Plenty of data is being collected about asteroids. Typically, classification is outsourced to citizen science projects, volunteers. This is manual. Our proposed solution is an automated mechanism for processing photometric studies of Asteroids (i.e. Light curves) using pattern recognition. We will then present the analysis with an interactive, easy to understand data visualization.

This project is solving the Neuromorphic Studies of Asteroid Imagery challenge.


The heavy lift in understanding asteroids is in the classification. We took existing classifications and use machine learning to train algorithms to automatically classify larger data sets. We also enabled a front end submission tool so that citizen scientists and backyard astronomers can participate in the effort to understand asteroids. We want to build a platform where anyone can upload their images of the night sky and utilize the power of our classifier. They will then get feedback based on whether they've discovered a new object, found an existing object, or whether they should continue the hunt! Citizens and scientists alike can now become asteroid heroes. Heroes can compete by capturing and submitting more images and earning points for their discoveries and identifications. The classifier brings together the citizen scientist and the professional scientific community with the common goal of discovering, identifying, and learning more about asteroids.

Alternate Link to Video:

To view our video embedded on this page: Enable all scripts for playing flash. Click the PLAY triangle button. Scroll to the lower right hand of the screen and click the icon to enable "full screen". This is the best viewing option.

Project Information

License: Academic Free License 3.0 (AFL-3.0)

Source Code/Project URL:


Asteroid Classification -


  • Korina Ysabel
  • Sergii Kostenko
  • Naoya Oishi
  • Naureen Sayani
  • alex huang
  • Mohib Hassan
  • Norihito Naka
  • Korina Ysabel