Interactive classification of satellite imagery. By using the human capacity of pattern recognition we aim to create an application that will help identifying natural events through tags.

This project is solving the Volcanoes, Icebergs, and Cats from Space challenge.


The tagEarth Project proposes, a network of colaboration where users will help us identifying and tagging Satellite imagery provided by NASA’s Earthexplorer data base. The main objective of the Project is to optimize the analysis speed of NASA’s Satellite imagery. By dividing the problem of Earth changes detection in small problems, users can help solving them in a fun and interactive manner. After combining a meaningful sample size, we can achieve a reliable data.

Inspiration for our challenge:


Being humans capable of images pattern recognition we could obtain a feedback of natural events going on in our planet. By storing the historic classification of picture by tags, we expect to possibly compare those tags and detect changes. Later use this information to minimize, or even anticipate forthcoming of natural disasters.

In a future step, we plan is to transform this idea in a nice mobile app, with great design and a positive feedback system to help users not only to keep their motivation, but also help them to improve detection skills, raising the bar of our analysis accuracy.

Our goal is provide to the user a pleasant emerging experience where one could get to learn more about earth in several fields of knowledge, help protecting the planet while having the sensation of traveling around the globe.

Making off:


Project Information

License: GNU General Public License version 3.0 (GPL-3.0)

Source Code/Project URL: https://github.com/acbarbosa/tagEarth



  • Alexandre Barbosa
  • Thiago Abrahão Pereira
  • Michel Kluger
  • Alexandre Barbosa