DESIGN AND IMPLEMENTATION OF AN OBJECT DETECTION ALGORITHM FOR VISUALLY IMPAIRED PEOPLE (A CASE STUDY OF NOVENA UNIVERSITY, DELTA STATE)

ABSTRACT Visually impaired individuals often face significant challenges in navigating their environments and recognizing objects around them. This study proposes a real-time object detection system that integrates computer vision, deep learning, and audio feedback to enhance the mobility and independence of the visually impaired. The research explores and evaluates various object detection algorithms—Viola-Jones, SSD, YOLO, and notably Faster R-CNN—highlighting their strengths and limitations in terms of accuracy, speed, and suitability for real-world application. The proposed system leverages the Faster R-CNN model combined with monocular distance estimation to deliver accurate object recognition and depth perception. It is designed to run on mobile devices and provide real-time audio feedback, allowing users to identify and locate surrounding objects and people. A comprehensive literature review and gap analysis reveal that existing solutions often fall short in accuracy, usability, and adaptability to dynamic environments. Through an object-oriented design and a structured SDLC methodology, the system ensures modularity, reusability, and user-centric feedback.