Automatic Satellite Identification in Digital Images

Research output: Contribution to journalArticlepeer-review

Abstract / Description of output

Using a digital camera on the roof of the Royal Observatory, Edinburgh, images are continuously captured of the sky on a nightly basis to image satellites which pass overhead. Such satellites are seen in images as short lines since satellites move much quicker in the sky than stars, and up to 100 satellites can be imaged in one night. Using these images, software was developed (GitHub repository for code & documentation: https://github.com/Jack6228/STM_Final, or DOI: https://doi.org/10.5281/zenodo.7304283.) to use known catalogs of artificial objects which orbit the Earth to determine which satellites can be seen in each image. This was achieved by comparing the expected positions of satellites in the catalog to those observed by the camera to identify matches. Overall, the project was a success and satellites are able to be identified approximately 85% of the time, which could help track space debris in a cheaper manner in a world where space is becoming increasingly more polluted.
Original languageEnglish
JournalResearch Notes of the American Astronomical Society
Volume6
Issue number234
DOIs
Publication statusPublished - 10 Nov 2022

Fingerprint

Dive into the research topics of 'Automatic Satellite Identification in Digital Images'. Together they form a unique fingerprint.

Cite this