Real-Time Obstacle Detection Using YOLOv8 on Raspberry Pi 4 for Visually Challenged People

Visual impairment poses significant challenges to individuals’ daily lives, particularly in navigating and comprehending their surroundings. In recent years, notable advancements in computer vision technology have facilitated the development of novel applications specifically designed to cater to the requirements of individuals with visual impairments. Developing a system that allows users to identify and locate obstructing objects in their immediate surroundings requires real-time and accurate object detection. Among the cutting-edge algorithms, You Only Look Once (YOLO) marks itself due to its remarkable efficiency and precision. This research article aims to implement YOLOv8, a state-of-the-art computer vision algorithm on Raspberry Pi 4, an affordable, portable, and small single-board computer. The system's ability to detect and recognise common indoor objects is evaluated through extensive testing with multiple real-case scenarios.