Multi-Agent Fusion With Different Limited Fields-of-View
Bailu Wang, Suqi Li, Giorgio Battistelli, Luigi Chisci, Wei Yi
Abstract
A key objective of multi-agent surveillance systems is to monitor a much larger region than the limited <i>field-of-view</i> (FoV) of any individual agent by successfully exploiting cooperation among multiple agents. Whenever either a centralized or a distributed approach is pursued, this goal cannot be achieved unless an appropriately designed fusion strategy is adopted. This paper presents a novel information fusion approach by considering for each agent a known limited, and possibly different, FoV. The proposed method, named <i>Bayesian-operation InvaRiance on Difference-sets</i> (BIRD) fusion, relies on <i>Generalized Covariance Intersection</i> (GCI) and exploits a general and exact decomposition of each multi-object posterior by partitioning the global FoV, i.e. the union of the FoVs of the fusing agents, into common and exclusive FoVs. It is shown how BIRD fusion can be used to perform multi-object estimation based on random finite sets on both a centralized and a distributed peer-to-peer sensor network. Simulation experiments on realistic multi-object tracking scenarios demonstrate the effectiveness of BIRD fusion.