Visionai:Real-Time Object Detection with Audio Assistance for Visually Impaired People
International Journal of Science, Strategic Management and Technology · 18.03.2026 · 02 (03) · 1-9
DOI: 10.55041/ijsmt.v2i3.131 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Science, Strategic Management and Technology, online erschienen am 18.03.2026. Verfasst von Akilan A und 1 weiteren Autor:innen.
Abstract (Originalsprache)
According to the World Health Organization, 2.2 billion people are visually impaired worldwide, and 1 billion of the visually impaired cannot move around their environments using conventional mobility aids. In this study, the author suggests a real-time object detector audio assistant called VisionAI, which will assist visually impaired persons in moving around in a self-sufficient way by continually surveying the environment. The proposed solution is based on real-time video streaming at 1280720 pixels and identifying 80 classes of objects using the YOLOv10 model in 4.2 ms/frame. Objects are made to measure the distance using the Pinhole Camera Model, which estimates the distance at an average error of 0.5 m. The horizontal frame consists of three areas, and it identifies the direction of objects with an accuracy rate of 93 percent. The solution proposed here applies a three-level urgency classification model called CRITICAL, HIGH, and MODERTE to categorize objects based on their distance. Audio-visual feedback was delivered based on pyttsx3, text-to-speech, Web Audio API, and React 18 front-end. The solution is based on FastAPI, WebSocket, and React stack, and the latency is less than one second between end-to-end, and the detection of the object is 91 percent in clear and 84 percent in the low-light environment.
Zitieren
APA
A, A., & S.V., D. A. (2026). Visionai:Real-Time Object Detection with Audio Assistance for Visually Impaired People. International Journal of Science, Strategic Management and Technology, 02(03), 1-9. https://doi.org/10.55041/ijsmt.v2i3.131
Vancouver
A A, S.V. DA. Visionai:Real-Time Object Detection with Audio Assistance for Visually Impaired People. International Journal of Science, Strategic Management and Technology. 2026;02(03):1-9. doi:10.55041/ijsmt.v2i3.131
BibTeX
@article{A2026,
author = {A, Akilan and S.V., Dr. Anandhi},
title = {Visionai:Real-Time Object Detection with Audio Assistance for Visually Impaired People},
journal = {International Journal of Science, Strategic Management and Technology},
year = {2026},
volume = {02},
number = {03},
pages = {1--9},
doi = {10.55041/ijsmt.v2i3.131}
}
EUVIMED