This tutorial is part of a series where I will explore deep learning applications in several areas, each with its own project.
The focus here is on learning deep learning and image classification.
We could have chosen to classify cats, dogs or flowersbut for this project, we chose to work with satellite images as they provide a unique and exciting challenge.
For this we will use the EuroSAT_RGB data setgenerously shared by my friend Blanchon on face hugging. Special thanks to him for making this data set available to the community, enabling projects like this.
In this case, the goal is to teach a model to categorize different types of environments (such as forests, rivers or urban areas) based on satellite images.
However, the potential applications of this type of technology go far beyond satellite imagery, including:
- medical imagehow to detect tumors on x-rays or MRIs.
- autonomous vehiclesallowing them to recognize roads, obstacles and pedestrians.
- Retailfor the analysis of customer behavior through video sequences.
