GroundGazer: Camera-based indoor localization of mobile robots with millimeter accuracy at low cost

TL;DR

GroundGazer addresses the need for mm‑level indoor localization at low cost by combining a ground‑facing monocular fisheye camera, a chessboard floor, and an optional laser crosshair to estimate 2D position $(x,y)$ and heading $\theta$ with $mm$ precision and $<1^\circ$ accuracy. The method transforms the fisheye image to a top‑down view, detects crosshair and floor codes, localizes chessboard squares, and computes absolute pose using both local geometry and global references, enabling scalable, room‑size localization. Experiments on a low‑cost Raspberry Pi robot show mm‑scale position errors and sub‑degree heading with robust performance, while remaining portable and easy to set up. The approach promises a practical reference system for evaluating other indoor localization methods and can be extended to $3D$ with acknowledged trade‑offs, broadening applicability to robot swarms and larger environments.

Abstract

Highly accurate indoor localization systems with mm positioning accuracy are currently very expensive. They include laser trackers, total stations, and motion capture systems relying on multiple high-end cameras. In this work, we introduce a high-accuracy, planar indoor localization system named GroundGazer (GG) for autonomous mobile robots (AMRs). GG estimates the AMR's position with mm and its heading with sub-degree accuracy. The system requires only a monocular (fisheye) camera, a chessboard floor, and an optional laser diode. Our system is simple and low-cost, easy to set up, portable, robust, scalable to large areas and robot swarms, and extendable to 3D position and orientation estimation.

Figures (6)

• Figure 1: Overview of our GG localization system in a) and conceptual depiction of the localization method in b).
• Figure 2: Transformation of an original image recorded with an heading slightly below $90^\circ$ in a) through undistortion in b) and perspective transformation in c).
• Figure 3: Crosshair detection process.
• Figure 4: Square detection process in the perspective transformed top-down image.
• Figure 5: Trajectory evaluation