A distributed vision-based infrastructure for multi-robot navigation

Abstract

With an accurate estimate of the robot pose, robots will be able to navigate in their environment autonomously with the aid of dynamic path planning. In this paper, an infrastructure of Distributed Vision-Based System (DVS) for multi-robot navigation is presented. The infrastructure is comprised of Localization and Motion Planning Process. The Localization Process receives images from cameras mounted in the environment and then utilizes ID tags to localize multiple robots. Motion Planning Process is designed to generate paths and steer the robots to target positions. Experiments with a network of small robot platforms and overhead cameras show that the proposed infrastructure could simultaneously localize and navigate multiple robots with localization errors within 0.1 meters.