AI System

MonitorCOC AI: Intelligent Smoke & Fire Detection

Computer vision system with YOLOv8 and OpenCV to detect smoke and fire in real-time from Hikvision and Dahua CCTV cameras.

View Live Project

The Client

COC Control / Sivilla

Operations centers needed an automatic smoke and fire detection system to complement human surveillance, reducing response times to emergencies and improving safety in critical facilities.

The Challenge

Emergency detection automation was required:

Detect smoke and fire before human operators
Process multiple camera streams simultaneously
Integrate with existing Hikvision and Dahua cameras
Minimize false positives
Alert in real-time upon detection

The Solution

We developed an AI-powered detection system:

YOLOv8 model trained for smoke and fire
Video processing with OpenCV
Support for Hikvision and Dahua cameras
Real-time analysis with GPU
Configurable confidence and threshold system
Instant alerts with evidence capture
Integration with Sivilla Core via API
Real-time monitoring dashboard

Technologies Used

Tech stack for AI system:

Python

YOLOv8

OpenCV

CUDA/GPU

REST API

RTSP

Results Achieved

The system significantly improves response times:

<3s Detection time

95%+ Accuracy

24/7 Continuous monitoring

GPU Processing