Vision-Based Cooperative Localization for Small Networked Robot Teams

We describe the development of a vision-based cooperative localization and tracking framework for a team of small autonomous ground robots operating in an indoor environment. The objective is to enable a team of small mobile ground robots with limited sensing to explore, monitor, and search for objects of interest in regions in the workspace that may be inaccessible to larger mobile ground robots. In this work, we describe a vision-based cooperative localization and tracking framework for a team of small networked ground robots. We assume each robot is equipped with an LED-based identifier/marker, a color camera, wheel encoders, and wireless communication capabilities. Cooperative localization and tracking is achieved using the on-board cameras and local inter-agent communication. We describe our approach and present experimental and simulation results where a team of small ground vehicles cooperatively track other ground robots as they move around the workspace.