Quick Start¶
Get up and running with DeepGEE in just a few minutes!
Prerequisites¶
- DeepGEE installed (Installation Guide)
- Google Earth Engine account
- Google Cloud Project ID
Step 1: Authenticate with GEE¶
First time only - authenticate with Google Earth Engine:
Step 2: Initialize GEE¶
In every script, initialize GEE with your project ID:
Step 3: Download Satellite Data¶
from deepgee import GEEDataDownloader
# Create downloader
downloader = GEEDataDownloader()
# Define region of interest (bounding box)
roi = [85.0, 20.0, 87.0, 22.0] # [lon_min, lat_min, lon_max, lat_max]
# Create composite
composite = downloader.create_composite(
roi=roi,
start_date='2023-01-01',
end_date='2023-12-31',
sensor='landsat8',
add_indices=True, # Add spectral indices
add_elevation=True # Add elevation data
)
# Download to file
downloader.download_image(
composite,
output_path='my_composite.tif',
roi=roi,
scale=30 # 30m resolution for Landsat
)
Step 4: Visualize (Optional)¶
Create an interactive map:
# Visualize on interactive map
Map = downloader.visualize_map(
composite,
vis_params={'min': 0, 'max': 0.3, 'bands': ['B5', 'B4', 'B3']},
name='False Color Composite'
)
# Display map (in Jupyter notebook)
Map
Complete Example¶
Here's a complete working example:
import deepgee
from deepgee import GEEDataDownloader
# Initialize
deepgee.initialize_gee(project='your-project-id')
# Create downloader
downloader = GEEDataDownloader()
# Define region
roi = [85.0, 20.0, 87.0, 22.0]
# Download data
print("Creating composite...")
composite = downloader.create_composite(
roi=roi,
start_date='2023-01-01',
end_date='2023-12-31',
sensor='landsat8'
)
print("Downloading...")
downloader.download_image(composite, 'output.tif', roi=roi, scale=30)
print("✓ Done! Image saved to 'output.tif'")
What's Next?¶
Learn More¶
- GEE Setup Guide - Detailed GEE configuration
- User Guide - Comprehensive documentation
- API Reference - Detailed API docs
Try Examples¶
- Land Cover Classification - Complete ML workflow
- Change Detection - Temporal analysis
- Crop Monitoring - Time series analysis
Common Tasks¶
Download Different Sensors¶
# Sentinel-2
composite = downloader.create_composite(
roi=roi,
start_date='2023-01-01',
end_date='2023-12-31',
sensor='sentinel2', # Change sensor
add_indices=True
)
Calculate Specific Indices¶
from deepgee import SpectralIndices
# Add NDVI only
composite_ndvi = SpectralIndices.add_ndvi(composite, sensor='landsat8')
# Add all indices
composite_all = SpectralIndices.add_all_indices(composite, sensor='landsat8')
Extract Training Samples¶
# Extract samples from points
samples = downloader.extract_training_samples(
composite,
points=training_points, # ee.FeatureCollection
scale=30
)
Tips¶
Project ID
Always use your Google Cloud Project ID when initializing GEE.
Scale
Use 30m for Landsat, 10m for Sentinel-2.
ROI Format
ROI should be [lon_min, lat_min, lon_max, lat_max].
Authentication
You only need to authenticate once. After that, just initialize with your project ID.
Getting Help¶
- User Guide - Detailed documentation
- Examples - Complete workflows
- GitHub Issues - Report problems
- Email: pulakesh.mid@gmail.com