Smart Eye for Visually Impaired People

Abstract

Abstract: This paper presents a novel mobile application tailored to enhance the daily experiences of individuals with visual impairments. Leveraging cutting-edge real-time object detection and Optical Character Recognition (OCR) techniques, the app, constructed using the TensorFlow algorithm and the comprehensive Common Objects in Context (COCO) dataset, aims to empower users by providing invaluable assistance through their mobile devices. The application utilizes the device's camera to seamlessly identify objects in real-time via the object detection algorithm, while simultaneously extracting and vocalizing text from captured images through OCR technology. A crucial aspect of the app lies in its commitment to user-friendliness, ensuring a straightforward interface tailored specifically for the needs of visually impaired individuals. The efficacy of the system was rigorously evaluated, focusing on both accuracy and real-time performance. Results conclusively demonstrate the app's effectiveness and efficiency in offering essential support to visually impaired users, marking a significant stride toward fostering independence and ease in their daily lives. Remarkably, this mobile application stands out for its affordability and accessibility, providing a practical and inclusive solution for individuals with visual impairments. By delivering audio feedback about identified objects and recognized text, this innovative app contributes to a more accessible and autonomous lifestyle for its users, contributing positively to the broader landscape of assistive technology for the visually impaired.