Distributed tracking in a large-scale network of smart cameras
Honggab Kim, Marilyn C. Wolf · 2010
This paper describes a new distributed algorithm for track-ing in distributed camera networks. This algorithm oper-ates without a centralized server that collects all the mea-surements over the entire network. With the observations sent from its neighbors and the local probabilistic transition model, each camera independently estimates local paths in its neighborhood. The conflicts on locally estimated paths among cameras are resolved by a voting algorithm, and the agreed local paths are finally combined into global paths. Our experiments with simulated data demonstrate that the proposed distributed tracking algorithm is fast and scalable without degrading tracking accuracy.