IOT - Based Smart Guide for Blind People with Obstacle & Object Detection

Visually impaired people have very limited environmental awareness and detect hazards with a significant delay, which seriously challenges their independent navigation. Traditional mobility aids, such as white canes, provide ground-level support only and cannot detect overhead or far-ahead obstacles. This paper proposes an AI and IoT-based Smart Guide system for enhancing mobility and safety by offering real-time object detection and audio feedback to its visually impaired users. This system integrates the camera module, Raspberry Pi, object detection model of YOLOv5, GPS module, and a text-to-speech engine to name the surrounding objects that it detects and sends voice information via headphones to the user. Transcribing such detected objects into voice alerts will help users understand their environment. Location data is simultaneously transmitted to a Flask-based web application, enabling caregivers to keep track of the user while receiving emergency notifications, if required. Stress is put on the low cost, portability, and modular design of the system; therefore, it is very practical for deployment in the real world. The tests in the lab proved that the system was capable of detecting objects correctly and the user was getting the voice prompt in time under different conditions. The Smart Guide is intended to increase the independence of visually impaired people, lower the possibility of accidents, and generally raise the quality of their life by merging AI, computer vision, and IoT technologies into one assistive platform.