🚦 Smart Traffic Management System

AI-powered traffic signals that adapt in real time — using YOLOv8 computer vision and dynamic timing logic.

---

Traditional traffic lights run on fixed timers — regardless of whether 1 car or 50 cars are waiting. This system uses YOLOv8 to count vehicles per lane in real time and gives the busiest lane the longest green light. Emergency vehicles get an instant override.

---

✨ Features

🚗 Vehicle Detection YOLOv8 detects cars, motorcycles, buses & trucks Works on live camera, video file, or simulation Rolling-average smoothing prevents signal thrashing	🧠 Dynamic Signal Control Green time scales from 10s → 60s with vehicle count Busiest lane always gets priority 4 density levels: LOW / MEDIUM / HIGH / CRITICAL
🚨 Emergency Override Detects ambulances & fire trucks via color analysis Instantly overrides signal to give emergency vehicles a clear path Logs all override events	🖥️ Simulation Mode Run without any camera or hardware Synthetic traffic data generator included Great for testing and demos

---

⚡ Quick Start

git clone https://github.com/aasimansari1/smart-traffic-system.git
cd smart-traffic-system

pip install -r requirements.txt

# Run in simulation mode (no camera needed)
python main.py

# Single image detection
python main.py --image path/to/traffic.jpg

# Video file
python main.py --mode video --source path/to/video.mp4

# Live camera
python main.py --mode camera

---

🏗️ How It Works

  📷 Camera / Video / Simulation
          │
          ▼
  ┌──────────────────┐
  │  VehicleDetector │  ← YOLOv8 — counts cars/buses/trucks per lane
  └────────┬─────────┘
           │
           ▼
  ┌──────────────────┐
  │ DensityAnalyzer  │  ← rolling average → LOW/MEDIUM/HIGH/CRITICAL
  └────────┬─────────┘
           │
           ▼
  ┌──────────────────┐     ┌───────────────────┐
  │ SignalController │     │ EmergencyDetector │
  │ (dynamic timing) │◄────│ (color analysis)  │
  └────────┬─────────┘     └───────────────────┘
           │
           ▼
  🟢 Green time: 10s–60s (scales with density)
  🚨 Emergency override: instant green

---

⏱️ Signal Timing Logic

Density	Vehicles	Green Time
LOW	0 – 5	10 – 18s
MEDIUM	6 – 15	19 – 32s
HIGH	16 – 25	33 – 47s
CRITICAL	25+	48 – 60s

All thresholds and lane names are configurable in config/settings.py.

---

🗂️ Project Structure

smart-traffic-system/
├── config/
│   └── settings.py              # All tunable params (timing, thresholds, lanes)
├── models/
│   ├── vehicle_detector.py      # YOLOv8 detection pipeline
│   ├── density_analyzer.py      # Density classification + smoothing
│   ├── signal_controller.py     # Green time calculation + priority logic
│   └── emergency_detector.py    # Color-based emergency detection
├── utils/
│   ├── simulator.py             # Synthetic traffic frame generator
│   └── logger.py                # Event logging
├── tests/
│   ├── test_density_analyzer.py
│   └── test_signal_controller.py
├── app.py                       # Flask dashboard (real-time monitoring)
├── main.py                      # CLI entry point
└── yolov8n.pt                   # Pre-trained YOLOv8 nano weights (~6.5MB)

---

🧪 Tests

python tests/test_density_analyzer.py
python tests/test_signal_controller.py

---

🌐 Web Dashboard

python app.py
# → http://localhost:5000

Real-time view of vehicle counts, density levels, and signal state per lane.

---

🛠️ Tech Stack

Layer	Technology
Object Detection	YOLOv8 (Ultralytics)
Computer Vision	OpenCV
Numerical	NumPy
Web Dashboard	Flask
Pre-trained Model	yolov8n.pt (included)

---

🗺️ Roadmap

• IoT hardware integration (ESP32 / Arduino signal controllers)
• Accident detection module
• Google Maps real-time traffic data feed
• Multi-intersection coordination
• Mobile app for traffic authority alerts

---

🤝 Contributing

git checkout -b feature/your-feature
git commit -m 'Add your feature'
git push origin feature/your-feature

---

📄 License

MIT © Mohd Aasim Ansari

---

Making roads smarter, one intersection at a time. If this interests you, please ⭐ star the repo!