Mobile-based indoor navigation application with object detection for visually impaired

Kuek, Levin (2026) Mobile-based indoor navigation application with object detection for visually impaired. Final Year Project, UTAR.

PDF Download (2793Kb)

Abstract

Currently, visually impaired individuals face significant challenges in achieving safe indoor navigation. Existing solutions are often impractical and not cost-effective. Therefore, this project proposes an indoor navigation mobile application to enhance the mobility and independence of visually impaired users. The system uses pedestrian dead reckoning (PDR) to estimate user movement and position. The YOLO model further enhances the system's capability for obstacle detection and object recognition. The integration of distance estimation into the YOLO model enables users to identify and avoid hazards more effectively, thereby enhancing navigation. By using the A* pathfinding algorithm, the system computes optimal routes and provides real-time guidance through auditory feedback. This solution aims to offer reliable, accessible, and user-friendly navigation assistance, improving safety and autonomy for visually impaired users in indoor environments.

Actions (login required)