Radio Galaxies detection and characterization using deep learning techniques

Abstract

Future radio telescopes will generate data volumes that are increasingly difficult to analyse using traditional statistical methods, motivating the adoption of machine-learning techniques. In this work, we present YOLO-Chars (YOLO-based Detection and Characterisation of Radio Sources), a two-stage deep-learning framework for the automated detection and characterisation of radio galaxies in survey images. The framework is developed and evaluated using the Square Kilometre Array Science Data Challenge 1 (SKA SDC1) dataset. In the first stage, customised YOLO-based multi-scale detection models are used to localise compact and extended sources across large sky maps. In the second stage, a dedicated source-characterisation network estimates the physical properties of the detected sources. We focus on three key parameters: flux density, angular size, and position angle. Our results show that YOLO-Chars achieves competitive detection and characterisation performance on the SKA SDC1 benchmark, demonstrating its potential as a scalable framework for next-generation radio continuum surveys.