Trinetra: The Divine Eye of Mahakumbh

Team ID: TH1296

🌟 Overview

Trinetra is an AI-powered surveillance and assistance system designed for large-scale religious gatherings like Mahakumbh. The platform ensures real-time crowd monitoring, rapid emergency response, and seamless navigation for millions of pilgrims through integrated mobile and web interfaces. It features advanced AI analytics including facial recognition for lost & found, GPS-based family tracking with automated alerts, disaster prediction, and dynamic emergency routing.

🎯 Problem & Solution

Problem Statement

Mass gatherings at events like Mahakumbh face critical challenges including:

Family separation in dense crowds
Uncontrolled crowd movement leading to stampede risks
Delayed emergency detection and response
Overstretched infrastructure with limited real-time visibility
Difficulty in guiding people safely during crises

Solution

Trinetra provides a comprehensive AI-powered solution integrating:

Facial Recognition Technology for rapid lost person recovery
Real-time GPS Family Tracking with automated distance alerts
Predictive Analytics for disaster forecasting
Live Crowd Density Heatmaps for optimal resource allocation
Instant SOS Emergency Response system with ambulance booking and tracking
Offline AI Navigation for uninterrupted access
Centralized Dashboards for unified monitoring and response coordination

🏗️ System Architecture

🔄 Logic & Workflow

Data Collection

User registration with family member details and missing person images
Real-time GPS location data from registered family members
Live CCTV streams and sensor data from multiple surveillance points
Weather data, satellite imagery, and crowd movement patterns

Processing

AI Facial Recognition using FaceNet technology for matching uploaded images with live CCTV feeds
Machine Learning Algorithms analyzing crowd density and movement patterns using YOLO and OpenCV [[memory:8341815]]
Real-time Processing generating automated alerts for family separation (500m+ distance), emergency situations, and anomaly detection

Output

Instant notifications for lost family members with precise location coordinates
Real-time crowd density heatmaps and risk zone alerts
Dynamic route suggestions and traffic analysis
Emergency response coordination with ambulance dispatch and tracking
Multilingual chatbot assistance for pilgrim guidance

👥 User Roles

Pilgrim Interface

- Register and manage family member profiles - Report missing persons with image upload - Receive automated alerts for family separation - Access real-time locations of nearby hospitals - Book ambulance services with live tracking - Use offline AI navigation for Kumbh area mapping

Admin Dashboard

- Monitor live CCTV feeds with automated facial recognition - Analyze real-time crowd density heatmaps - Receive instant alerts for detected anomalies - Coordinate emergency response and resource allocation - Access advanced disaster prediction tools

Medical Admin Interface

- Process ambulance booking requests - Track and coordinate ambulance dispatch - Manage medical camp information - Monitor emergency response metrics

🛠️ Tech Stack

Frontend

React.js - Main web application framework
Next.js - Server-side rendering and optimization
React Native - Mobile application development
Tailwind CSS - Utility-first CSS framework

Backend

Node.js + Express - Web server and API development
FastAPI - High-performance Python API framework
Flask - Lightweight Python web framework

Database

PostgreSQL - Primary relational database
Firestore - NoSQL database for real-time data
Redis - In-memory data structure store

AI/ML

Python - Core programming language
OpenCV - Computer vision library
YOLO - Real-time object detection
FaceNet - Face recognition technology
Gemini AI - Advanced AI capabilities

Real-time Communication

Socket.io - Real-time bidirectional event-based communication
WebRTC - Peer-to-peer communication
Kafka - Distributed streaming platform

Security & Deployment

JWT Authentication - Secure token-based authentication
Docker - Containerization platform
AWS EC2 - Cloud computing service
Prometheus & Grafana - Monitoring and visualization

📁 Project Structure

Trinetra/
├── ai-guided-map/          # AI-powered navigation system
├── crowd-detection/        # Real-time crowd monitoring
├── gun-detection/          # Weapon detection system
├── image-recognition/      # Image processing and analysis
├── lost-and-found/         # Facial recognition for missing persons
├── mahakumbh-disaster-prediction/  # Disaster prediction system
├── peoples/               # People counting and analysis
├── Trinetra/              # Main React application
└── images/                # Project documentation images

📱 Mobile App (APK)

Download and Install

The Trinetra mobile application is available as an APK file for Android devices.

