StairGuide: A Staircase Detection and Guidance System for Visually Impaired People using ML

Visually impaired individuals face significant challenges when navigating environments with staircases, often resulting in accidents and decreased mobility. Traditional aids such as white canes and guide dogs are limited in their ability to detect staircases in real time, reducing independence and safety. This study presents StairGuide, a real-time staircase detection and guidance system aimed at improving the mobility and safety of visually impaired individuals. The system combines advanced image processing techniques and machine learning algorithms to detect staircases and obstacles while providing real-time auditory feedback to guide users. Using a wide-angle camera, the system captures the environment and processes the images with algorithms such as edge detection and the Hough Transform to identify staircases. The YOLOv3 object detection model is employed to detect nearby individuals, ensuring safe navigation. The system’s effectiveness was evaluated in various environments, demonstrating high accuracy and real-time performance. Our findings highlight the potential of integrating machine learning and image processing to create affordable and accessible navigation aids for visually impaired users.