A large portion of the global population suffers from vision-related disabilities, which restricts their capacity to move freely and engage with their surroundings. Standard assistance tools such as walking canes and trained canines offer restricted support and fail to deliver detailed environmental data like text reading or item recognition. This research proposes an intelligent wearable device powered by artificial intelligence, combining edge computing, deep learning, and visual processing technologies. The prototype hardware comprises an image sensor, distance measuring sensors, a Raspberry Pi processing board alongside a Google Coral Edge TPU, and an audio output unit. By employing cutting-edge detection algorithms like YOLO together with Optical Character Recognition methods, the device processes live visual feeds, recognizes surrounding items and barriers, extracts written content, and transforms this data into spoken words. In contrast to opaque commercial AI products, our method prioritizes low response time and human-centric design. Testing outcomes indicate reliable item detection, barrier alerts, and text interpretation with negligible delay. This device delivers an economical, transportable, and transparent support solution that improves movement, security, and self-reliance for blind and partially sighted users.

This document presented an AI-enabled wearable vision assistant system designed for persons with sight impairment. The suggested system joins computer vision, deep learning, and edge computing technologies to supply live environmental awareness through spoken feedback. By integrating YOLO-based item detection, OCR text recognition, and ultrasonic barrier detection, the system furnishes complete assistance that resolves the constraints of conventional aids like white canes and guide dogs. The principal innovations of this work consist of a lightweight, economical hardware design using Raspberry Pi and edge TPU acceleration; combination of various sensing methods for full scene comprehension; real-time on-device processing safeguarding user privacy; and natural language spoken feedback for straightforward user interaction. Testing outcomes indicate that the suggested system realizes precise item detection, instantaneous barrier warnings, and dependable text reading ability with negligible delay. The system furnishes an economical substitute to costly commer-cial products, making assistance technology more attainable to sight-impaired persons worldwide. Subsequent efforts will concentrate on expanding item categories, enhancing low-light performance, and incorporating facial recognition for social interaction support.

[1]World Health Organization, “Blindness and vision im-pairment,” WHO Fact Sheets, 2023. [Online]. Available: https://www.who.int/newsroom/fact-sheets/detail/blindness-and-visual-impairment [2]R. C. Joshi, N. Singh, A. K. Sharma, R. Burget, and M. K. Dutta, “AI-SenseVision: A Low-Cost Artificial-Intelligence-Based Robust and Real-Time Assistance for Visually Impaired People,” IEEE Transactions on Human-Machine Systems, vol. 54, no. 3, pp. 325-336, Jun. 2024. [3]Y. Song, Z. Li, G. Li, B. Wang, M. Zhu, and P. Shi, “Multi-Sensory Visual-Auditory Fusion of Wearable Navigation Assistance for People With Impaired Vision,” IEEE Transactions on Automation Science and Engineering, vol. 22, pp. 4724-4736, 2025. [4]J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You Only Look Once: Unified, Real-Time Object Detection,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 2016, pp. 779-788. [5]A. Kumar, R. Patra, and M. M. Mahbub, “Smart Assistive System for Visually Impaired People Using Raspberry Pi,” in 2022 International Conference on IoT and Embedded Systems, Chennai, India, 2022, pp. 45-50. [6]S. Patil, P. More, and K. Kulkarni, “Ultrasonic Smart Glasses for Visually Impaired Navigation,” International Journal of Engineering Research & Technology, vol. 12, no. 4, pp. 234-239, 2023. [7]W. Li, J. Zhang, and S. Chen, “Real-Time Object Detection and Recog-nition for Visually Impaired Assistance Using YOLOv5,” in Proceedings of IEEE Conference on Artificial Intelligence, Beijing, China, Apr. 2024, pp. 112-118. [8]H. Wang, L. Liu, and Y. Zhao, “IoT-Enabled Smart Navigation System for Visually Impaired People,” IEEE Internet of Things Journal, vol. 10, no. 8, pp. 6789-6801, 2023. [9]J. Redmon and A. Farhadi, “YOLOv3: An Incremental Improvement,” arXiv preprint arXiv:1804.02767, 2018. [10]A. Bochkovskiy, C. Y. Wang, and H. Y. M. Liao, “YOLOv4: Optimal Speed and Accuracy of Object Detection,” arXiv preprint arXiv:2004.10934, 2020. [11]R. Smith, “An Overview of the Tesseract OCR Engine,” Google Re-search, Technical Report, 2023. [12]Google Coral Team, “Edge TPU: Accelerating ML at the Edge,” Google AI Blog, 2024. [Online]. Available: https://coral.ai [13]Microsoft Research, “Seeing AI: App for Blind and Low Vision Community,” Microsoft, 2024. [Online]. Available: https://www.microsoft.com/en-us/ai/seeing-ai [14]OrCam Technologies, “OrCam MyEye: Wearable Assistive Technology,” 2024. [Online]. Available: https://www.orcam.com [15]L. Deng, Y. Liu, and M. Zhang, “Deep Learning for Assistive Technolo-gies: A Comprehensive Survey,” ACM Computing Surveys, vol. 56, no. 5, pp. 1-35, 2024.