Noise-based Detection and Segmentation of Nebulous Signal: A review

摘要

Because of the rich dynamic history of internal and external processes, galaxies display a very diverse variety of shapes or morphologies. Added with their low surface brightness features this diversity can cause various systematic biases in their detection and photometry. This method imposes statistically negligible constraints on the to-be-detected targets. It is able to apply a sub-sky threshold (-0.5σ) to the image for the first time. This allows for very accurate non-parametric detection of the low surface brightness structure in the outer wings of bright galaxies or the intrinsically faint objects that remain wholly below the commonly used thresholds (> Iσ). NoiseChisel is the software we have created to apply this algorithm. The software infrastructure hosting NoiseChisel (GNU Astronomy Utilities) is also introduce here.