Raspberry Pi based Smart Assistance for Visually Impaired People

Reading is the essential life skill in today's society as the reports, bills, bank-related files, classroom worksheets, pharmaceutical product instructions, and other printed documents are present everywhere. The main objective of this research work is to convert a real text into speech and to develop collision avoidance for the blind. It entails capturing the image with the Raspberry Pi camera, processing it with image processing, and then simply reading it aloud. The text in the captured image is predicted by an artificial neural network. The output generated by the proposed device will be in the form of speech, with OCR to help the visually impaired people to understand it. The speech output of the software is sent to the blind's headphones, allowing them to hear the information from the text present in front of them. For blind people, this technology is both effective and cost-effective. Moreover, this technology is highly beneficial to blind students in schools and colleges. As an artificial intelligence application, the developed application is useful for illiterate people, as well as extremely important to society.