Precise localization of indoor mobile robots in FMS based on distributed vision

Abstract

The knowledge of the pose and the orientation of mobile robot in its operating environment is of utmost importance for an autonomous robot. In this paper, we present a new robot localization method integrates distributed vision sensors with Monte Carlo localization (MCL) method. Firstly, an improved MCL method is used to estimates the posterior distribution of robot poses conditioned on sensor data. Secondly, the system uses distributed vision sensors to extract the feature of each running robotpsilas mark, and then use vision algorithm to determine each robotpsilas pose and orientation. Finally, Monte Carlo method is extended to integrate the detection information coming from vision sensors to localize robot. The result obtained in simulation and with real robots show that the method can measure the pose and orientation of each robot in the system accurately. Meanwhile, the reliability, real-time property and robustness of the system have been validated.