Home /Research /Robust Monocular Camera-Based Localization in Industrial Environments
OTHER

Robust Monocular Camera-Based Localization in Industrial Environments

Yungu Won, Yein Choi, Hwangryol Ryu, Youngmin Moon, Sung Soo Hwang

Year
2024
Citations
2

Abstract

This paper proposes a camera-based system for robust localization of workers or robots in large-scale factories and industrial sites. The system is designed with the convenience and efficiency of workers in mind, providing stable tracking capabilities. To achieve this, a monocular camera is attached to a safety helmet, enabling localization without restricting the worker's range of motion. Additionally, by combining Visual Simultaneous Localization and Mapping and Visual Place Recognition, the system tracks the location of workers or robots while minimizing tracking loss, ensuring consistent localization. Our methodology has enhanced the accuracy and reliability of position estimation by approximately 46.0%. By accurately identifying the location of workers or robots, the system aims to prevent accidents and ensure swift responses in emergency situations.

Keywords

Computer visionComputer scienceArtificial intelligenceMonocularRobustness (evolution)Computer graphics (images)

Related papers

Browse all OTHER papers