Intelligent Glasses for Visually Impaired People

Vision impairment and blindness are a big concern in the research field of assistive technologies. An individual with sight impairment suffers from performing tasks that seem simple to a sighted person. However, these effortless tasks for sighted humans are very complex to be performed by machines. With fast processors and high-resolution cameras, researchers can develop accurate algorithms to assist blind and visually impaired (BVI) individuals in performing their daily activities. In this research, we designed an intelligent eyeglass that integrates various models to perform face recognition, object detection, and path navigation. When the system is powered, a welcoming message is played to the user asking him/her to choose one of the three modes. When the user pronounces the desired mode, the system uses Google speech recognition API to identify the request, tunes the required mode, and responds accordingly. Face recognition and object detection modes provide audio feedback of the names of the surrounding known people and objects, respectively. On the other hand, the navigation mode triggers a slight vibration whenever an obstacle is within a range of half a meter. Raspberry Pi 4 is used to control the system components because it can perform all the sophisticated computations needed in a short time, offering the user a realistic experience. Moreover, the low cost of Raspberry Pi, its small size, and its lightweight result in practical and simple-design eyeglasses.