AI-Powered Precision Agriculture Analytics Using Drone Imagery

A complete real-time crop disease detection and decision support system built with YOLOv8, Firebase, and React.

📋 Table of Contents

Overview
System Architecture
Features
Tech Stack
Prerequisites
Installation Guide
Project Structure
Configuration
Usage Guide
Troubleshooting
API Reference
Contributing
License

🎯 Overview

This system revolutionizes precision agriculture by combining:

Real-time AI detection of crop diseases using drone imagery
Spatial intelligence for precision spray path planning
Economic impact analysis with ROI calculation
Multimodal fusion of vision + sensor data
Intelligent alerts and decision recommendations
Mission reporting with complete audit trail

Problem Solved: Farmers lose 30-40% of crops to disease and waste 78% of chemicals by spraying entire fields. Our solution detects diseased zones in real-time and generates precision treatment paths, saving money and reducing environmental impact.

🏗️ System Architecture

┌─────────────────┐
│  Drone Camera   │
│  (Image Feed)   │
└────────┬────────┘
         │
         ▼
┌─────────────────────────────┐
│  Edge AI (YOLOv8)           │
│  - Real-time detection      │
│  - Runs on drone/ground PC  │
│  - No cloud dependency      │
└────────┬────────────────────┘
         │
         ▼
┌─────────────────────────────┐
│  Firebase Realtime DB       │
│  - Event streaming          │
│  - Sub-500ms latency        │
└────────┬────────────────────┘
         │
         ▼
┌─────────────────────────────┐
│  React Dashboard            │
│  - Live map visualization   │
│  - Path planning            │
│  - Economic analysis        │
│  - Decision support         │
│  - PDF report generation    │
└─────────────────────────────┘

✨ Features

Core Capabilities

✅ Real-time crop disease detection using YOLOv8
✅ Live GPS-mapped visualization with Leaflet
✅ Automated precision spray path generation
✅ Economic impact & ROI calculation
✅ Multimodal fusion (vision + soil sensors)
✅ Intelligent alert system with prioritization
✅ Mission report generation (PDF export)
✅ Sub-500ms detection-to-dashboard latency

Dashboard Panels

Map View - Interactive field map with drone position, detection pins, spray path overlay
Spray Path Generator - Automated path planning for infected zones
Economic/ROI Panel - Cost comparison, savings calculation, net benefit
Multimodal Intelligence - Fused diagnosis from vision + sensor data
Analytics - Detection statistics and disease distribution
Detection Feed - Live stream of all detections with confidence scores
Alerts & Decisions - Prioritized alerts with recommended actions

🛠️ Tech Stack

Backend (Edge AI)

Python 3.8+
YOLOv8 (Ultralytics)
OpenCV
Firebase Admin SDK
NumPy

Frontend (Dashboard)

React 18
React Leaflet (Maps)
Firebase SDK
jsPDF (Report generation)
date-fns
CSS3 Grid Layout

Infrastructure

Firebase Realtime Database
Firebase Authentication (optional)
Node.js 16+

📦 Prerequisites

Before installation, ensure you have:

Python 3.8 or higher (Download)
Node.js 16 or higher (Download)
npm or yarn (comes with Node.js)
Git (Download)
Firebase account (Sign up free)
Webcam or video file (for testing)
4GB+ RAM (for YOLOv8 inference)

Check Installations

python --version    # Should show 3.8+
node --version      # Should show 16+
npm --version       # Should show 8+
git --version       # Any recent version

📥 Installation Guide

1. Clone Repository

# Clone the project
git clone https://github.com/yourusername/precision-agriculture-analytics.git
cd precision-agriculture-analytics

# Project structure
# precision-agriculture-analytics/
# ├── backend/          # Edge AI inference
# ├── dashboard/        # React frontend
# ├── models/           # YOLOv8 weights
# └── README.md

2. Backend Setup (Edge AI)

Step 2.1: Create Python Virtual Environment

cd backend

# Create virtual environment
python -m venv venv

# Activate virtual environment
# On Windows:
venv\Scripts\activate

# On macOS/Linux:
source venv/bin/activate

# Your prompt should now show (venv)

Step 2.2: Install Python Dependencies

# Upgrade pip
pip install --upgrade pip

# Install required packages
pip install ultralytics opencv-python firebase-admin python-dotenv numpy pillow

# Verify installations
pip list

Expected packages:

ultralytics       8.0.0+
opencv-python     4.8.0+
firebase-admin    6.2.0+
python-dotenv     1.0.0+
numpy             1.24.0+
pillow            10.0.0+

