Smart braille translator: Integrating braille-to-speech and speech-to-braille using raspberry PI

ABSTRACT: In the Philippines, an estimated 500,000 individuals experience visual impairments, with a considerable number lacking proficiency in Braille. This study aims to address challenges in developing a Braille translation system, specifically ensuring accurate recognition, interpretation, and conversion of Braille characters into text and speech. The primary concerns involve (1) how effectively the system identifies Braille patterns, (2) maintains accuracy in text conversion, (3) and scamlessly transforms text or speech into its respective braille pattern on the braille pad. To achieve these objectives, the system employs a Rasberry Pi 4 microcontroller, integrating Convolutional Neural Networks (CNN) for Braille character recognition, followed by Natural Language Processing (NLP) and Braille mapping techniques for precise translation. The researchers adopted the Agile Scrum methodology throughout the system’s development process. The development cycle was structured into sprints, each representing a distinct phase of progress. Every sprint encompassed key elements, including defining the project vision, maintaining backlogs using Trello for task management, planning, implementation, daily scrum meetings, reviews, retrospectives, and continuous development. During data collection, researchers engaged with respondents from the Philippine Blind Union community, revealing that many had only a fundamental understanding of Braille. These insights further emphasized the necessity of an efficient and accessible Braille translation system. The final system demonstrated full functionality, offering an effective tool for converting Braille into text and speech. By addressing the critical challenges in Braille translation, this study contributes to improving literacy and communication for visually impaired individuals, fostering greater independence and accessibility in their daily lives.

Filipiniana