eo-learn
A Python framework for processing spatio-temporal satellite imagery and extracting features for machine learning applications.
What is eo-learn?
eo-learn is an open-source Python framework for processing and analyzing Earth observation satellite imagery. It provides tools to handle spatio-temporal data, extract features, and prepare datasets for machine learning applications, addressing the challenge of automatically deriving insights from large volumes of satellite data.
Remote sensing scientists, geospatial data engineers, and machine learning practitioners working with satellite imagery for applications like land cover monitoring, disaster control, and environmental analysis.
Developers choose eo-learn for its modular, reusable task system that simplifies complex EO workflows, its seamless integration with the Python data science stack, and its active community that shares and improves processing components.
Overview
Earth observation processing framework for machine learning in Python
Use Cases
Best For
- Automating land cover classification from satellite time-series data
- Building custom cloud masking and image preprocessing pipelines
- Extracting spatio-temporal features for environmental monitoring models
- Processing Copernicus Sentinel and Landsat imagery at scale
- Prototyping machine learning workflows for remote sensing research
- Creating reproducible geospatial data processing workflows
Not Ideal For
- Real-time or low-latency satellite data analysis for applications like live disaster monitoring
- Teams working exclusively in non-Python ecosystems such as R or Java
- Projects requiring only basic image visualization without spatio-temporal processing or machine learning integration
Pros & Cons
Pros
EOTasks and EOWorkflow enable building reusable processing chains, as illustrated in the README for water mapping with NDVI thresholding, fostering collaboration and code reuse.
EOPatch structure manages time-series satellite imagery efficiently, which is essential for handling large datasets from fleets like Copernicus and Landsat.
Leverages NumPy, scikit-learn, and other Python libraries, making it easy to incorporate machine learning and data science tools into remote sensing workflows.
Supports custom task development and community contributions through a structured package system, with extra tasks available in separate repositories for flexibility.
Cons
Requires installing system-specific libraries (e.g., GDAL on Windows via unofficial wheels) and managing extra dependencies, which can be error-prone and time-consuming.
Examples are housed in a separate repository (eo-learn-examples) and may not be up-to-date, complicating the learning curve for new users.
Relies on standard Python libraries without built-in distributed computing support; scaling to petabyte-level data requires additional tools like RAY or custom infrastructure.
Frequently Asked Questions
Related Projects
detectron2Detectron2 is a platform for object detection, segmentation and other visual recognition tasks.
EasyOCRReady-to-use OCR with 80+ supported languages and all popular writing scripts including Latin, Chinese, Arabic, Devanagari, Cyrillic and etc.
imgaugImage augmentation for machine learning experiments.
meshroomNode-based Visual Programming Toolbox
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.