Kinetic-Triggered Visual Triage for Real-Time Post-Crash Driver Assessment on Smart Highways
This research presents Kinetic-Triggered Visual Triage (KTVT), a Cyber-Physical System for real-time post-crash driver assessment on smart highways. The proposed framework integrates RFID infrastructure, IoT-based vibration sensing, ESP32-CAM image acquisition, and CNN-based driver classification to enable rapid emergency triage after vehicle collisions. Experimental evaluation achieved 88.7% classification accuracy with end-to-end latency below 275 ms.
Authors
- Imasha Perera
- Ravindu Lakshan
- Keshani Konara
- Sathira Porogama
Institutions
- Sri Lanka Institute of Information Technology (LK)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-05-28
- DOI
- https://doi.org/10.5281/zenodo.20431899
- Primary Topic
- IoT and GPS-based Vehicle Safety Systems
- Type
- article
- Field-Weighted Citation Impact
- 0.00