Errors and mistakes in automated player tracking
This paper examines errors, which can affect the accuracy of computer vision based people tracker. After a successful development of an automated player tracking system for use in team sports, the set of experiments was designed to investigate its accuracy. The authors take advantage of the controlled environment they use to obtain the "ground truth" information. This information is used to measure the accuracy of the tracking system. The results obtained are analyzed, and conclusions, including specification of the overall tracker accuracy, are given.
© Copyright 2001 Proceedings of Sixth Computer Vision Winter Workshop, Bled, Slovenia, February 7-9,. All rights reserved.
| Subjects: | |
|---|---|
| Notations: | technical and natural sciences sport games |
| Published in: | Proceedings of Sixth Computer Vision Winter Workshop, Bled, Slovenia, February 7-9, |
| Language: | English |
| Published: |
Bled
2001
|
| Online Access: | http://vision.fe.uni-lj.si/docs/janezp/Pers-wwk2001.pdf |
| Pages: | 25-36 |
| Document types: | congress proceedings |
| Level: | advanced |