Satellite Image Classification Using Deep Learning | Written by Leo Anello | In the middle

This tutorial is part of a series where I will explore deep learning applications across different domains, each with its own project.

The focus here is on deep learning Reading and classifying images.

We could choose to separate cats, dogs, or flowersbut for this project, we chose to work with it satellite images as they offer a unique and exciting challenge.

In this case, we will use the EuroSAT_RGB datasetgenerously shared by my friend Blanchon at Hugging Face. Special thanks to him for making this dataset available to the public, which enables projects like this one.

In this case, the goal is to teach the model to distinguish different types of terrain – such as forests, rivers, or urban areas – based on a satellite image.

However, the potential uses of this type of technology go beyond satellite imagery, including:

• Medical imagingsuch as finding tumors on X-rays or MRIs.
• Private carsenabling them to see roads, obstacles, and pedestrians.
• Sellingto analyze customer behavior via video.