Improving Gamma-ray Source Search with Image Processing

TL;DR

This paper proposes an image processing-based method to improve gamma-ray source searches, aiming to reduce computational time and enhance detection of faint sources in HAWC data analysis.

Contribution

It introduces a novel image processing technique to seed source locations, streamlining the likelihood-based search process for gamma-ray sources.

Findings

Reduces computational time for source detection.

Improves detection sensitivity for faint sources.

Streamlines the source search pipeline.

Abstract

The analysis of HAWC data is done using a likeihood-based systematic multi-source search procedure utilizing the threeML software package and the HAL Plugin. This approach was inspired by the extended source search described in the Fermi-LAT Extended Source Search Catalog. The pipeline to search for point sources and extended sources within the region of interest (ROI) is described in the recent HAWC papers. This procedure is computationally intensive and often requires multiple days to produce a final model for a region. Often this approach misses fainter sources, which need to be added manually later. This blind search could be complemented by providing a method to seed source locations, which can be assessed and evaluated by likelihood analysis, thereby significantly reducing the computational time and resources spent on finding a model.