Projections on moving objects have a problem in that the projection may slip because of the delay between tracking and projection. Here, we propose a new prediction method combining a Kalman filter and a three-frame feedback model that switches between these models according to the ball’s state of motion. We developed a real-time tracking and projection system named “Ballumiere”, which uses motion capture cameras for tracking and multiple projectors for spherical projection. We conducted a comparative experiment with an existing prediction model and showed that our method minimizes slipping and increases the accuracy of the projection.
Shio Miyafuji, Masato Sugasaki, and Hideki Koike. 2016. Ballumiere: Real-Time Tracking and Spherical Projection for High-Speed Moving Balls. In Proceedings of the 2016 ACM International Conference on Interactive Surfaces and Spaces (ISS ‘16). Association for Computing Machinery, New York, NY, USA, 33–37. https://doi.org/10.1145/2992154.2992181