Cooperation between two omnidirectional perception systems for mobile robot localization

In this paper, an absolute localization paradigm based on the cooperation of an omnidirectional vision system composed of a conical mirror and a CCD camera and a low cost panoramic range finder system is reported. These two sensors, which have been used independently until now, provide some complementary data. This association enables us to build a robust sensorial model which integrates an important number of significant primitives. We can thus realize an absolute localization of the mobile robot in particular configurations, like symmetric environments, where it is not possible to determine the position with the use of only one of the two sensors. In a first part, we present our global perception system. In a second part, we describe our sensorial model building approach. Finally we present an absolute localization method which uses three matching criteria fused thanks to the combination rules of the Dempster-Shafer theory. The basic probability assignment got for each primitive matching enables to estimate the reliability of the localization. We test our global absolute localization system on several robots' elementary moves in an indoor and symmetric environment.