APK File: base.apk

Installation Instructions

Method 1: Direct Installation (Android)

Enable Unknown Sources:
- Go to Settings → Security → Unknown Sources
- Enable "Install from Unknown Sources" or "Allow from this source"
Download APK:
- Download the base.apk file from the repository
- Transfer to your Android device via USB, email, or cloud storage
Install:
- Open the APK file on your Android device
- Tap "Install" when prompted
- Grant necessary permissions (Location, Camera, Storage)

Method 2: Using ADB (Advanced)

# Install via ADB (Android Debug Bridge)
adb install base.apk

# Or install and grant permissions
adb install -g base.apk

Mobile App Features

Real-time Family Tracking with GPS coordinates
Emergency SOS with one-tap ambulance booking
Lost Person Reporting with photo upload
Offline Navigation for Kumbh area mapping
Multilingual Support for diverse pilgrim community
Push Notifications for alerts and updates

System Requirements

Android Version: 6.0 (API level 23) or higher
RAM: 2GB minimum, 4GB recommended
Storage: 100MB available space
Network: WiFi or 4G/5G for real-time features
Permissions: Location, Camera, Storage, Network Access

🚀 Quick Start

Prerequisites

Docker and Docker Compose
Node.js 18+ (for local development)
Python 3.8+ (for local development)

Using Docker (Recommended)

Clone the repository

git clone <repository-url>
cd Trinetra

Start all services

docker-compose up -d

Access the applications

Main Application: http://localhost:3000
AI Guided Map: http://localhost:3001
Crowd Detection: http://localhost:5000
Gun Detection: http://localhost:5001
Image Recognition: http://localhost:5002
Lost & Found: http://localhost:5003
Disaster Prediction Backend: http://localhost:5004
Disaster Prediction Frontend: http://localhost:3002
User Frontend: http://localhost:3003
People Counting: http://localhost:5005

Manual Setup

1. Main React Application (Trinetra)

cd Trinetra
npm install
npm run dev

2. AI Guided Map

cd ai-guided-map
npm install
npm run dev

3. Crowd Detection

cd crowd-detection
pip install -r requirements.txt
python app.py

4. Gun Detection

cd gun-detection
pip install -r requirements.txt
python main.py

5. Image Recognition

cd image-recognition
pip install -r requirements.txt
python app.py

6. Lost & Found

cd lost-and-found
pip install -r requirements.txt
python app.py

7. Disaster Prediction Backend

cd mahakumbh-disaster-prediction/backend
pip install -r requirements.txt
python app.py

8. Disaster Prediction Frontend

cd mahakumbh-disaster-prediction/frontend
npm install
npm run dev

🔧 Configuration

Environment Variables

Each service requires specific environment variables. Copy the example files and configure:

# For Python services
cp gun-detection/env.example gun-detection/.env
cp image-recognition/env.example image-recognition/.env

# Configure your API keys and database connections

API Keys Required

Google Maps API (for navigation)
Gemini AI API (for AI features)
Weather API (for disaster prediction)
Satellite imagery API

📊 Features Showcase

Real-time Crowd Monitoring

Facial Recognition System

Emergency Response Dashboard

Disaster Prediction Analytics

GPS Tracking Interface

Medical Emergency System

Navigation Interface

Admin Control Panel

Real-time Analytics

Security Monitoring

🔮 Future Scope

Advanced Features

Drone Integration for aerial surveillance covering blind spots
Advanced CCTV Integration for automated video flow processing
Voice Monitoring capabilities for distress call detection
Machine Learning Models with reinforcement algorithms for incident prediction
City Infrastructure Integration with traffic systems and public transportation

Scalability

Adaptation for other large-scale events
Smart city implementations
International religious gatherings with multi-language support
Cultural customization for different regions

🤝 Contributing

Fork the repository
Create a feature branch (git checkout -b feature/AmazingFeature)
Commit your changes (git commit -m 'Add some AmazingFeature')
Push to the branch (git push origin feature/AmazingFeature)
Open a Pull Request

📄 License

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

👥 Team