In recent years, the scientific community has become increasingly interested in the study of Unidentified Aerial Phenomena (UAP), otherwise known as UFOs. With the release of the UFO Report by the ODNI in 2021, the study of these phenomena has transitioned from a clandestine affair to a scientific pursuit. In an effort to detect and understand possible “visitors” from beyond our planet, researchers from the University of Strathclyde have proposed the use of hyperspectral imaging and Machine Learning.

Led by Professor Massimiliano Vasile, a team of researchers from the University of Strathclyde and the Fraunhofer Center for Applied Photonics in Glasgow has explored the potential of hyperspectral imaging paired with machine learning in the identification and classification of space objects. Their findings are detailed in a paper titled “Space Object Identification and Classification from Hyperspectral Material Analysis,” which is currently under review for publication in Scientific Reports.

Hyperspectral imaging involves collecting and analyzing data from across the electromagnetic spectrum to identify different objects or materials. This imaging technique combined with machine learning algorithms could help narrow down the search for possible technosignatures by eliminating false positives caused by human-made debris objects in space. By analyzing the spectra of UAP, researchers can understand their material composition and motion, providing valuable insights into their nature.

To create a data processing pipeline for UAP images, the researchers propose the need for a diverse dataset of time-series spectra of space objects, including satellites and debris. Since such a comprehensive dataset is not readily available, the team created numerical physics simulation software to generate training data for the machine learning models. They also employed a two-pronged approach, using both machine learning and mathematical regression analysis, to associate a spectrum with a set of materials generating it.

The team conducted various tests, including laboratory tests with a mockup of a satellite and observations using a telescope. The results of these tests were promising and provided valuable insights for future study. Ultimately, the goal is to enhance our understanding of UAP through advanced imaging techniques and AI, potentially shedding light on their origin and purpose.

Sources:
– “Space Object Identification and Classification from Hyperspectral Material Analysis” by M. Vasile et al. (2023) – arXiv preprint
– “Intelligent characterization of space objects with hyperspectral imaging” – Acta Astronautica, February 2023

The post Using Hyperspectral Imaging and Machine Learning to Study Unidentified Aerial Phenomena appeared first on TS2 SPACE.