Multi-view table tennis video sequence for ball detection purposes

<p dir="ltr">This video database contains 3 multi-view table tennis sequences which were captured during local league matches. The sequences were filmed using up to 4 high speed cameras (Casio EX-F1). The original purpose of these sequences is for ball tracking research. A ground truth, which contains the ball location in each frame identified by human volunteers, is also available for some of the sequences. These sequences can be used for non-commercial research purposes only.</p><p dir="ltr">If you use these data files for your research, please cite the following paper of ours:</p><p dir="ltr"><a href="http://oro.open.ac.uk/view/person/kcpw2.html" target="_blank">Wong, Patrick</a>; <a href="http://oro.open.ac.uk/view/person/hom27.html" target="_blank">Myint, Hnin</a>; <a href="http://oro.open.ac.uk/view/person/lsd3.html" target="_blank">Dooley, Laurence</a> and <a href="http://oro.open.ac.uk/view/person/aah3.html" target="_blank">Hopgood, Adrian</a> (2023).A multi-view automatic table tennis umpiring framework. <i>Proceedings of the Institution of Mechanical Engineers, Part P: Journal of Sports Engineering and Technology</i> (early access).</p><p dir="ltr">DOI: <a href="https://doi.org/10.1177/17543371231171433" target="_blank">https://doi.org/10.1177/17543371231171433</a></p><p><br></p>