Step 2.3: Download YOLOv8 Model

# Create models directory
mkdir -p ../models
cd ../models

# Option A: Use pre-trained YOLOv8 (for testing)
# The model will auto-download on first run

# Option B: Use custom-trained model (recommended for production)
# Place your trained weights here:
# models/crop_disease_yolov8.pt

Step 2.4: Backend Configuration

cd ../backend

# Create .env file
touch .env

# Edit .env with your settings
nano .env  # or use any text editor

backend/.env content:

# Firebase Configuration
FIREBASE_CREDS_PATH=./firebase-credentials.json
FIREBASE_DATABASE_URL=https://your-project.firebaseio.com

# Video Source
VIDEO_SOURCE=0  # 0 for webcam, or path to video file
# VIDEO_SOURCE=test_videos/field_video.mp4

# Model Configuration
MODEL_PATH=../models/crop_disease_yolov8.pt
CONFIDENCE_THRESHOLD=0.5

# Session Configuration
SESSION_ID=FIELD-001
FIELD_CENTER_LAT=28.6139
FIELD_CENTER_LNG=77.2090

3. Firebase Setup

Step 3.1: Create Firebase Project

Go to Firebase Console
Click "Add project"
Enter project name: precision-agriculture
Disable Google Analytics (optional)
Click "Create project"

Step 3.2: Enable Realtime Database

In Firebase Console, go to "Build" → "Realtime Database"
Click "Create Database"
Select location: United States or closest to you
Start in "Test mode" (for development)
Click "Enable"

Set up security rules:

{
  "rules": {
    ".read": "auth != null || true",
    ".write": "auth != null || true"
  }
}

⚠️ Note: For production, restrict access with proper authentication.

Step 3.3: Generate Service Account Key

In Firebase Console, go to Project Settings (⚙️ icon)
Go to "Service accounts" tab
Click "Generate new private key"
Click "Generate key" (downloads JSON file)
Rename file to firebase-credentials.json
Move to backend/ directory

# Move downloaded key to backend folder
mv ~/Downloads/your-project-firebase-adminsdk-xxxxx.json backend/firebase-credentials.json

# Verify file exists
ls backend/firebase-credentials.json

Step 3.4: Get Database URL

In Realtime Database page, copy the database URL
It looks like: https://your-project-default-rtdb.firebaseio.com/
Update FIREBASE_DATABASE_URL in backend/.env

4. Frontend Setup (Dashboard)

Step 4.1: Install Node Dependencies

cd dashboard

# Install all npm packages
npm install

# This installs:
# - react
# - react-leaflet
# - leaflet
# - firebase
# - jspdf
# - date-fns
# - and all other dependencies

If you get errors, try:

# Clear npm cache
npm cache clean --force

# Delete node_modules and reinstall
rm -rf node_modules package-lock.json
npm install

Step 4.2: Configure Firebase in Frontend

# Get Firebase config from Firebase Console
# Project Settings → General → Your apps → Web app

# Create .env file
touch .env.local

# Edit .env.local
nano .env.local

dashboard/.env.local content:

REACT_APP_FIREBASE_API_KEY=AIzaSyXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
REACT_APP_FIREBASE_AUTH_DOMAIN=your-project.firebaseapp.com
REACT_APP_FIREBASE_DATABASE_URL=https://your-project.firebaseio.com
REACT_APP_FIREBASE_PROJECT_ID=your-project
REACT_APP_FIREBASE_STORAGE_BUCKET=your-project.appspot.com
REACT_APP_FIREBASE_MESSAGING_SENDER_ID=123456789012
REACT_APP_FIREBASE_APP_ID=1:123456789012:web:xxxxxxxxxxxxx

To get Firebase config:

Firebase Console → Project Settings
Scroll to "Your apps" section
Click "Web app" icon (</>)
Copy the config values

Step 4.3: Update Firebase Config File

Edit dashboard/src/firebase/config.js:

import { initializeApp } from 'firebase/app';
import { getDatabase } from 'firebase/database';

const firebaseConfig = {
  apiKey: process.env.REACT_APP_FIREBASE_API_KEY,
  authDomain: process.env.REACT_APP_FIREBASE_AUTH_DOMAIN,
  databaseURL: process.env.REACT_APP_FIREBASE_DATABASE_URL,
  projectId: process.env.REACT_APP_FIREBASE_PROJECT_ID,
  storageBucket: process.env.REACT_APP_FIREBASE_STORAGE_BUCKET,
  messagingSenderId: process.env.REACT_APP_FIREBASE_MESSAGING_SENDER_ID,
  appId: process.env.REACT_APP_FIREBASE_APP_ID
};

const app = initializeApp(firebaseConfig);
export const database = getDatabase(app);

5. Running the System

Step 5.1: Start Backend (Edge AI)

cd backend

# Activate virtual environment if not already active
source venv/bin/activate  # macOS/Linux
# OR
venv\Scripts\activate     # Windows

# Run inference
python inference.py

# Expected output:
# ✅ Firebase initialized
# ✅ YOLOv8 model loaded
# 🎥 Video source opened
# 📡 Streaming detections to Firebase...

