| Authors : | Atharva Bagul, Disha Sawant, Atharva Pathak, Atreyee Saha, Snehal Sadalage, Sudhanshu Barway, Ajit Kembhavi, Ashish Mahabal
|
| Abstract : | We present galaxy morphology identification from images observed by the HSC Subaru survey. Each galaxy may have several features, like spiral, dust lane, bar, indications rings, tidal tails etc. We aim to catalogue one million galaxies with such features. We aim to involve citizens to overcome the difficulty of examining the vast numbers of galaxy images needed to generate the catalogue. We have set up an interactive online platform for citizens to participate and detect features in the galaxy images. A pilot program has been developed and tested online with the help of amateur astronomers, college students, homemakers, senior citizens, and others. We also include experts from the field and use their responses for the ambiguous cases which arise from the rigorous statistics applied to citizens' responses. We use a condition-based statistical model to identify features of the galaxies, which includes grouping the answers and fitting a binomial distribution to them to study the probability of future reactions on different features of galaxies. Based on this analysis, we obtain the minimum number of citizens required to identify each feature in a galaxy with some confidence value (like 95%). We plan to develop an artificial intelligence model for the identification of morphological features using the data obtained through the One Million Galaxies program. The project would serve two goals: spreading knowledge and awareness among amateur astronomers and building a full-fledged catalogue for scientific purposes. |
