Facial Recognition and Obstacle Detection Smart Glasses for Visually Challenged People

Abstract: This study presents the design and development of facial recognition and obstacle detection smart glasses for visually challenged people using Raspberry Pi. The proposed system utilizes a Raspberry Pi 4 Model B, a Pi 3 camera, an ultrasonic sensor and a speaker to recognize and identify pre-registered individuals along with obstacle distance measurement. The facial recognition algorithm is based on the OpenCV library. The system provides real-time audio feedback to the user, announcing the name of the recognized individual and also provides audio warning if the obstacle is less than 18 cm. Our results show that the system achieves an accuracy of 92.5% in recognizing the pre-registered individuals. User feedback and usability testing revealed significant improvements in social interactions and confidence levels among visually challenged individuals. This study demonstrates the potential of low-cost, Raspberry Pi-based facial recognition and obstacle detection systems to empower visually challenged people.

Keywords: Raspberry Pi 4, Pi 3 Camera, Real Time Face Detection, Audio feedback system, Object detection for accessibility