LunarAlign

Multi-modal, sun-angle & scale invariant image correspondence using Chandrayaan-2 optical images (OHRC, TMC-2, IIRS).

Getting Started for New Contributors

Welcome! Here is the quick setup workflow to get up and running:

1. Python Virtual Environment:
   • Create: python -m venv .venv
   • Activate (PowerShell): .\.venv\Scripts\Activate.ps1 (or source .venv/bin/activate on bash)
   • Install dependencies: pip install -r requirements.txt
2. Frontend Setup:
   • cd frontend && npm install
3. Run Locally:
   • Backend: python -m uvicorn backend.main:app --reload --port 8000 (API on http://localhost:8000)
   • Frontend: npm run dev inside frontend/ (UI on http://localhost:5173)
4. Documentation:
   • Deep-dive documentation: docs/ (architecture, algorithms, evaluation)

Prerequisites

• Python 3.10+ (tested with 3.11)
• Node.js 18+ (tested with 20.20.0)
• npm 10+

Backend

# Create a virtual environment
python -m venv .venv

# Activate the virtual environment (Windows PowerShell)
.\.venv\Scripts\Activate.ps1

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

# Install dependencies
pip install -r requirements.txt

# Start the API server
python -m uvicorn backend.main:app --reload --port 8000

Verify: open http://localhost:8000/health

Note: If .venv is not activated, prefix commands with .\.venv\Scripts\python.exe -m (e.g., .\.venv\Scripts\python.exe -m uvicorn ...).

Frontend

cd frontend
npm install
npm run dev

Verify: open http://localhost:5173

Tests

# With .venv activated:
python -m pytest tests/ -v

# Without activation:
.\.venv\Scripts\python.exe -m pytest tests/ -v

Documentation

• Architecture
• Algorithms
• Evaluation