Crop Classification of Belle-Île Island using Sentinel-1 and Sentinel-2 Data

In this project, I used vegetation indices derived from both Sentinel-1 and Sentinel-2 imagery for the crop classification of Belle-Île Island, France. This end-to-end project demonstrates the complete workflow, including data acquisition, preprocessing, feature engineering, and modeling.

Key Features and Tools

• Data Acquisition and Processing: Leveraged the Google Earth Engine (GEE) Python API for querying and processing satellite imagery.
• Feature Engineering:
  • Sentinel-2 Vegetation Indices: NDVI, GNDVI, NBR, and EVI.
  • Sentinel-1 Indices: RVI, DPSVI, and DPSVIo.
• Modeling: Implemented crop classification using Random Forest and XGBoost machine learning algorithms.
• Visualization: Utilized geemap for interactive visualizations.