Give Eye: A Navigation Guidance System for Visually Impaired

Visually impaired individuals face significant challenges in navigating unknown and dynamic environments, as traditional tools like white canes or guide dogs provide limited assistance. Indoor navigation is particularly challenging because of the lack of GPS signals and the need for real-time obstacle avoidance. To overcome these challenges, this paper introduces an IoT-based navigation system that provides accurate, real-time guidance indoor for visually impaired users. The system proposed employs RSSI-based trilateration using ESP8266 modules and utilizes Dijkstra's algorithm in a grid environment to compute the shortest path within. A mechanism for haptic feedback, motor vibration, has been designed so that the calculated path can intuitively guide a user safely while keeping the required directions in sight. Real-time data transmission of the target device-central server via Socket.IO ensures real-time updates while the path recalculation happens dynamically. An object detection module integrated with an ESP32-CAM scans for nearby obstacles, providing adaptive responses to ensure user safety. This modular approach enhances navigation for visually impaired individuals by improving mobility and reducing errors in real-time obstacle avoidance. This cost-effective and modular approach gives greater potential to further improve the system with advanced computer vision algorithms, increasing the use of AI-based pathfinding that addresses critical mobility challenges for greater independence and safety in navigating, especially for visually impaired individuals.