Skip to content

sohamgudewar/AI-DocParser

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🌿 KrishiNetra: AI-Powered Precision Agriculture

KrishiNetra is an end-to-end MVP designed to empower farmers and agricultural stakeholders with data-driven insights. Built for the Maharashtra Agrihackathon 2026, it combines satellite imagery analysis with Generative AI to bridge the gap between complex data and actionable farming decisions.

🚀 Key Features

  • 🛰️ Satellite NDVI Analysis: Visualizes vegetation health using pre-computed NDVI overlays to identify high-stress zones.
  • 📄 AI Document Extraction: Uses Google Gemini to automatically parse and verify complex PDF land records (7/12 extracts).
  • 🌾 Crop Prediction: Context-aware crop identification based on geographic coordinates and local lookup data.
  • 📍 Interactive Mapping: Built with Streamlit and Folium for a seamless, location-based user experience.
  • 🛡️ Stress Detection: Automated calculation of "at-risk" farm areas to prioritize irrigation and intervention.

🛠️ Quick Start

# Navigate to the MVP directory
cd krishinetra_mvp

# Activate venv (from repo root)
..\krishinetra_venv\Scripts\Activate.ps1

# Install deps
pip install -r requirements.txt

# Run app
streamlit run app.py

📂 Project Structure

krishinetra_mvp/
├── app.py                 # Main app
├── modules/               # Python modules
│   ├── map_viz.py         # Folium map
│   ├── ndvi_loader.py     # NDVI PNG loader
│   ├── crop_predictor.py  # Lookup crop classifier
│   ├── stress_detector.py # Threshold stress zones
│   ├── area_calculator.py # Shapely area math
│   └── doc_extractor.py   # Gemini PDF extractor
├── data/
│   ├── ndvi_overlays/     # Pre-computed NDVI PNGs
│   ├── sample_documents/  # Demo land record PDF
│   └── crop_lookup.json   # Coordinate→crop map
├── .env                   # Gemini API key (optional)
└── requirements.txt

🤖 Gemini API (Optional)

For document AI, add your key to .env:

GEMINI_API_KEY=your_key_here

Without it, the app uses fallback demo data.


📸 Demo Scenarios

Select a demo scenario from the sidebar:

  • A: Cotton Farm Stress - Vidarbha — NDVI stress analysis
  • B: Onion Fields - Nashik — Mixed vegetation health
  • C: Land Document Verification - Pune — AI document extraction

About

AI-powered precision agriculture platform featuring NDVI satellite analysis, crop stress detection, and automated land record extraction for smarter farming

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors