Omnidirectional sensors cooperation for multi-target tracking

In this paper, we present a multi-sensor cooperation paradigm between an omnidirectional vision system and a panoramic range finder system, to localize dynamically a mobile robot. These two sensors provide some complementary data which enable us to build a robust sensorial model which integrates an important number of significant primitives. Based on this model we treat the problem of maintaining a matching and propagating uncertainties on each matched primitive. Then we developed a localization method based on the tracking of significant landmarks of the environment. Our tracking method integrates a multi-level uncertainty propagation stage based on the use of the Dempster-Shafer theory.