R.E.T.I.N.A. Real-Time Environmental Transcription and Interactive Navigation Assistant for Visually Impaired

RETINA (Real-time Environmental Transcription and Interpretation for Navigation and Assistance for Visually Impaired) is an innovative project aimed at enhancing the navigation and environmental awareness of visually impaired individuals through advanced assistive technology. The system integrates a portable, headband-mounted camera with a Raspberry Pi, enabling real-time capture and interpretation of visual information from the user's surroundings. Utilizing cutting-edge computer vision and machine learning algorithms, along with the Google Coral Accelerator for rapid processing, RETINA facilitates swift object recognition and detailed auditory feedback. The Mobilenet tflite model plays a crucial role in generating accurate, contextually relevant descriptions of identified objects, enhancing the quality of information provided to users. By continuously capturing video footage and processing it in real-time, RETINA delivers immediate audio cues about objects, obstacles, and environmental features through connected earphones. This innovative approach addresses key challenges faced by visually impaired individuals, such as limited environmental awareness and reliance on external assistance, ultimately promoting greater independence and safety. With a functional prototype already developed, RETINA stands at the forefront of assistive technology, potentially transforming the lives of visually impaired users by enabling more confident navigation and interaction with their surroundings.