Simultaneous localization and mapping of a mobile robot with surveillance and flying cameras

Localization and exploration of robotic systems has been a highly active research area in recent years. In order to explore unknown environments with better precision, a robotic system employing surveillance and flying cameras is proposed in this paper. Specifically, a mobile robot equipped with an onboard monocular camera performs visual simultaneous localization and mapping with extended Kalman filtering (EKF-VSLAM) and moves from indoor to outdoor environments. To correct the estimation errors of the system during the task, a least squares-based correction approach is proposed. The proposed robotic system employs surveillance cameras in indoor environment and a flying camera in outdoor environment to correct estimation error from the onboard monocular camera. The proposed correction approach, integrated with the conventional EKF-VSLAM, has been validated by experiments demonstrating better precision.

1. Introduction

New technology allowing the detection in new environments is being developed for the purpose of robotic mapping and navigation. Many methods of localization and mapping have been proposed and employed in robot systems for exploring unknown area. However, each environment has its own different complexities and details which is challenging in developing the systems. In either indoor or outdoor environments, there should be different devices located in each environment that are suitable to fulfil specifications and parameters of the place. This will be challenging for development of robotic systems.