Step 5.2: Start Frontend (Dashboard)

Open a NEW terminal window:

cd dashboard

# Start React development server
npm start

# Dashboard will open automatically at:
# http://localhost:3000

# Expected output:
# Compiled successfully!
# webpack compiled with 0 warnings
# Dashboard running on http://localhost:3000

Step 5.3: Verify System is Working

Backend Terminal: Should show detections being processed

Frame 45 | Detections: 3 | Uploaded to Firebase ✅
Frame 46 | Detections: 2 | Uploaded to Firebase ✅

Dashboard: Should show:
- ✅ LIVE connection badge (green)
- ✅ Detection markers appearing on map
- ✅ Detection feed updating in real-time
- ✅ Live metrics incrementing

📁 Project Structure

precision-agriculture-analytics/
│
├── backend/                      # Edge AI inference
│   ├── inference.py              # Main detection script
│   ├── firebase_handler.py       # Firebase upload logic
│   ├── gps_simulator.py          # GPS coordinate generation
│   ├── firebase-credentials.json # Service account key (private)
│   ├── .env                      # Environment variables
│   ├── requirements.txt          # Python dependencies
│   └── venv/                     # Virtual environment
│
├── dashboard/                    # React frontend
│   ├── public/
│   │   └── index.html
│   ├── src/
│   │   ├── components/
│   │   │   ├── Dashboard.jsx     # Main dashboard component
│   │   │   ├── MapView.jsx       # Interactive map
│   │   │   ├── PathPlanningPanel.jsx
│   │   │   ├── EconomicImpactPanel.jsx
│   │   │   ├── FusionInsightPanel.jsx
│   │   │   ├── AlertsDecisionPanel.jsx
│   │   │   ├── DetectionFeed.jsx
│   │   │   ├── StatsPanel.jsx
│   │   │   ├── LiveStatus.jsx
│   │   │   └── MissionReportPanel.jsx
│   │   ├── hooks/
│   │   │   └── useDetections.js  # Firebase data subscription
│   │   ├── utils/
│   │   │   ├── pathPlanner.js    # Spray path algorithm
│   │   │   ├── economicCalculator.js
│   │   │   ├── fusionEngine.js
│   │   │   ├── alertRuleEngine.js
│   │   │   └── reportGenerator.js
│   │   ├── firebase/
│   │   │   └── config.js         # Firebase client config
│   │   ├── App.js
│   │   └── index.js
│   ├── .env.local                # Firebase config (private)
│   ├── package.json
│   └── package-lock.json
│
├── models/                       # YOLOv8 model weights
│   └── crop_disease_yolov8.pt
│
├── test_videos/                  # Sample videos for testing
│   └── field_video.mp4
│
└── README.md                     # This file

⚙️ Configuration

Backend Configuration (backend/.env)

Variable	Description	Example
`FIREBASE_CREDS_PATH`	Path to service account JSON	`./firebase-credentials.json`
`FIREBASE_DATABASE_URL`	Firebase Realtime DB URL	`https://project.firebaseio.com`
`VIDEO_SOURCE`	Video input (0=webcam, path=file)	`0` or `test_videos/field.mp4`
`MODEL_PATH`	YOLOv8 model weights path	`../models/crop_disease_yolov8.pt`
`CONFIDENCE_THRESHOLD`	Detection confidence filter	`0.5` (0.0 to 1.0)
`SESSION_ID`	Mission/field identifier	`FIELD-001`
`FIELD_CENTER_LAT`	Field center latitude	`28.6139`
`FIELD_CENTER_LNG`	Field center longitude	`77.2090`

Frontend Configuration (dashboard/.env.local)

All variables must start with REACT_APP_:

Variable	Description
`REACT_APP_FIREBASE_API_KEY`	Firebase Web API key
`REACT_APP_FIREBASE_AUTH_DOMAIN`	Firebase auth domain
`REACT_APP_FIREBASE_DATABASE_URL`	Realtime Database URL
`REACT_APP_FIREBASE_PROJECT_ID`	Firebase project ID
`REACT_APP_FIREBASE_STORAGE_BUCKET`	Storage bucket URL
`REACT_APP_FIREBASE_MESSAGING_SENDER_ID`	FCM sender ID
`REACT_APP_FIREBASE_APP_ID`	Firebase app ID

