SentryVision - Critical Region Object Throw Detection System

Overview

SentryVision is an advanced system that detects fast-moving thrown objects (such as sacks, stones, or tools) entering a user-defined critical region in real-time video feeds. It is designed to minimize false positives from birds, animals, and natural environmental motion using state-of-the-art computer vision techniques.

Features

• Real-time video input from webcam, RTSP streams, or video files
• Interactive ROI (Region of Interest) selection with visual feedback
• Advanced background subtraction (MOG2) with adaptive learning
• High-speed motion and size filtering with configurable thresholds
• Morphological noise reduction and contour analysis
• Visual and console alerts with comprehensive event logging
• Automatic snapshot saving on detection with metadata
• Robust error handling and system recovery
• Modern web-based dashboard with real-time monitoring
• Event database with search and export capabilities
• Video recording and playback functionality

File Structure

• main.py: Main application and GUI interface
• app.py: Web-based dashboard and API endpoints
• motion_detector.py: Advanced motion detection logic with object tracking
• roi_manager.py: ROI selection and management with persistence
• alert_system.py: Alert generation and logging system
• event_db.py: Event database management and querying
• config.py: Configurable parameters and system settings
• requirements.txt: Python dependencies
• templates/: Web interface HTML templates
• api/: API endpoints and dashboard logic

Setup

1. Install dependencies:

   pip install -r requirements.txt

2. Run the web dashboard:

   python app.py

3. Run the desktop GUI (optional):

   python main.py

Usage

• Web Interface: Open browser to http://localhost:5000
• ROI Selection: Draw region of interest on video feed
• Detection Control: Start/stop detection and recording
• Event Review: Browse detected events and snapshots
• Settings: Adjust detection parameters in real-time

Production Deployment

SentryVision is production-ready with Gunicorn and Nginx support.

Setup

1. Install Nginx for reverse proxy
2. Run web interface:

   gunicorn --worker-class eventlet -w 1 -b 0.0.0.0:5000 app:app

3. Run with multiple workers for scalability:

   gunicorn --worker-class eventlet -w 4 -b 0.0.0.0:5000 app:app

Nginx Configuration

server {
    listen 80;
    server_name yourdomain.com;

    location / {
        proxy_pass http://127.0.0.1:5000/;
        proxy_http_version 1.1;
        proxy_set_header Upgrade $http_upgrade;
        proxy_set_header Connection "Upgrade";
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
        proxy_set_header X-Forwarded-Proto $scheme;
    }

    # Static file serving
    location /static/ {
        alias /path/to/sentryvision/static/;
        expires 30d;
    }
}

Performance Optimization

• Use load balancer with sticky sessions for multiple workers
• Enable gzip compression in Nginx
• Configure proper caching headers
• Monitor system resources and adjust worker count accordingly

Security Features

• Input validation and sanitization
• Secure file upload handling
• SQL injection prevention
• XSS protection
• Rate limiting support

Monitoring and Maintenance

• Event logging with structured data
• Performance metrics collection
• Automatic error recovery
• Database maintenance utilities