Smart Wearable Device for Enhanced Mobility and Safety of the Visually Impaired Person

Visually impaired individuals encounter significant challenges in navigating unfamiliar environments, identifying obstacles, and accessing written information, which impact their safety and independence. This work presents a smart wearable device that integrates real-time object detection, text recognition, distance measurement, and an SOS alert system into a compact, affordable solution. The system is powered by a Raspberry Pi 4, equipped with a camera module, and provides audio feedback to the user. Yolov5n is employed for object detection, achieving $\boldsymbol{\sim} \mathbf{8 8 \%}$ accuracy under standard lighting and $\boldsymbol{\sim} \mathbf{9 2 \%}$ in optimal conditions, while PaddleOCR delivers $\sim 97 \%$ text recognition accuracy. Distance to nearby obstacles is estimated using a calibrated focal length method with an error margin of $\pm 5 \mathrm{~cm}$, and the SOS system ensures emergency communication within 3 seconds. Field testing demonstrated improved mobility, situational awareness, and safety among visually impaired users. This work contributes a multifunctional, accessible assistive device to promote independence and confidence.