Implementation of Yolo-Based Object Detection and Sensor Integration Model for Visually Challenged Person
M. Kamarunisha, K. Jayasri, Saraban Tasnim Sharin, M. Priyadharshini, T. K. Anusuya, R. Malathi · 2024
It is suggested that visually impaired individuals use an assistive device to receive automatic direction and navigation. The gadget may provide obstacle-free navigation and use real-time image processing to detect impediments. The goal of the strategy is to maximise information flow by completing all activities in real-time and gathering all the information regarding path impediments. The suggested navigation stick is an inexpensive item, obstacle, and obstacle identifier. For safety and navigation, the suggested assistance system for the blind makes use of cutting-edge technology such as tilt and ultrasonic sensors, as well as YOLO(YouOnly Look Once) object recognition. It updates its location via GPS, haptic feedback, and real-time object detection. The user interface is independent and accessible through non-visual means. The suggested work focuses on a multi-label method for detecting items in indoor and outdoor settings utilising vision technology. It improves accuracy and cuts down on recognition time by using classification and clustering techniques.