• Deeksha D
• Suresha D
• H N Prakash

Visually impaired people often struggle with navigation and many environmental awareness. This paper presents an improved smart glasses system that leverages real-time scene analysis [1], object detection, and OCR [2][3], using lightweight CNN models optimized for edge AI platforms. Building on prior work with Google Glass and Raspberry Pi, the proposed system enhances performance through on-device deep learning, enabling faster and more accurate detection. It offers audio feedback for contextual guidance, supporting safe navigation in complex environments. Experimental results show the system performs reliably in both indoor and outdoor settings, validating its potential as a practical, cost-effective assistive solution.

Smart Glasses, Visually Impaired Assistance, Real-Time Scene Analysis, Object Detection, Deep Learning, Embedded Systems, Edge Computing, Google Glass, Optical Character Recognition (OCR [2][3]) , Edge AI, Assistive Technology, Human-Computer Interaction (HCI), Computer Vision, Voice Feedback Systems, Navigation Aid, Sensory Substitution, Wearable AI Devices.

" Survey on AI-Powered Smart Glasses: Real-Time Scene Analysis and Navigation Support for the Visually Impaired", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 7, page no.949-954, July-2025, Available :http://www.jetir.org/papers/JETIRGX06176.pdf

" Survey on AI-Powered Smart Glasses: Real-Time Scene Analysis and Navigation Support for the Visually Impaired", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 7, page no. pp949-954, July-2025, Available at : http://www.jetir.org/papers/JETIRGX06176.pdf