Computer vision-based gate crossing detection for timing and analysis of ski and snowboard races
- Author
- Robbe Decorte (UGent) , Jelle De Bock (UGent) , Maarten Slembrouck (UGent) and Steven Verstockt (UGent)
- Organization
- Abstract
- This paper proposes an algorithm that detects skiers and flags during a ski slalom race, by using object detection on video footage. The algorithm uses the detection of skiers and flags to implement gate-to-gate timing and analysis. The result of the detection and gate timing is used to analyze metrics such as the total time of the skier's race, the time between every two consecutive flags and anomalies in the skier's trajectory. Furthermore, the data of all skiers competing in the slalom race can be used to create a heatmap of the trajectory and to calculate average statistics. To detect the objects, a custom-trained YOLO model is used. Training is done on a custom created dataset, containing images from various slalom races together with footage recorded on an indoor ski slope. Compared to the current equipment for gate timing using infrared photocells, video based timing is less accurate, as the algorithm is on the one hand limited by the frames per second of a video and on the other hand by the accuracy of the object detection. However, for recreational and training purposes, it delivers accuracy within a reasonable margin and can provide quick feedback.
Downloads
-
fears 2022 poster.pdf
- full text (Author's original)
- |
- open access
- |
- |
- 623.91 KB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-01GMVB1ARQ7FP9RAGF3FS242ZZ
- MLA
- Decorte, Robbe, et al. “Computer Vision-Based Gate Crossing Detection for Timing and Analysis of Ski and Snowboard Races.” Faculty of Engineering and Architecture Research Symposium 2022 (FEARS 2022), Abstracts, 2022, doi:10.5281/zenodo.7405854.
- APA
- Decorte, R., De Bock, J., Slembrouck, M., & Verstockt, S. (2022). Computer vision-based gate crossing detection for timing and analysis of ski and snowboard races. Faculty of Engineering and Architecture Research Symposium 2022 (FEARS 2022), Abstracts. Presented at the Faculty of Engineering and Architecture Research Symposium 2022 (FEARS 2022), Ghent, Belgium. https://doi.org/10.5281/zenodo.7405854
- Chicago author-date
- Decorte, Robbe, Jelle De Bock, Maarten Slembrouck, and Steven Verstockt. 2022. “Computer Vision-Based Gate Crossing Detection for Timing and Analysis of Ski and Snowboard Races.” In Faculty of Engineering and Architecture Research Symposium 2022 (FEARS 2022), Abstracts. https://doi.org/10.5281/zenodo.7405854.
- Chicago author-date (all authors)
- Decorte, Robbe, Jelle De Bock, Maarten Slembrouck, and Steven Verstockt. 2022. “Computer Vision-Based Gate Crossing Detection for Timing and Analysis of Ski and Snowboard Races.” In Faculty of Engineering and Architecture Research Symposium 2022 (FEARS 2022), Abstracts. doi:10.5281/zenodo.7405854.
- Vancouver
- 1.Decorte R, De Bock J, Slembrouck M, Verstockt S. Computer vision-based gate crossing detection for timing and analysis of ski and snowboard races. In: Faculty of Engineering and Architecture Research Symposium 2022 (FEARS 2022), Abstracts. 2022.
- IEEE
- [1]R. Decorte, J. De Bock, M. Slembrouck, and S. Verstockt, “Computer vision-based gate crossing detection for timing and analysis of ski and snowboard races,” in Faculty of Engineering and Architecture Research Symposium 2022 (FEARS 2022), Abstracts, Ghent, Belgium, 2022.
@inproceedings{01GMVB1ARQ7FP9RAGF3FS242ZZ,
abstract = {{This paper proposes an algorithm that detects skiers and flags during a ski slalom race, by using object detection on video footage. The algorithm uses the detection of skiers and flags to implement gate-to-gate timing and analysis. The result of the detection and gate timing is used to analyze metrics such as the total time of the skier's race, the time between every two consecutive flags and anomalies in the skier's trajectory. Furthermore, the data of all skiers competing in the slalom race can be used to create a heatmap of the trajectory and to calculate average statistics. To detect the objects, a custom-trained YOLO model is used. Training is done on a custom created dataset, containing images from various slalom races together with footage recorded on an indoor ski slope. Compared to the current equipment for gate timing using infrared photocells, video based timing is less accurate, as the algorithm is on the one hand limited by the frames per second of a video and on the other hand by the accuracy of the object detection. However, for recreational and training purposes, it delivers accuracy within a reasonable margin and can provide quick feedback.}},
author = {{Decorte, Robbe and De Bock, Jelle and Slembrouck, Maarten and Verstockt, Steven}},
booktitle = {{Faculty of Engineering and Architecture Research Symposium 2022 (FEARS 2022), Abstracts}},
language = {{eng}},
location = {{Ghent, Belgium}},
pages = {{1}},
title = {{Computer vision-based gate crossing detection for timing and analysis of ski and snowboard races}},
url = {{http://doi.org/10.5281/zenodo.7405854}},
year = {{2022}},
}
- Altmetric
- View in Altmetric