Raspberry Pi based Assistive Device for Visually Impaired

Recent research reveals that 36 million individuals are entirely blind, while 217 million experience moderate to severe visual impairment (MSVI). Navigating daily life poses significant challenges for visually impaired individuals, necessitating the development of advanced assistive technologies. We introduce an innovative supportive device aimed at increasing the self-sufficiency of visually impaired individuals. This device integrates Optical Character Recognition (OCR), object detection, and door detection capabilities. It further employs a Raspberry Pi microcontroller along with a camera module to capture images, which are subsequently processed to deliver auditory feedback to the user. The OCR system converts printed text into speech using Text-to-Speech (TTS) synthesis, enabling users to access written information. The object detection module identifies and announces obstacles and landmarks, while the door detection module informs users about the status of doors, enhancing navigation and safety. Performance metrics such as precision, accuracy, recall, and F1 score were assessed, demonstrating the efficacy of each model. An 89% of accuracy was attained by the OCR model, while the object detection and door detection models recorded accuracies of 87.6% and 82%, respectively. Despite limitations such as restricted handwritten text recognition and limited computational power, the device offers a user-friendly and cost-effective solution for visually impaired individuals, promoting greater autonomy and ease in daily activities.