.
[0070] when At that time, it was determined to be a tabletop impact. At that time, a ground impact was determined. When the fusion confidence level is in the fuzzy region, the evidence is insufficient.
[0071] For example, the calculated overall signal quality coefficient Q is 0.5, and the base threshold... The support level is 0.2, the highest among all support levels. Second support The product is 0.119, so the width coefficient β is taken as... , The calculated interval adjustment amount is 0.5. The upper boundary value of the decision interval is 0.25. The lower boundary value is 0.45. It is -0.45.
[0072] In some embodiments, such as Figure 4 As shown, the weights are adjusted based on the audio signal-to-noise ratio (SNR) and vibration clarity (C), including:
[0073] S4.1. Normalize the audio signal-to-noise ratio and vibration intelligibility and calculate the difference rate D. Difference rate The closer to 0, the closer the audio signal is to the vibration signal; the closer the difference rate D is to 1, the better the quality of one signal is compared to the other.
[0074] S4.2. Multiply the difference rate D by the preset adjustment coefficient k to obtain the weight adjustment range. Weight adjustment range Used to control the first weight in each iteration Second weight The adjustment step size, and the adjustment coefficient k is the weight adjustment range. The larger the scaling factor D, the more trustworthy the signal source should be.
[0075] S4.3. Compare the normalized audio signal-to-noise ratio and vibration clarity Size: If the audio signal-to-noise ratio Greater than vibration clarity Then the first weight Increase the weight adjustment range and the second weight Reduce weight adjustment range If the vibration clarity Greater than the audio signal-to-noise ratio Then the second weight Increase the weight adjustment range and the first weight Reduce weight adjustment range .
[0076] For example, the calculated difference rate D is approximately 0.883, the adjustment coefficient k is taken as 0.1, and the calculated weight adjustment range... The value is 0.0883, the audio signal-to-noise ratio (SNR) is 6.0, the vibration clarity (C) is 0.7, and the new first weight is... The new second weight is 0.3883. The value is 0.6117, and the iteration continues.
[0077] In some embodiments, generating a scoring result based on the fusion confidence level includes:
[0078] If the fusion confidence level Greater than the upper boundary of the decision interval If there is sufficient confidence that the ball has struck the table, it is considered a table impact, and the score is awarded as a valid ball within the table.
[0079] If the fusion confidence level At the lower boundary of the decision interval If there is sufficient confidence that the ball has hit the ground and is out of bounds, it is judged as a ground impact and the score is out of bounds.
[0080] If the preset maximum number of iterations is reached, the fusion confidence level If the evidence still falls within the judgment range, it is deemed insufficient.
[0081] For example, if the final fusion confidence level It is 0.46, which is greater than the upper boundary. If the number of valid balls in the table is reached, the score will be calculated. If the final fusion confidence level is... It is -0.50, which is less than the lower boundary. If the maximum number of iterations is reached, then the boundary ball is output. If the fusion confidence score is not reached, then the boundary ball is output. If the value is 0.44, it still falls within the judgment range, then the output is insufficient evidence.
[0082] In some embodiments, this application provides an intelligent table tennis scoring system based on wireless touch and a mobile terminal, including:
[0083] Data acquisition module: used to acquire audio and vibration signals, which are generated by a mobile terminal attached to the side of the table tennis table during the impact of the ping-pong ball on the table or ground.
[0084] Feature extraction module: used to calculate audio signal-to-noise ratio and vibration clarity based on audio and vibration signals, and to extract audio and vibration features.
[0085] Confidence generation module: used to calculate the first comprehensive support of audio features to table impact and the second comprehensive support of vibration features to table impact, assign weights to the first comprehensive support and the second comprehensive support and sum them up to obtain the fused confidence; determine the judgment interval based on audio signal-to-noise ratio and vibration clarity.
[0086] Iterative arbitration module: It is used to compare the fusion confidence score with the judgment interval. If the fusion confidence score falls into the judgment interval, the weights are adjusted according to the audio signal-to-noise ratio and vibration clarity, and the calculation of the fusion confidence score is returned until the preset maximum number of iterations is reached or the fusion confidence score falls outside the judgment interval. If the fusion confidence score falls outside the judgment interval, a scoring result is generated based on the fusion confidence score.
[0087] In some embodiments, this application provides a readable storage medium storing computer program instructions, which are read and executed by a processor to perform the steps of a smart table tennis scoring method based on wireless touch and a mobile terminal.
[0088] It should be noted that, in this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0089] Any references to memory, storage, database, or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory.
[0090] Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.
Claims
1. A smart table tennis scoring method based on wireless touch control and a mobile terminal, characterized in that, include: Acquire audio and vibration signals, which are generated by a mobile terminal attached to the side of the table tennis table during the impact of the ping-pong ball on the table or the ground. The audio signal-to-noise ratio and vibration clarity are calculated based on the audio signal and vibration signal, and the audio features and vibration features are extracted. Calculate the first comprehensive support of audio features for table impact and the second comprehensive support of vibration features for table impact, assign weights to the first comprehensive support and the second comprehensive support and sum them by weight to obtain the fusion confidence; determine the judgment interval based on audio signal-to-noise ratio and vibration clarity. The fusion confidence is compared with the decision interval: if the fusion confidence falls into the decision interval, the weights are corrected according to the audio signal-to-noise ratio and vibration clarity, and the step of calculating the fusion confidence is returned until the preset maximum number of iterations is reached or the fusion confidence falls outside the decision interval. If the fusion confidence score falls outside the judgment range, a scoring result is generated based on the fusion confidence score.
2. The method according to claim 1, characterized in that, Calculating audio signal-to-noise ratio and vibration intelligibility based on audio and vibration signals includes: The signal power of the audio signal is calculated as the ratio of the signal power to the background noise power to obtain the audio signal-to-noise ratio. The vibration clarity is obtained by calculating the proportion of vibration energy that matches the characteristic frequency band of the ping-pong ball impact to the total vibration energy.
3. The method according to claim 1, characterized in that, The audio features and vibration features are extracted, including: the audio features include the proportion of high-frequency energy, pulse width, and time difference between the two ends; the vibration features include high-frequency vibration energy, attenuation constant, and waveform symmetry.
4. The method according to claim 3, characterized in that, Calculate the first comprehensive support of audio features for table impact and the second comprehensive support of vibration features for table impact, including: The audio features are matched with a preset table impact audio feature template to obtain a first matching degree. The audio features are matched with a preset ground impact audio feature template to obtain a second matching degree. The difference between the first matching degree and the second matching degree is used as the first comprehensive support degree. The vibration feature is matched with a preset table impact vibration feature template to obtain a third matching degree. The vibration feature is matched with a preset ground impact vibration feature template to obtain a fourth matching degree. The difference between the third matching degree and the fourth matching degree is used as the second comprehensive support degree.
5. The method according to claim 1, characterized in that, Weights are assigned to the first and second comprehensive support scores, and the weighted sum is obtained to obtain the fusion confidence score, including: A first weight is assigned to the first overall support based on the audio signal-to-noise ratio, and a second weight is assigned to the second overall support based on the vibration intelligibility. The first weight is positively correlated with the audio signal-to-noise ratio, and the second weight is positively correlated with the vibration intelligibility. The product of the first comprehensive support and the first weight is added to the product of the second comprehensive support and the second weight to obtain the fusion confidence.
6. The method according to claim 1, characterized in that, The judgment range is determined based on the audio signal-to-noise ratio and vibration clarity, including: Calculate the overall signal quality coefficient based on the audio signal-to-noise ratio and vibration clarity. Based on the preset basic threshold and the comprehensive signal quality coefficient, the upper and lower boundary values of the determination interval are calculated.
7. The method according to claim 6, characterized in that, Based on the preset basic threshold and the comprehensive signal quality coefficient, the upper and lower boundary values of the judgment interval are calculated, including: The bandwidth coefficient is calculated based on the first support and the second support. If the product of the first overall support and the second overall support is greater than zero, the bandwidth coefficient is the preset first coefficient; otherwise, the bandwidth coefficient is the preset second coefficient, which is greater than the first coefficient. The upper and lower boundary values of the judgment interval are calculated based on the preset basic threshold, bandwidth coefficient, and comprehensive signal quality coefficient.
8. The method according to claim 5, characterized in that, The weights are adjusted based on the audio signal-to-noise ratio and vibration clarity, including: The audio signal-to-noise ratio and vibration clarity are normalized and the difference rate is calculated; Multiply the difference rate by a preset adjustment coefficient to obtain the weight adjustment range; Compare the normalized audio signal-to-noise ratio and vibration intelligibility: if the audio signal-to-noise ratio is greater than the vibration intelligibility, increase the weight of the first weight by the adjustment range and decrease the weight of the second weight by the adjustment range; if the vibration intelligibility is greater than the audio signal-to-noise ratio, increase the weight of the second weight by the adjustment range and decrease the weight of the first weight by the adjustment range.
9. The method according to claim 1, characterized in that, The scoring results are generated based on the fusion confidence level, including: If the fusion confidence level is greater than the upper boundary of the judgment interval, it is judged as a table collision, and the score is a valid ball inside the table and a score is awarded. If the fusion confidence level is less than the lower boundary of the judgment interval, it is judged as a ground impact and the score is out of bounds. If the fusion confidence level still falls within the determination interval when the preset maximum number of iterations is reached, it is determined that the evidence is insufficient.
10. A smart table tennis scoring system based on wireless touch control and a mobile terminal, characterized in that, include: Data acquisition module: used to acquire audio signals and vibration signals, which are generated by a mobile terminal attached to the side of the table tennis table during the process of the table tennis ball hitting the table or the ground; Feature extraction module: used to calculate the audio signal-to-noise ratio and vibration clarity based on the audio signal and vibration signal, and to extract audio features and vibration features; Confidence generation module: used to calculate the first comprehensive support of audio features to table impact and the second comprehensive support of vibration features to table impact, assign weights to the first comprehensive support and the second comprehensive support and sum them by weight to obtain the fused confidence; determine the judgment interval based on audio signal-to-noise ratio and vibration clarity; Iterative arbitration module: used to compare the fusion confidence with the decision interval. If the fusion confidence falls into the decision interval, the weights are corrected according to the audio signal-to-noise ratio and vibration clarity, and the calculation of fusion confidence is returned until the preset maximum number of iterations is reached or the fusion confidence falls outside the decision interval. If the fusion confidence score falls outside the judgment range, a scoring result is generated based on the fusion confidence score.