Assistive Technology for the Visually Impaired Using Computer Vision and Image Processing

Abstract

Blindness and vision loss, whether congenital or acquired, have a negative impact on people's quality of life. Visually impaired children may face delays in their motor, social, and cognitive development. This has long-term implications and lowers educational achievement. Adults with vision impairment, on the other hand, have decreased engagement and productivity in the workforce. Visual impairments have a strong association with social isolation, despair, and anxiety, which contributes to an increased likelihood of early admission to nursing and care homes. Our proposed technology is intended to assist people with complete or partial vision loss in doing daily tasks without the need for assistance from others. In the form of wearable spectacles, the system combines a Raspberry Pi with a camera, microphone, and sensors. The system can perform five major functions based on computer vision, image processing, and speech recognition: obstacle detection, currency recognition, face recognition, color detection, object detection, and text recognition. The system's operation is modified for a more user-friendly experience with the help of a mobile application and cloud services. The system was put to the test, and the overall accuracy was 88.2%.