Real-Time Visual Navigation Framework for Edge-Level Systems Using Classical Vision Algorithms

Basit Öğe Kaydı
dc.contributor.authorKaya, Baris
dc.contributor.authorTekelioglu, Melih Ilter
dc.contributor.authorSarioglu, Baykal
dc.date.accessioned2026-04-04T18:48:33Z
dc.date.available2026-04-04T18:48:33Z
dc.date.issued2025
dc.description19th International Conference on Innovations in Intelligent Systems and Applications, INISTA 2025 -- 29 October 2025 through 31 October 2025 -- Ras Al Khaimah -- 217522
dc.description.abstractThis paper presents a novel real-time navigation assistance system that enhances driver awareness by visually embedding directional guidance directly into the roadway scene. Instead of displaying instructions on a separate map interface, our system renders navigational cues - such as arrows - within the actual road view captured by a forward-facing camera. These cues are aligned with the detected lane geometry, enabling drivers to follow directions more intuitively without shifting focus away from the road. The system relies on lightweight computer vision techniques using OpenCV, avoiding the need for complex neural networks or high-performance hardware. It operates efficiently on minimal CPU resources and requires only a simple camera. Real-time integration with Google Maps provides up-to-date route information, which is seamlessly transformed into visual overlays projected within the driving lane. This design significantly improves driver focus and reaction time, particularly at higher speeds or in unfamiliar environments. By bridging map data with on-road visual guidance, our method offers a practical step forward in intelligent driver assistance systems. The approach supports future extensions such as windshield-projected interfaces, but its current implementation already demonstrates how smart, vision-based systems can make navigation safer, more accessible, and more human-centered. © 2025 IEEE.
dc.description.sponsorshipAmerican University of Ras Al Khaimah; Huawei; IEEE UAE Section ? Advancing Technology for Humanity; OpenCEMS Industrial Chair; Yildiz Technical University
dc.identifier.doi10.1109/INISTA68122.2025.11249600
dc.identifier.isbn979-833157024-8
dc.identifier.scopus2-s2.0-105030447532
dc.identifier.scopusqualityN/A
dc.identifier.urihttps://doi.org/10.1109/INISTA68122.2025.11249600
dc.identifier.urihttps://hdl.handle.net/11411/10228
dc.indekslendigikaynakScopus
dc.language.isoen
dc.publisherInstitute of Electrical and Electronics Engineers Inc.
dc.relation.ispartof19th International Conference on Innovations in Intelligent Systems and Applications, INISTA 2025 - Proceedings
dc.relation.publicationcategoryKonferans Öğesi - Uluslararası - Kurum Öğretim Elemanı
dc.rightsinfo:eu-repo/semantics/closedAccess
dc.snmzKA_Scopus_20260402
dc.subjectAugmented Navigation Cues
dc.subjectComputer Vision
dc.subjectDriver Assistance
dc.subjectEmbedded Directions
dc.subjectIntelligent Transportation Systems
dc.subjectLane Detection
dc.subjectMap Integration
dc.subjectOpencv
dc.subjectReal-Time Navigation
dc.titleReal-Time Visual Navigation Framework for Edge-Level Systems Using Classical Vision Algorithms
dc.typeConference Paper

Dosyalar

Koleksiyon

Scopus Indexed Publications