Estimation of Court Boundary and Showing of Player Tra- jectory using a Broadcast Handball Game Video
DOI:
https://doi.org/10.52731/liir.v004.172Keywords:
Computer Vision, Handball, Sports Analysis, Court Boundary Estimation, Player Trajectory ShowingAbstract
This study proposes a method for estimating court boundary for player tracking and show- ing of player trajectory for visualization in a broadcast handball video. The video captured by a camera shows approximately one-third of the court, and shifts from left to right to fol- low the players. Therefore, the camera direction is estimated for the player trajectory with a taken goal object position in the video, and the goal object is detected with the object de- tection algorithm YOLOv5. The court boundary is estimated in two ways: one is by using the goal object detection result, and the other is by using a feature of a red painted area on the court. The player is detected and classified to own team by YOLOv5 and uniform color, and its trajectory is shown on a top view of the court. We experimented with three videos to confirm the accuracy of the two proposed methods of the court boundary estimation. The results indicates that the estimation rates of the two methods were 59.2 % and 73.8 %, respectively.
References
Data Volley: “https://www.dataproject.com” (confirmed in Jul. 2023)
SPLYZA Teams: “https://products.splyza.com/teams” (confirmed in Mar. 2023)
M. Perse, M. Kristan, S. Kovacic, G. Vuckovic, and J. Pers, ”A Trajectory-based Analy-sis of Coordinated Team Activity in a Basketball Ggame,” Computer Vision and Image Understanding, vol. 113, 2009, pp. 612–621
G. Hasegawa, and T, Takahashis, ”Shoot Motion Analysis in Basketball Formation Practice,” Proc. of the SPIE, vol. 11049, 2019, doi:10.1117/12.2521673.
P. Rudelsdorfer, N. Schrapf, H. Possegger, T. Mauthner, H. Bischof, and M. Tilp, ”A Novel Method for the Analysis of Sequential Actions in Team Handball,” Int’l J. of Computer Science in Sport, vol. 13, 2014, pp. 69–84.
R. M. L. Barros, R. P. Menezes, T. G. Russomanno, M. S. Misuta, B. C. Brandao, P.
J. Figueroa, N. J. Leite, and S.K. Goldenstein, ”Measuring Handball Players Trajecto-ries using an Automatically Trained Boosting Algorithm,” Comput Methods Biomech Biomed Engin, vol. 14, 2011, pp. 53–63.
X. Nie, S. Chen, and R. Hamid: “A Robust and Efficient Framework for Sports-Field Registration”, IEEE Workshop on Applications of Computer Vision (WACV), 2021, pp. 1936–1944.
Ultralytics: YOLOv5, https://github.com/ultralytics/yolov5, (confirmed in Jul. 2023)
Nwojke: DeepSort, https://github.com/nwojke/deep_sort, (confirmed in Jul. 2023)
