E. Borsato, F. Fonzo, N. Bellucco, E. Iodice, E. M. Corsini, M. Spavone, S. Pasquato, C. Buttitta, M. Cantiello, M. D'Onofrio, M. Gullieuszik, A. La Marca, A. Moretti, A. Nucita, M. Paolillo, A. Pizzella, E. Portaluri, C. Tortora

Context. Ultra-diffuse galaxies (UDGs) lie at the extreme end of the size-luminosity distribution of low surface-brightness (LSB) galaxies. Their detection and characterization require deep imaging and reliable source detection techniques that can handle low signal-to-noise ratios and severe source blending. Aims. We aim at improving the detection and characterization of the LSB galaxies and UDG candidates in different environments. To this end, we have developed a new automated detection Python-based tool, named ATTILA. Methods. We use deep g- and r-band imaging from the VST Early-type GAlaxy Survey (VEGAS), covering the central region of Hydra I and three new additional fields. Sources are identified combining tiling processing, source detection, and iterative deblending. The structural parameters are derived through surface brightness profile analysis and Sérsic modelling. Cluster membership is determined using the early-type galaxies colour-magnitude relation. Results. We identify 24 new UDGs, doubling the known population in the Hydra-I cluster to 48, consistent with expectations from halo mass scaling relations, and 92 additional LSB galaxies. In real data, ATTILA recovers more than 80% of previously known LSB galaxies and significantly improves the automated detection rate relative to standard methods. Conclusions. By improving the recovery of faint and diffuse sources while mitigating blending and contamination effects, ATTILA enables a more complete census of the LSB galaxy population, including UDGs.