A Route Planning and Localization System Integrated into A Computer Vision System to Assist Navigation for the Visually Impaired

In daily life, visually impaired individuals face numerous challenges in independently navigating and orienting themselves in real-world environments. To enhance their autonomy, this paper proposes a real-time route planning and localization system that enables users to interact via voice to select destinations and receive audio-based navigation instructions. The system comprises three main components: (1) voice-based address input, which is converted into geographic coordinates (longitude and latitude) using a geocoding service; (2) a Global Navigation Satellite System (GNSS) module that provides high-accuracy real-time positioning; (3) a Dijkstra shortest path algorithm implemented on OpenStreetMap (OSM) to determine the optimal route to the desired destination. Navigation information is converted into audio signals and played through a speaker to guide users along the correct path. The system is designed to be user-friendly, low-cost, and suitable for outdoor navigation applications for the visually impaired.