AgroVision is an advanced geospatial agriculture intelligence dashboard that combines satellite remote sensing, GIS visualization, environmental analytics, and AI-assisted crop monitoring to analyze agricultural land directly from space.
The platform uses live Sentinel-2 satellite imagery and environmental intelligence to monitor vegetation health, detect crop stress, analyze environmental conditions, and generate agricultural insights in real time.
- 🛰️ Live Sentinel-2 Satellite Monitoring
- 🌱 NDVI Crop Health Analysis
- 🗺️ Interactive GIS Polygon Selection
- 📈 Time-Series Vegetation Monitoring
- 🌧️ Weather Intelligence Integration
- 🚨 AI-Based Health Alerts
- 💧 Soil Moisture Analysis
- 🌊 Flood Detection Analytics
- 🤖 AI Crop Advisory Engine
- 📊 Real-Time Dashboard Visualization
- 🌍 Multi-Layer Satellite Mapping
| Domain | Technology |
|---|---|
| Frontend | Streamlit |
| GIS Visualization | Folium |
| Satellite APIs | Sentinel Hub |
| Data Processing | NumPy |
| Data Visualization | Plotly / Matplotlib |
| Weather APIs | Open-Meteo |
| Geospatial APIs | OpenStreetMap |
| Future AI Stack | TensorFlow / Scikit-learn |
The primary objective of AgroVision is to transform raw Earth observation data into actionable agricultural intelligence using geospatial analytics and remote sensing technologies.
The platform enables:
- crop health monitoring
- vegetation stress detection
- drought analysis
- environmental intelligence
- smart farming visualization
- AI-assisted agricultural recommendations
without requiring physical field inspection.
The platform uses live Earth observation data from:
- European Space Agency (ESA) Sentinel-2
- Sentinel Hub APIs
| Band | Purpose |
|---|---|
| B02 | Blue |
| B03 | Green |
| B04 | Red |
| B08 | Near Infrared (NIR) |
| B11 | Shortwave Infrared (SWIR) |
| SCL | Scene Classification Layer |
NDVI = (NIR - Red) / (NIR + Red)
Healthy vegetation strongly reflects Near Infrared wavelengths while absorbing red light because of chlorophyll activity.
| NDVI Range | Interpretation |
|---|---|
| 0.6 – 1.0 | Healthy vegetation |
| 0.3 – 0.6 | Moderate vegetation |
| 0.1 – 0.3 | Weak vegetation |
| < 0.1 | Bare soil / unhealthy crop |
Cloud-contaminated pixels are automatically removed using Sentinel-2 Scene Classification Layer (SCL) masking to improve vegetation analysis accuracy.
The masking engine filters:
- cloud pixels
- cloud shadows
- cirrus contamination
- invalid pixels
NDVI = (NIR - Red) / (NIR + Red)
Used for:
- crop health monitoring
- chlorophyll analysis
- vegetation intensity
NDMI = (NIR - SWIR) / (NIR + SWIR)
Used for:
- drought analysis
- water stress detection
- soil moisture estimation
NDWI = (Green - NIR) / (Green + NIR)
Used for:
- flood detection
- water body analysis
- irrigation monitoring
Current version includes:
- rule-assisted crop stress analysis
- environmental recommendation engine
- simulated predictive analytics pipeline
- CNN Crop Classification
- Disease Detection AI
- Yield Prediction Models
- SAR Radar Soil Moisture Analysis
- Drone-Based Crop Monitoring
- Tamil Voice AI Farming Assistant
git clone https://github.com/Swethancyber/AgroVision.git
cd AgroVisionpip install -r requirements.txtstreamlit run app.pypip install streamlit sentinelhub numpy matplotlib folium streamlit-folium plotly requests- Real CNN Crop Classification
- Disease Detection AI
- Real-Time Satellite Monitoring
- Climate Risk Prediction
- Multi-Farm Monitoring
- Voice-Based Tamil AI Assistant
- Drone + Satellite Hybrid Monitoring
- Predictive Agricultural Intelligence
- Precision Agriculture
- Smart Farming Systems
- Crop Health Monitoring
- Environmental Intelligence
- Climate Analytics
- Agricultural Decision Support
- Water Stress Analysis
- Disaster & Flood Monitoring
This project is intended for educational, research, and innovation purposes.
swethan
Geospatial AI | Remote Sensing | Precision Agriculture | Environmental Intelligence
GitHub: https://github.com/Swethancyber