← All projects
CCTV surveillanceBETA
Fire & Smoke Detection + EMNS
Camera-based fire, smoke, and anomaly detection that automatically triggers Emergency Mass Notification across PA, SMS, and mobile channels
Overview
This system makes every camera an emergency sensor. Deep-learning models detect smoke, flames, falls, and crowd anomalies in live video; confirmed events trigger the Emergency Mass Notification System within seconds - broadcasting to PA speakers, sirens, SMS, and mobile apps with escalation workflows.
The same stack serves environmental monitoring: long-range cameras watch for wildfire smoke, illegal dumping, and rising water levels at rivers and reservoirs, fused with IoT sensor data for reliable early warning.
Key features
Smoke and flame detection from live video
Fall and crowd anomaly detection
Multi-channel EMNS broadcast in seconds
Wildfire and water-level monitoring
IoT sensor fusion for fewer false alarms
Tech stack
YOLOv8 · aiPyTorch · aiMQTT · otherFastAPI · backendTimescaleDB · databaseNVIDIA Jetson · devops