Abstract
Taking an inspiration from psychological studies of visual attention, the contribution of this paper lies in prediction of the critical points of the trajectory using the structure of a scene and physical motion model. On one side, we present our approach for video analysis that differs from traditional tracking techniques by predicting future states of the moving object rather than its next consecutive position using the physically-based motion functionality. On the other side, we propose to use the structure of the scene, which contains the information about the obstacles and space limits, for discovering the critical points of the trajectory. As a proof of concept we developed the use case application for analysing snooker footage.
Reference
Gabdulkhakova, A., & Kropatsch, W. G. (2014). Video analysis of a snooker footage based on a kinematic model. In Structural, Syntactic, and Statistical Pattern Recognition (pp. 223–232). Springer. https://doi.org/10.1007/978-3-662-44415-3_23