Advancing Earth Observation Through Machine Learning: A TorchGeo Tutorial

2026-03-02

AI Summary

This work addresses the unique challenges in Earth observation machine learning—such as georeferenced imagery, heterogeneous labels, and spatially aware sampling—and the lack of standardized tools that seamlessly integrate with mainstream deep learning frameworks. Building upon the TorchGeo library, this study presents the first systematic integration of geospatial data processing within the PyTorch ecosystem, offering an end-to-end, reproducible development paradigm. The methodology is demonstrated through a semantic segmentation case study on water bodies using Sentinel-2 imagery, showcasing a cohesive workflow that combines unified coordinate transformations, spatially aware samplers, pretrained models, and interactive Jupyter Notebook tutorials. The pipeline directly produces GeoTIFF prediction outputs suitable for downstream geospatial analysis. All code and tutorials are publicly released.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Machine Learning: Time-Series/Data Streams
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Security and Privacy: Data transparency and provenance

Abstract

Earth observation machine learning pipelines differ fundamentally from standard computer vision workflows. Imagery is typically delivered as large, georeferenced scenes, labels may be raster masks or vector geometries in distinct coordinate reference systems, and both training and evaluation often require spatially aware sampling and splitting strategies. TorchGeo is a PyTorch-based domain library that provides datasets, samplers, transforms and pre-trained models with the goal of making it easy to use geospatial data in machine learning pipelines. In this paper, we introduce a tutorial that demonstrates 1.) the core TorchGeo abstractions through code examples, and 2.) an end-to-end case study on multispectral water segmentation from Sentinel-2 imagery using the Earth Surface Water dataset. This demonstrates how to train a semantic segmentation model using TorchGeo datasets, apply the model to a Sentinel-2 scene over Rio de Janeiro, Brazil, and save the resulting predictions as a GeoTIFF for further geospatial analysis. The tutorial code itself is distributed as two Python notebooks: https://torchgeo.readthedocs.io/en/stable/tutorials/torchgeo.html and https://torchgeo.readthedocs.io/en/stable/tutorials/earth_surface_water.html.

Problem

• Earth observation
• machine learning
• geospatial data
• semantic segmentation
• TorchGeo

Innovation

• TorchGeo
• Earth observation
• geospatial machine learning
• semantic segmentation
• spatially-aware sampling