Virtual Assistance for Visually Impared

Abstract

Vision is essential for perceiving the environment, and its absence greatly impacts the visually impaired, making mobility and access to information difficult. To address this, we propose an assistive system using a Convolutional Neural Network (CNN) for object recognition. Implemented on a multimedia processor with OpenCV, the system identifies objects in real-time and delivers audio feedback. This enables visually challenged individuals to make decisions independently, enhancing their safety, mobility, and quality of life without constant reliance on others.