BRAHMa: Bar Recognition And Hatching using Machine learning -- a framework for oriented bounding box detection of galactic bars

Abstract

Galactic bars are fundamental components in studies of dark matter, galaxy evolution, and secular dynamics. However, current methods for measuring bar properties suffer from significant limitations. Manual and crowdsourced catalogs are not scalable to next-generation surveys, while recent machine learning (ML) approaches based on pixel-wise segmentation are computationally intensive and require complex post-processing to extract physical parameters. In this paper, we introduce BRAHMa (Bar Recogn…