Cooperative Multi-Camera Tracking with Edge AI

Cooperative Multi-Camera Tracking with Edge AI

A cooperative traffic-surveillance concept combining multiple cameras, edge computing, and representation learning for more continuous vehicle tracking.

University of Washington · Research project

Project story

From question to working system.

This research explores how observations from multiple cameras can be coordinated to maintain vehicle identities and movement histories across a transportation network.

The project connects edge artificial intelligence with representation learning so processing can happen close to the sensors while information from different viewpoints is reconciled for traffic surveillance and safety analysis.

Multi-camera

Coordinated observations across multiple fields of view

Edge AI

Processing designed to operate close to the sensing source

Re-identification

Representation learning for more continuous tracks

Safety ready

Movement data organized for operations and risk analysis

My contributions

Roles and technical work.

  • Cooperative multi-camera system design
  • Edge-AI workflow development
  • Vehicle detection and tracking
  • Representation-learning methods
  • Traffic-surveillance analysis

Project gallery

Systems, fieldwork, and results.

Resources

Presentations, publications, reports, and links.

Public materials for this project are being prepared. Request project materials ↗