Vol 21, No 10 (2021) / Eassa

Automated classification technique for edge-on galaxies based on mathematical treatment of brightness data

Mohamed Eassa, Ibrahim Mohamed Selim, Walid Dabour, Passent Elkafrawy

Abstract

t Classification of edge-on galaxies is important to astronomical studies due to our Milky Way galaxy being an edge-on galaxy. Edge-on galaxies pose a problem to classification due to their less overall brightness levels and smaller numbers of pixels. In the current work, a novel technique for the classification of edge-on galaxies has been developed. This technique is based on the mathematical treatment of galaxy brightness data from their images. A special treatment for galaxies’ brightness data is developed to enhance faint galaxies and eliminate adverse effects of high brightness backgrounds as well as adverse effects of background bright stars. A novel slimness weighting factor is developed to classify edge-on galaxies based on their slimness. The technique has the capacity to be optimized for different catalogs with different brightness levels. In the current work, the developed technique is optimized for the EFIGI catalog and is trained using a set of 1800 galaxies from this catalog. Upon classification of the full set of 4458 galaxies from the EFIGI catalog, an accuracy of 97.5% has been achieved, with an average processing time of about 0.26 seconds per galaxy on an average laptop.

Keywords


techniques: image processing — methods: data analysis — galaxies: formation

Full Text:

PDF


DOI: https://doi.org/10.1088/1674-14527/21/10/264

Refbacks

  • There are currently no refbacks.