Assistive Reading Device for Blind Individuals
Anthony Huynh, Zoltan Marchant, Inderdeep Singh, Federica Aveta · 2023
Completely blind individuals face significant challenges when it comes to reading text. Braille is the most used method for reading and writing, but braille may not always be available. Therefore, many blind individuals rely on assistive technologies to help them navigate the world around them. One such technology is smart glasses that can read text aloud to the wearer. However, the high cost of these devices is a significant barrier to access for many individuals. Our solution aims to provide a low-cost alternative for blind individuals who need a reliable text-reading device. By using open-source libraries and commercial hardware, we have been able to reduce the cost for our system. Our prototype is made of a camera, microcontroller, speaker, amplifier, battery, and power supply. Utilizing Pytesseract for Optical Character Recognition (OCR), OpenCV libraries for image preprocessing, and Google’s text-to-speech synthesis technology, the program captures an image, extracts the English text as a string, and plays the audio back to the wearer. The OCR obtained an accuracy of 99% during our testing. With a high accuracy rate, the listener can be confident that the text is accurately transcribed, making it easy to comprehend.