This project provides an AI-powered solution for real-time safety monitoring in construction zones. It detects safety gear violations (PPE compliance), danger zone breaches, and notifies supervisors through visual dashboards and SMS alerts.
📊 Features
Real-time detection of:
No Hardhat
No Mask
No Safety Vest
Danger Zone Intrusion
DeepSORT-based person tracking
CLAHE-enhanced video processing for better accuracy
SMS alert system via Twilio
Web dashboard for detection log viewing
Violation summary with tracked counts per frame
🚀 Getting Started
Prerequisites
Python 3.8+
Django
OpenCV
Ultralytics YOLO
DeepSORT (deep_sort_realtime)
Twilio
Installation
Clone the repository:
git clone https://github.com/your-username/construction-safety-monitoring.git cd construction-safety-monitoring
Create virtual environment and install dependencies:
python -m venv venv source venv/bin/activate # or venv\Scripts\activate pip install -r requirements.txt
Run migrations:
python manage.py makemigrations python manage.py migrate
Start the development server:
python manage.py runserver
🚧 Model Weights
Due to file size, the YOLOv11s model (best.pt) is not included. Download it from Google Drive and place it inside the APP/ folder.
🔧 How It Works
Web camera feed is processed in real-time.
CLAHE enhances frame contrast.
YOLO detects objects; DeepSORT assigns consistent IDs.
If a person is detected in the danger zone or violates PPE rules, a log is saved with:
Frame number
Track ID
Violation type
Timestamp
Person count
SMS is sent via Twilio if alerts are enabled.
🌟 Violation Summary Fields
Total persons tracked
No Mask
No Hardhat
No Safety Vest
Danger Zone Intrusions
Collaborators (if any)
🔧 Future Enhancements
Crowd density monitoring
Posture-based fatigue detection
Tool and material misuse detection
Attendance with PPE compliance
Centralized cloud dashboard
🌟 Accuracy
Model accuracy: 87.2% (YOLOv11s variant chosen after testing multiple models)