08 / Research
Research Prototype
MS-ADA Intelligent Traffic Surveillance
A CPU-first academic computer-vision prototype for vehicle tracking, ALPR, accident reasoning, hit-and-run monitoring, and evidence capture.
PythonYOLO11OpenCVTrackingALPROptical FlowDocker
Case study
What this project demonstrates
MS-ADA is an academic research prototype that combines computer vision components into an interpretable temporal event pipeline. It is intentionally presented as research rather than as deployment-grade traffic safety infrastructure.
Architecture
Video → YOLO detection → persistent tracking → motion/ALPR signals → temporal collision reasoning → evidence capture + alerts
Engineering highlights
- Real-time vehicle detection and tracking with Pakistani plate recognition and evidence outputs.
- Deterministic multi-signal accident reasoning using contact/proximity, relative motion, deceleration, optical flow, and trajectory convergence.
- Hit-and-run monitoring, event CSVs, processed video, snapshots/clips, and alert workflows.
- Final Year Project at NED University; publication details will be added once the research paper is formally published.
Boundaries / limitations
Academic prototype; no claim of production readiness or universal accident-detection accuracy. Research-paper citation will be added after formal publication.