Desarrollo e implementación de un sistema de lectura asistida mediante voz para personas invidentes con técnicas de visión artificial e inteligencia artificial

Abstract

This project addresses the design, development, and implementation of a voice-assisted reading system specifically aimed at visually impaired individuals. By leveraging advanced techniques in computer vision and artificial intelligence, the system accurately and efficiently converts printed text into audio format, enabling inclusive access to information in educational, workplace, and daily environments. The work involved selecting and configuring hardware components such as a Raspberry Pi 4, a Logitech C920s camera, and Genius SP-U115 speakers, along with software tools like the EAST model for text detection, Tesseract OCR for optical character recognition, and gTTS for voice synthesis. Additionally, interaction mechanisms, including keyboard commands and auditory feedback, were integrated to ensure an accessible and autonomous user experience. Tests conducted in various lighting scenarios confirmed the system’s remarkable performance, achieving accuracies of up to 95 % under optimal conditions. Processing times and audio clarity were also evaluated, demonstrating the device’s feasibility for real-time use. This project represents a significant step towards technological inclusion, enhancing the quality of life for visually impaired individuals and laying the groundwork for future improvements and expansions.