IoT-Driven Refreshable Braille Architecture Utilizing Vision-Based OCR and Cloud Storage

This paper introduced an IoT-enabled architecture for Braille reading that is based on vision-based OCR, cloud synchronization, embedded processing, and tactile Braille creation. The designed technology uses the ESP32-S3 CAM to read printed documents, FastAPI web service with Tesseract OCR to get text out of it, and Firebase Realtime Database to synchronize this information. The ESP32-S3 processes this information, creating six-dot patterns from the retrieved characters, which it then renders into Braille dots by controlling six push-pull solenoids through a DRV8833 driver. Interaction with LCD and navigation buttons was also supported by the device. As well as Android application, which allows to upload other types of content – images and voice