Revealing Astronomical and Archaeological Information from Satellite Imaging Data (AA - Reveal)
As part of our astronomical research we are developing computer methods to make automated searches for very low signal to noise sources.
These include ultra-diffuse galaxies, features resulting from galaxy tidal interactions and inter-galactic light. We are using pixel clustering algorithms, distributed computing and machine learning to identify objects of interest.
These are computer intensive methods particularly as our intention is to apply them to new state-of-the-art sky surveys such as LSST and EUCLID. We have also been collaborating with archaeologists on the use of satellite imaging of the Earth and how it might be used to discover new archaeological sites.
The problem has many similarities to our astronomical work. The proposed project is to look at the further development of the astronomical software and to particularly consider its application to archaeological surveys. In particular we will use the very high resolution multi-band WorldView-3 data and concentrate on Roman sites such as Hadrian’s wall, Carnuntum and Portus/Ostia.
Supervisor: Professor Jonathan Davies