Skip to content

Moonshot idea to train ML model on synthetically generated data to detect and trace real asteroids.

License

Notifications You must be signed in to change notification settings

LazaUK/MachineLearning-AsteroidHunting

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

45 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Asteroid Hunting with synthetic data

In some sitiations, data may be rare or hard to find. For example, NASA’s discovery statistics database as of 10th of March 2023 contained records of 31,519 Near Earth Objects (NEO) collected over the period of 20 years. Still, none of them even closely matched characteristics of asteroid `Oumuamua, that was discovered in 2017.

For this project, with a great help and support from my colleagues (Pedro Urbina, Shannon Monroe, Jon Hanzelka and Luke Tiday), I was able to generate synthetic asteroid images using some of the best 3D modelling platforms. The object detection model was then trained in Azure Custom Vision, before being exported into TensorFlow format. Export process also generated a template class to perform object (asteroid) detection that was then utilised in a custom asteroids.py inference code1.

  • You can check this YouTube video for an output of ML model, tested on the video stream of an asteroid's fly-by.

Img alt text

Img alt text

Footnotes

  1. This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

    This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

    You should have received a copy of the GNU General Public License along with this program. If not, see https://www.gnu.org/licenses/.

  2. This short video shows an artist's impression of 2004 EW95, the first carbon-rich asteroid confirmed to exist in the Kuiper Belt and a relic of the primordial Solar System. The video shows a fly-by of the enigmatic asteroid as it tumbles through the icy outer reaches of the Solar System due to past interactions with migrating planets.

    More information and download options: http://www.eso.org/public/videos/eso1.

    Credit: ESO/M. Kornmesser

About

Moonshot idea to train ML model on synthetically generated data to detect and trace real asteroids.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages