Research on pattern recognition and sports performance of table tennis game strategy based on big data mining technology
21 mar 2025
INFORMAZIONI SU QUESTO ARTICOLO
Pubblicato online: 21 mar 2025
Ricevuto: 30 ott 2024
Accettato: 30 gen 2025
DOI: https://doi.org/10.2478/amns-2025-0694
Parole chiave
© 2025 Ku Duan, published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Figure 1.
Figure 2.
Figure 3.
Figure 4.
Analysis of the specific route characteristics of the variation line
| Number | Variant line combination form | frequency | score | Lose point | Usage rate(%) | Scoring rate(%) | Contribution rate(%) |
|---|---|---|---|---|---|---|---|
| 1 | Reverse skew - anyway straight | 150 | 105 | 45 | 16.59% | 70.00% | 11.62% |
| 2 | Middle reverse oblique - reverse straight | 23 | 16 | 7 | 2.54% | 69.57% | 1.77% |
| 3 | Anti-center oblique – center Positive and oblique | 58 | 45 | 13 | 6.42% | 77.59% | 4.98% |
| 4 | Medium straight - medium positive oblique | 23 | 18 | 5 | 2.54% | 78.26% | 1.99% |
| 5 | Anti-integrity – right Positive and oblique | 134 | 90 | 44 | 14.82% | 67.16% | 9.96% |
| 6 | | medium oblique – Positive and oblique | 49 | 40 | 9 | 5.42% | 81.63% | 4.42% |
| 7 | Anticlinal. - Anticlinal | 36 | 30 | 6 | 3.98% | 83.33% | 3.32% |
| 8 | Straight forward and backward mesoblinia | 7 | 6 | 1 | 0.77% | 85.71% | 0.66% |
| 9 | · Anti-center oblique - center straight | 15 | 13 | 2 | 1.66% | 86.67% | 1.44% |
| 10 | Median oblique - median straight | 9 | 7 | 2 | 1.00% | 77.78% | 0.77% |
| 11 | Positive oblique - median oblique | 25 | 22 | 3 | 2.77% | 88.00% | 2.43% |
| 12 | Counterup rightness - median oblique | 17 | 17 | 0 | 1.88% | 100.00% | 1.88% |
| 13 | Positive oblique - positive straight | 65 | 53 | 12 | 7.19% | 81.54% | 5.86% |
| 14 | Medium oblique - straight forward and backward | 25 | 20 | 5 | 2.77% | 80.00% | 2.21% |
| 15 | Straight forward and backward | 112 | 80 | 32 | 12.39% | 71.43% | 8.85% |
| 16 | Middle antiskew - antiskew | 83 | 50 | 33 | 9.18% | 60.24% | 5.53% |
| 17 | Median straight - median anticlinal | 38 | 25 | 13 | 4.20% | 65.79% | 2.77% |
| 18 | Median oblique - median anticline | 35 | 24 | 11 | 3.87% | 68.57% | 2.65% |
The application of the variation and the non-linear strategy
| / | Frequency | Score | Lose point | Scoring rate | Usage rate | Contribution rate |
|---|---|---|---|---|---|---|
| Overall data | 6154 | 3077 | 3077 | 50.00% | 100.00% | 50.00% |
| Variation line | 2463 | 1536 | 927 | 62.36% | 40.02% | 24.96% |
| Invariant line | 3691 | 1625 | 2066 | 44.03% | 59.98% | 26.41% |
| T | 21.756 | 14.564 | 1.485 | |||
| P | 0.000 | 0.000 | 0.135 |
Various forms of variation
| / | Frequency | Score | Lose point | Scoring rate | Usage rate | Contribution rate |
|---|---|---|---|---|---|---|
| Variation line (total) | 2463 | 1536 | 927 | 62.36% | 100.00% | 50.00% |
| Meet... Intermediate level | 615 | 452 | 163 | 73.50% | 24.97% | 18.35% |
| Meet... Become forehand | 926 | 592 | 334 | 63.93% | 37.60%** | 24.04%** |
| Meet... Variable backhand bit | 922 | 452 | 470 | 49.02% | 37.43% | 18.35% |
| F | 0.785 | 21.264 | 11.265 | |||
| P | 0.422 | 0.000 | 0.000 |
Comparison of table tennis positioning and tracking method
| Method | Accuracy rate | Single frame detection time /ms | Model size /Mibyte |
|---|---|---|---|
| Color detection | 55.56% | 8.15 | 0.2 |
| Contour detection | 64.22% | 9.23 | 0.2 |
| YOLOv4 | 97.26% | 27.01 | 244.66 |
| YOLOv5s | 96.18% | 15.21 | 13.75 |
| Faster regional convolution neural network | 75.12% | 198.25 | 113.48 |
| YOLOv8 | 98.39% | 15.23 | 13.24 |
| Method | Central position pixel error | Detection speed /(frame • |
Model size /Mibyte |
| YOLOv4 | 14 pixels | 43 | 244.66 |
| YOLOv5s | 36 pixels | 75 | 13.75 |
| Faster regional convolution neural network | - | 3 | 113.48 |
| YOLOv8 | 11 pixels | 64 | 13.24 |