Trajectory segmentation using dynamic programming
Richard Mann, Allan D. Jepson, Thomas El-Maraghi · 2003
We consider the segmentation of a trajectory into piecewise polynomial parts, or possibly other forms. Segmentation is typically formulated as an optimization problem which trades off model fitting error versus the cost of introducing new segments. Heuristics such as split-and-merge are used to find the best segmentation. We show that for ordered data (e.g., single curves or trajectories) the global optimum segmentation can be found by dynamic programming. The approach is easily extended to handle different segment types and top down information about segment boundaries, when available. We show segmentation results for video sequences of a basketball undergoing gravitational and nongravitational motion.