📖 Usage Guide

Basic Workflow

Start Backend - Begin inference on drone/ground station
Open Dashboard - Monitor detections in real-time
Analyze Field - Watch detection markers populate map
Generate Path - Click "Generate Precision Path" button
Review Economics - Check ROI and cost savings
Check Alerts - Review prioritized recommendations
Generate Report - Click "Generate Mission Report" for PDF

Using with Real Drone

# In backend/inference.py, modify video source:

# Option 1: RTSP stream from drone
VIDEO_SOURCE = "rtsp://drone_ip:8554/stream"

# Option 2: USB camera
VIDEO_SOURCE = 0

# Option 3: IP camera
VIDEO_SOURCE = "http://camera_ip:8080/video"

Testing with Sample Data

# Place test video in test_videos/
cp /path/to/your/video.mp4 test_videos/field_video.mp4

# Update backend/.env
VIDEO_SOURCE=test_videos/field_video.mp4

# Run backend
python inference.py

🔧 Troubleshooting

Common Issues

1. "Firebase initialization failed"

Symptom: Backend crashes with authentication error

Solution:

# Verify credentials file exists
ls backend/firebase-credentials.json

# Check .env has correct path
cat backend/.env | grep FIREBASE_CREDS_PATH

# Ensure Firebase Admin SDK is installed
pip show firebase-admin

2. "Dashboard shows OFFLINE"

Symptom: Connection badge is red, no data flowing

Solution:

# Check backend is running
# Backend terminal should show "Streaming detections..."

# Verify Firebase Database URL matches in both:
# - backend/.env
# - dashboard/.env.local

# Check Firebase Realtime Database rules allow read/write

# Test Firebase connection in browser console:
# Open dashboard → F12 → Console → Check for errors

3. "YOLOv8 model not found"

Symptom: FileNotFoundError: model path not found

Solution:

# Option A: Let YOLOv8 auto-download
# In inference.py, use:
model = YOLO('yolov8n.pt')  # Auto-downloads

# Option B: Download manually
cd models
wget https://github.com/ultralytics/assets/releases/download/v0.0.0/yolov8n.pt

# Update MODEL_PATH in .env
MODEL_PATH=../models/yolov8n.pt

4. "Camera not opening"

Symptom: Error: Cannot open video source

Solution:

# Test camera access
python -c "import cv2; print(cv2.VideoCapture(0).read())"

# Try different camera indices
VIDEO_SOURCE=1  # or 2, 3, etc.

# On Linux, check permissions
sudo usermod -a -G video $USER
# Logout and login again

# On macOS, grant terminal camera access
# System Preferences → Security & Privacy → Camera

5. "npm install fails"

Symptom: Dependency installation errors

Solution:

# Clear npm cache
npm cache clean --force

# Remove node_modules
rm -rf node_modules package-lock.json

# Update npm
npm install -g npm@latest

# Reinstall
npm install

# If still failing, use yarn
npm install -g yarn
yarn install

6. "Port 3000 already in use"

Symptom: Something is already running on port 3000

Solution:

# Option A: Kill process on port 3000
# On macOS/Linux:
lsof -ti:3000 | xargs kill -9

# On Windows:
netstat -ano | findstr :3000
taskkill /PID <PID> /F

# Option B: Use different port
PORT=3001 npm start

7. "Map not loading"

Symptom: Blank white panel where map should be

Solution:

# Check Leaflet CSS is imported
# In dashboard/src/index.js or App.js:
import 'leaflet/dist/leaflet.css';

# Verify internet connection (map tiles load from OpenStreetMap)

# Check browser console for errors
# F12 → Console → Look for 404 errors

# Try clearing browser cache
# Ctrl+Shift+R (hard refresh)

8. "GPS coordinates all the same"

Symptom: All detection markers at same location

Solution:

# This is normal if using simulated GPS
# GPS simulator adds random offsets to FIELD_CENTER

# To use real GPS:
# 1. Ensure drone sends GPS in detection payload
# 2. Modify backend/inference.py to include actual GPS
# 3. Remove GPS enrichment from Dashboard.jsx

Performance Optimization

Slow Inference Speed

# Use smaller YOLOv8 model
model = YOLO('yolov8n.pt')  # Nano (fastest)
# vs
model = YOLO('yolov8x.pt')  # Extra large (slowest)

# Reduce frame processing
# In inference.py, skip frames:
if frame_count % 3 == 0:  # Process every 3rd frame
    results = model(frame)

