Combination of Adaptive Object Model for Basketball Tracking

Abstract

Theoretically it is certified that basketball tracking question belongs to the NP-Hard question. In this paper, by combining the bipartite graph of adaptive object model, the basketball tracking is formed, and by this way the calculation complexity can be reduced effectively, thus the combination of adaptive object model for basketball tracking (AOMBT) is put forward. After introducing the adaptive object model, the calculation can be obtained at the first time when conducting the basketball tracking; and the calculation complexity can be reduced by the incremental coverage method. Lastly, through the simulation experiment it shows that the method proposed in this paper represents relatively high detection degree in the basketball tracking, meanwhile the tracking error is kept in a lower level, and as a whole, even though some questions, such as false, unobservable etc., exist partly, this algorithm still has relatively strong advantage in the aspect of detection rate of basketball tracking, and can adapt large-scale basketball tracking.