🛡️ IntelliGuardAI - Smart Surveillance System

An AI-powered surveillance system with real-time threat detection. This streamlined codebase supports a modern web interface and a simple CLI stream. Desktop GUI has been removed for a cleaner structure.

✨ Key Features Implemented

🎥 Live Camera Feed Display

• Real-time CCTV footage streaming with low-latency playback
• Smooth video processing at configurable FPS (default: 30 FPS)
• Support for webcam and IP cameras with easy configuration
• Fullscreen mode support in web interface
• Responsive video display that adapts to different screen sizes

🔍 Real-Time Object Detection

• YOLOv8-based object detection with 80+ COCO dataset objects
• Real-time bounding box visualization with confidence scores
• Dynamic object labeling on video feed
• Color-coded detection boxes (red for threats, green for normal objects)
• Configurable confidence thresholds for detection accuracy

⚠️ Threat Alert System

• Automatic threat detection for weapons, violence, and suspicious activities
• Visual alert indicators with color coding and animations
• Audio alert notifications using Web Audio API
• SMS integration via Twilio for immediate notifications
• Emergency stop functionality for critical situations
• Alert cooldown system to prevent spam notifications

🎨 Modern UI/UX

• Clean, responsive web interface with live stats and logs
• Mobile/tablet responsive design for remote monitoring
• Dark mode styling and WebSocket real-time updates

🚀 Quick Start

Prerequisites

• Python 3.8+
• Webcam or IP camera
• CUDA-compatible GPU (optional, for faster processing)
• Twilio account (for SMS alerts)

Installation

1. Clone the repository

   git clone <repository-url>
   cd IntelliGuardAI

2. Install dependencies

   pip install -r requirements.txt

3. Configure environment variables

   # Copy the example environment file
   cp .env.example .env
   
   # Edit .env with your Twilio credentials (for SMS alerts)
   # Get credentials from https://console.twilio.com/

4. Configure the system (optional)

   # Edit config.py to customize settings
   # Or set environment variables
   export CAMERA_SOURCE=0
   export CONFIDENCE_THRESHOLD=0.5
   export SMS_ENABLED=true

🖥️ Usage

Web Interface (Recommended)

python main.py

• URL: http://localhost:8000/web
• Features: Live video feed, real-time alerts, statistics, logs
• Best for: Remote monitoring, multiple users, mobile access

Command Line Stream

python detect_stream.py

• Features: Terminal-based detection, SMS alerts, console output
• Best for: Server deployment, automated monitoring, logging

📱 Web Interface Features

Live Video Feed

• Real-time camera stream with detection overlays
• Fullscreen mode support
• Automatic refresh and reconnection
• Responsive design for all devices

Real-Time Alerts

• Visual threat indicators with animations
• Audio alert notifications
• Alert history and details
• Emergency stop button

Statistics Dashboard

• Total detections counter
• Active threats counter
• Real-time detection list
• System status indicators

Activity Log

• Live detection logs
• Threat-specific highlighting
• Timestamp information
• Scrollable history

🔧 Configuration

Camera Settings

# config.py
CAMERA_SOURCE = 0              # Camera index or IP camera URL
CAMERA_WIDTH = 640            # Video width
CAMERA_HEIGHT = 480           # Video height
CAMERA_FPS = 30              # Frames per second

AI Model Settings

MODEL_PATH = "yolov8n.pt"     # YOLO model file
CONFIDENCE_THRESHOLD = 0.5    # Detection confidence (0.0-1.0)
USE_GPU = True               # Enable GPU acceleration

Threat Detection

THREAT_OBJECTS = [
    'knife', 'gun', 'fight', 'snatching', 
    'weapon', 'pistol', 'rifle', 'assault'
]
ALERT_COOLDOWN = 30          # Seconds between alerts

SMS Alerts

SMS alerts require a Twilio account. Configure via environment variables:

# Create .env file from template
cp .env.example .env

# Edit .env with your Twilio credentials
# Get credentials from https://console.twilio.com/

Environment variables:

SMS_ENABLED=true
TWILIO_ACCOUNT_SID=your_account_sid
TWILIO_AUTH_TOKEN=your_auth_token
TWILIO_PHONE_NUMBER=+1234567890
TARGET_PHONE_NUMBER=+0987654321

⚠️ Never commit your actual credentials to version control!

🎯 Detection Capabilities

Supported Objects (80+ COCO Objects)

• People: person, child, adult, man, woman
• Vehicles: car, truck, bus, motorcycle, bicycle
• Animals: dog, cat, bird, horse, cow
• Objects: chair, table, phone, laptop, book
• And many more...

Threat Detection

• Weapons: knife, gun, weapon, pistol, rifle
• Violence: fight, assault, attack
• Crime: snatching, theft, robbery
• Custom: Add your own threat categories

📊 Performance Optimization

GPU Acceleration

• Automatic CUDA detection and usage
• Significant performance improvement with NVIDIA GPUs
• Fallback to CPU if GPU unavailable

Model Optimization

• Use smaller models (yolov8n) for speed
• Use larger models (yolov8l, yolov8x) for accuracy
• Adjustable confidence thresholds

System Requirements

• Minimum: 4GB RAM, CPU-only
• Recommended: 8GB+ RAM, CUDA GPU
• Optimal: 16GB+ RAM, RTX 3060+ GPU

🔒 Security Features

Production Deployment

• Configurable CORS settings
• Rate limiting support
• API key authentication (optional)
• HTTPS support

Privacy Compliance

• Local processing (no cloud dependencies)
• Configurable data retention
• Secure log storage
• User consent mechanisms

🐛 Troubleshooting

Common Issues

Camera not detected:

# Check camera permissions
# Try different camera index
# Update config.py: CAMERA_SOURCE = 1

Model loading error:

# Model will be downloaded automatically
# Or download manually:
wget https://github.com/ultralytics/assets/releases/download/v0.0.0/yolov8n.pt

Performance issues:

# Reduce resolution in config.py
CAMERA_WIDTH = 320
CAMERA_HEIGHT = 240

SMS not sending:

• Create .env file from .env.example
• Add your Twilio credentials to .env
• Verify phone number format (+country code)
• Ensure sufficient Twilio credits

Logs and Debugging

• Check console output for errors
• Monitor system resources
• Verify camera permissions
• Test with sample images

🛠️ Customization

Adding Custom Threat Categories

# In config.py
THREAT_OBJECTS = [
    'knife', 'gun', 'fight', 'snatching',
    'your_custom_threat'  # Add your own
]

Custom Alert Messages

ALERT_MESSAGES = {
    'knife': '⚠️ Knife detected! Potential threat identified.',
    'your_threat': '🚨 Your custom alert message!'
}

Custom Alert Actions

# Add custom alert handling in your interface
def custom_alert_handler(threat_type, confidence):
    # Your custom logic here
    pass

📈 API Endpoints

Core Endpoints

• GET / - Health check
• GET /web - Web interface
• GET /video_feed - Live video stream
• GET /api/status - System status
• GET /api/threats - Current threats
• GET /api/logs - Detection logs

Analysis Endpoints

• POST /analyze - Analyze uploaded image
• WebSocket /ws - Real-time updates

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Test thoroughly
5. Submit a pull request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Ultralytics for YOLOv8 implementation
• OpenCV for computer vision capabilities
• FastAPI for web framework
• Twilio for SMS integration
• Font Awesome for icons

---

⚠️ Important Notice: This system is designed for security and surveillance purposes. Please ensure compliance with local laws and regulations regarding camera usage and privacy rights.