The TECLA project aims to develop innovative techniques for the detection,
recognition, and classification of coherent structures and the discrimination of signal from noise in astrophysical images.
We plan to study astrophysical images using statistical physics techniques, time-series analysis techniques, information-theoretic techniques, and artificial intelligence techniques already successfully used in the analysis of various types of complex systems. Our focus is on using spatially resolved data to identify the main structures of sources of astrophysical interest.
On the one hand, we will be interested in investigating the correlations between pixels in X-ray emission images of different metals. This will be done by building a correlation-based pixel network and subsequently using filtering techniques to identify the most significant correlations between pixels. The idea is that these correlations indicate the presence of relevant spatial structures.
Furthermore, we will exploit the possibility of using machine learning techniques to enhance the power of the methodologies previously investigated. We will leverage the existing image database available within the various UNIPA research groups involved in the project to train the model.
Finally, particular attention will be paid to the software implementation of the above-mentioned protocols. We plan to have two types of software. First, we will consider a C and Python implementation specifically designed for researchers who wish to run the software on their own datasets and computing equipment. Second, we will also consider a web-based implementation for outreach activities within the scientific community and schools.
TECLA originally involved the following researchers: S. Calderaro (UNIPA-DiFC), C. Fazio (UNIPA-DiFC), G. Lo Bosco (UNIPA-DMI), R. N. Mantegna (UNIPA-DiFC), G. Marsella (UNIPA-DiFC), S. Miccichè (UNIPA-DiFC, PI), M. Miceli (UNIPA-DiFC), M. L. Saladino (UNIPA-STEBICEF),
TECLA Advisory Board involved Prof. C Baccigalupi (SISSA) and Prof. P. Mazzotta (UNITOV).