CIRASA

Innovative developments in data processing, archiving, analysis and visualization are nowadays unavoidable to deal with the data deluge expected in next-generation facilities for radio astronomy, such as the Square Kilometer Array (SKA) and its precursors (MeerKAT, ASKAP, LOFAR, MWA). In this context, integration of source extraction and analysis algorithms into data visualization tools could significantly improve and speed-up the cataloguing process of large area surveys, boosting astronomer productivity and shortening publication times. This project aims to address some of the highlighted challenges through the realization of a visual analytic platform, dubbed CIRASA (C​ollaborative and ​Integrated platform for ​Radio ​ Astronomical ​Source ​ Analysis), integrating state-of-the-art tools developed in INAF. The main components of the platform are the visualization client (VLVA) interfacing, through an authentication layer, with a series of services for source extraction, classification and analysis, image/catalog search and cutout (VLKB), deployed in a distributed computing infrastructure. A high-level architecture diagram of the proposed platform showing the main software components and expected data flows is shown below.

The proposed platform should reach the following objectives:

[O1] Integration of existing compact/extended source finders for radio images (continuum or spectral line) into the CIRASA platform, exploiting the peculiarity of each of them and possibly combining their outputs to improve detection capabilities and source measurement accuracy. Finders developed and maintained with INAF contribution (Caesar, ​Cutex, Sofia) will be integrated with higher priority with respect to other finders (e.g. Selavy,​ ​PyBDFS, ​Aegean);
[O2] Development and integration of new source classifiers, employing deep learning methodologies, to enhance the performances of traditional source finders and to support creation of added-value catalogues;
[O3] Upgrading ​VLVA client with source visualization and interactive/automated annotation functionalities;
[O4] Preparation of curated training datasets for source classification scopes, including single- or multi-channel image data (radio and infrared primarily) or source feature tables extracted from images or external catalogues;
[O5] Supporting source finding and classification processing on a distributed computing infrastructure integrated with the VLVA; [O6] Test​ing platform functionalities with end-to-end catalogue analysis on real SKA precursor survey data.