Radio Sources Segmentation and Classification with Deep Learning

Abstract

Modern large radio continuum surveys have high sensitivity and resolution, and can resolve previously undetected extended and diffuse emissions, which brings great challenges for the detection and morphological classification of extended sources. We present HeTu-v2, a deep learning-based source detector that uses the combined networks of Mask Region-based Convolutional Neural Networks (Mask R-CNN) and a Transformer block to achieve high-quality radio sources segmentation and classification. The sources are classified into 5 categories: Compact or point-like sources (CS), Fanaroff-Riley Type I (FRI), Fanaroff-Riley Type II (FRII), Head-Tail (HT), and Core-Jet (CJ) sources. HeTu-v2 has been trained and validated with the data from the Faint Images of the Radio Sky at Twenty-one centimeters (FIRST). We found that HeTu-v2 has a high accuracy with a mean average precision ($A{P}_{@50:5:95}$) of 77.8%, which is 15.6 points and 11.3 points higher than that of HeTu-v1 and the original Mask R-CNN respectively. We produced a FIRST morphological catalog (FIRST-HeTu) using HeTu-v2, which contains 835,435 sources and achieves 98.6% of completeness and up to 98.5% of accuracy compared to the latest 2014 data release of the FIRST survey. HeTu-v2 could also be employed for other astronomical tasks like building sky models, associating radio components, and classifying radio galaxies.

arXiv comment20 pages, 15 figures, Accepted by Astronomy and Computing

References in corpus (37)

• Observational Evidence of AGN Feedback
• LOFAR: The LOw-Frequency ARray
• The Sixteenth Data Release of the Sloan Digital Sky Surveys: First Release from the APOGEE-2 Southern Survey and Full Release of eBOSS Spectra
• The Murchison Widefield Array: the Square Kilometre Array Precursor at low radio frequencies
• The LOFAR Two-metre Sky Survey - I. Survey Description and Preliminary Data Release
• Active Galactic Nuclei: what's in a name?
• On the evolution of star forming galaxies
• EMU: Evolutionary Map of the Universe
• The Last of FIRST: The Final Catalog and Source Identifications
• Australian Square Kilometre Array Pathfinder: I. System Description
• Compact continuum source-finding for next generation radio surveys
• Image Data Augmentation for Deep Learning: A Survey
• ProFound: Source Extraction and Application to Modern Survey Data
• Radio Galaxy Zoo: host galaxies and radio morphologies derived from visual inspection
• Giant radio galaxies in the LOFAR Two-metre Sky Survey-I
• FRII{\sl{CAT}}: A FIRST catalog of FR~II radio galaxies
• Radio Galaxy Zoo: ClaRAN - A Deep Learning Classifier for Radio Morphologies
• Deblending and Classifying Astronomical Sources with Mask R-CNN Deep Learning
• FRICAT: A FIRST catalog of FRI radio galaxies
• Radio galaxies of the local universe: all-sky catalog, luminosity functions, and clustering
• Source-finding for the Australian Square Kilometre Array Pathfinder
• Morphological Annotations for Groups in the FIRST Database
• DeepSource: Point Source Detection using Deep Learning
• Discovery of a High-Redshift (z=0.96) Cluster of Galaxies Using a FIRST Wide-Angle Tailed Radio Source
• MIGHTEE: Are giant radio galaxies more common than we thought?
• First release of Apertif imaging survey data
• Radio Source Extraction with ProFound
• Deep Learning Based Detection of Cosmological Diffuse Radio Sources
• Status and progress of China SKA Regional Centre prototype
• Bent-Tailed Radio Sources in the Australia Telescope Large Area Survey of the Chandra Deep Field-South
• Deep Learning Assisted Data Inspection for Radio Astronomy
• Automated detection of extended sources in radio maps: progress from the SCORPIO survey
• A New Catalog of Head Tail Radio Galaxies from the VLA FIRST Survey
• Artificial intelligence for celestial object census: the latest technology meets the oldest science
• CAESAR source finder: recent developments and testing
• Radio source-component association for the LOFAR Two-metre Sky Survey with region-based convolutional neural networks
• Parallel implementation of w-projection wide-field imaging

Cited by in corpus (7)

• Self-supervised contrastive learning of radio data for source detection, classification and peculiar object discovery
• Identification of 4876 Bent-Tail Radio Galaxies in the FIRST Survey Using Deep Learning Combined with Visual Inspection
• Classification of Radio Sources Through Self-Supervised Learning
• A Machine Learning made Catalog of FR-II Radio Galaxies from the FIRST Survey
• Identifying Radio Active Galactic Nuclei with Machine Learning and Large-Area Surveys
• Transformer-Based Source Detection and Morphological Classification in LOFAR Deep-Field Continuum Images
• Radio Galaxy Zoo: Morphological classification by Fanaroff-Riley designation using self-supervised pre-training