Mining leader-follower realtionships from GPS trajectories

alt text 

The leader-follower phenomenon is natural for many mobile objects acting collectively. It is possible to find the leader of such collective motion from trajectory data and use it to understand the motion pattern of the group. This work proposes a technique for detecting leader-follower relationships from trajectories of mobile objects, and represents them as a graph sequence. We apply this method to analyze the leader-follower behavior of a herd of cows based on their GPS traces.

Publication

Y. Li, and L. Guibas, Leader-Follower Relationships from Trajectory Data – a Case Study, Computational Geometry: Young Researchers Forum (CG:YRF ’12), Chapter Hill, NC, U.S.A, June 16-20, 2012 pdf ppt
Bibtex
@inproceedings{lg-lfrtd-12,
  title = "Leader-Follower Relationships from Trajectory Data -- a Case Study",
  author = "Y. Li,  L. Guibas",
  booktitle = "Computational Geometry: Young Researchers Forum (CG:YRF `12)",
  year = 2012
 }

Project list