Real-time indoor localization with visual SLAM for in-building emergency response
Journal
Automation in Construction
Journal Volume
140
Date Issued
2022
Author(s)
Abstract
Localization is critical to but also challenging for an efficient first response during indoor emergency situations. This paper describes an economical indoor localization system with a low setup cost based on the visual Simultaneous Localization and Mapping (vSLAM). In addition, threshold for key features per image frame is studied for acceptable localization accuracy. The amount of image input users should collect is also analyzed for reliable localization. To address matching and accuracy issues under varying lighting conditions, a comprehensive approach for accurate localization utilizing vSLAM is also proposed. An actual building was tested for the proposed approach with an average localization success rate over 94%. © 2022 Elsevier B.V.
Subjects
Automated external defibrillator (AED); Emergency medical service (EMS); Indoor localization; Simultaneous localization and mapping (SLAM)
SDGs
Other Subjects
Indoor positioning systems; Mapping; Robotics; Automated external defibrillator; Emergency medical service; Emergency medical services; Indoor localization; Localisation; Real- time; Simultaneous localization and mapping; Visual simultaneous localization and mappings; Emergency services
Type
journal article