Design and Implementation of an AI-Based Assistive Smart Glass System for Visually Impaired with Real-Time Object Detection and Emergency Alert

DOI

No Doi

Published In

Year 2026 · Volume 3 · Issue 1

Published Online

January-April 2026

Pages

42-48

Article Type

Original Article

Check for updates

Visual impairment is a major global challenge that restricts independent mobility, environmental awareness, and personal safety. Traditional assistive devices provide limited support and lack intelligent perception and emergency handling capabilities. This work presents a vision-based assistive safety system developed using Raspberry Pi to enhance independence and safety for visually impaired individuals. This work proposes a vision-driven assistive system designed to improve mobility and safety for visually impaired individuals. The system combines real-time face recognition using the LBPH algorithm with object detection based on the YOLOv3-Tiny deep learning model to identify people and surrounding objects. Audio feedback is provided through a text-to-speech engine for user guidance. A GPS module continuously tracks live location, which is displayed through a web interface, while a touch-based emergency mechanism sends alert emails with a Google Maps location link to a caretaker. The system operates in real time using multithreading and demonstrates reliable performance in assistive and emergency scenarios, making it suitable for real-world applications.