YOLOv8 Smoke and Fire Detection

YOLOv8 Smoke and Fire Detection

Designed and deployed a live YOLOv8-based smoke and fire detection pipeline on Azure/Django at Capgemini for a public safety authority in Germany.

  • The Challenge: Real-time detection of fire and smoke in large-scale environments like stadiums, requiring high precision and low latency.
  • Technical Approach: Leveraged YOLOv8 for real-time object detection. The pipeline handles video stream ingestion, frame slicing, and model inference.
  • System Architecture: Deployed on Azure using a Django backend to provide a live monitoring dashboard and alert system.
  • Visual Representation: YOLOv8 Stadium Pipe

  • Keywords: YOLOv8, Computer Vision, Azure, Django, Real-time Systems.