A People-Localization Method for Multi-Robot Systems First Approach for Guiding-Tours

Abstract

Throughout this article we present a methodology to localize multiple people in a group by a multi-robot system (MRS). The aim of the MRS is to conduct people through hallways in indoors as a guided-tour service task. However, further than guidance process, we detail a method for humans' localization by sharing distributed sensor data arising from the team of robots instrumented with stereo vision. The robustness of the method is presented, and by matching the real environment against the computed results, error in human localization is showed as well. As a first approach of the entire MRS goal, this paper explains from a task approach the way for environment ranging, spatial noise filtering, distributed sensor data fusion and clustering based segmentation. Likewise, through the paper experimental results are shown to verify the feasibility of the method.

Keywords

• multi-robot system
• multi-people localization
• vision based ranging
• distributed sensor data sharing.