High Memory Usage

# In inference.py, add:
import gc
gc.collect()  # Periodically free memory

# Reduce detection history retention
# In Dashboard.jsx:
const recentDetections = detections.slice(0, 20)  # Keep only 20

🌐 API Reference

Firebase Data Structure

{
  "detections": {
    "SESSION-ID": {
      "detection-uuid-1": {
        "frame_id": 123,
        "timestamp": "2026-01-17T01:50:00.000Z",
        "detection_count": 2,
        "detections": [
          {
            "class_name": "Early Blight",
            "confidence": 0.87,
            "bbox": [100, 150, 200, 250]
          }
        ],
        "gps": {
          "lat": 28.6145,
          "lng": 77.2095
        }
      }
    }
  }
}

Detection Object Schema

Field	Type	Description
`id`	string	Unique detection ID (UUID)
`frame_id`	integer	Video frame number
`timestamp`	ISO 8601	Detection timestamp
`detection_count`	integer	Number of diseases in frame
`detections`	array	List of detected diseases
`gps`	object	GPS coordinates {lat, lng}

Disease Detection Schema

Field	Type	Description
`class_name`	string	Disease name
`confidence`	float	Confidence score (0.0-1.0)
`bbox`	array	Bounding box [x1, y1, x2, y2]

🤝 Contributing

We welcome contributions! Please follow these guidelines:

Fork the repository
Create feature branch: git checkout -b feature/your-feature
Commit changes: git commit -m 'Add your feature'
Push to branch: git push origin feature/your-feature
Submit Pull Request

Development Setup

# Install development dependencies
pip install black flake8 pytest  # Python
npm install --save-dev eslint prettier  # JavaScript

# Run linters
black backend/  # Format Python code
npm run lint  # Format JavaScript

📄 License

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

Name		Name	Last commit message	Last commit date
drone_edge		drone_edge
frontend		frontend
.gitignore		.gitignore
CHATBOT_QUESTIONS.md		CHATBOT_QUESTIONS.md
CHATBOT_SETUP.md		CHATBOT_SETUP.md
FIREBASE_INTEGRATION_SUMMARY.md		FIREBASE_INTEGRATION_SUMMARY.md
FIREBASE_QUICK_START.md		FIREBASE_QUICK_START.md
FIREBASE_SETUP_GUIDE.md		FIREBASE_SETUP_GUIDE.md
LOCAL_MODE_SETUP.md		LOCAL_MODE_SETUP.md
README.md		README.md
VIDEO_UPLOAD_GUIDE.md		VIDEO_UPLOAD_GUIDE.md
sample.jsx		sample.jsx

Latest commit

History

Folders and files

Repository files navigation

** AI-Powered Precision Agriculture Analytics Using Drone Imagery**

📋 Table of Contents

🎯 Overview

🏗️ System Architecture

✨ Features

Core Capabilities

Dashboard Panels

🛠️ Tech Stack

Backend (Edge AI)

Frontend (Dashboard)

Infrastructure

📦 Prerequisites

Check Installations

📥 Installation Guide

1. Clone Repository

2. Backend Setup (Edge AI)

Step 2.1: Create Python Virtual Environment

Step 2.2: Install Python Dependencies

Step 2.3: Download YOLOv8 Model

Step 2.4: Backend Configuration

3. Firebase Setup

Step 3.1: Create Firebase Project

Step 3.2: Enable Realtime Database

Step 3.3: Generate Service Account Key

Step 3.4: Get Database URL

4. Frontend Setup (Dashboard)

Step 4.1: Install Node Dependencies

Step 4.2: Configure Firebase in Frontend

Step 4.3: Update Firebase Config File

5. Running the System

Step 5.1: Start Backend (Edge AI)

Step 5.2: Start Frontend (Dashboard)

Step 5.3: Verify System is Working

📁 Project Structure

⚙️ Configuration

Backend Configuration (backend/.env)

Frontend Configuration (dashboard/.env.local)

📖 Usage Guide

Basic Workflow

Using with Real Drone

Testing with Sample Data

🔧 Troubleshooting

Common Issues

1. "Firebase initialization failed"

2. "Dashboard shows OFFLINE"

3. "YOLOv8 model not found"

4. "Camera not opening"

5. "npm install fails"

6. "Port 3000 already in use"

7. "Map not loading"

8. "GPS coordinates all the same"

Performance Optimization

Slow Inference Speed

High Memory Usage

🌐 API Reference

Firebase Data Structure

Detection Object Schema

Disease Detection Schema

🤝 Contributing

Development Setup

📄 License

About

Topics

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages

AI-Powered Precision Agriculture Analytics Using Drone Imagery